Abstract
Myo-inositol oxygenase (MIOX) is the first enzyme in the inositol route to ascorbate (L-ascorbic acid, AsA, vitamin C). We have previously shown that Arabidopsis plants constitutively expressing MIOX have elevated foliar AsA content and displayed enhanced growth rate, biomass accumulation, and increased tolerance to multiple abiotic stresses. In this work, we used a combination of transcriptomics, chromatography, microscopy, and physiological measurements to gain a deeper understanding of the underlying mechanisms mediating the phenotype of the AtMIOX4 line. Transcritpomic analysis revealed increased expression of genes involved in auxin synthesis, hydrolysis, transport, and metabolism, which are supported by elevated auxin levels both in vitro and in vivo, and confirmed by assays demonstrating their effect on epidermal cell elongation in the AtMIOX4 over-expresser plants. Additionally, we detected up-regulation of transcripts involved in photosynthesis that was validated by increased efficiency of the photosystem II and proton motive force. We also found increased expression of amylase leading to higher intracellular glucose levels. Multiple gene families conferring plants tolerance to cold, water limitation, and heat stresses were found to be elevated in the AtMIOX4 line. Interestingly, the high AsA plants also displayed up-regulation of transcripts and hormones involved in defense including jasmonates, defensin, glucosinolates, and transcription factors that are known to be important for biotic stress tolerance. These results overall indicate that elevated levels of auxin and glucose, and enhanced photosynthetic efficiency in combination with up-regulation of abiotic stresses response genes underly the higher growth rate and abiotic stresses tolerance phenotype of the AtMIOX4 over-expressers.
Introduction
L-Ascorbic acid (AsA) also known as vitamin C is important for human, animal, and plant health. In animals, vitamin C helps in collagen formation, decreases cholesterol content, boosts immune function, and improves wound healing. Vitamin-C deficiency on the other hand leads to scurvy. Humans and other primates are unable to produce vitamin C, so they depend on fruits and vegetables to satisfy their need for this vitamin (Leong 2017; Figueroa-Mendez and Rivas-Arancibia 2015; Weaver et al., 2014). In plants, ascorbate acts as a redox buffer and a cofactor of enzymes involved in various biochemical processes. Ascorbate also participates in vitamin E regeneration and acts as a precursor of organic acids including theronic and tartaric acids (Debolt et al., 2006; Gallie et al., 2013; Gest et al., 2013). Ascorbate modulates the cell cycle, flowering time, and signal transduction affecting plant growth and development (Ortiz-Espín et al., 2018). Vitamin-C deficient (vtc) mutants have lower AsA, slower growth, and show early senescence (Conklin et al., 1999; Dowdle et al., 2007; Pastori et al., 2003; Kerchev et al., 2011).
Despite the importance of vitamin C for human health it was until 1998 when a pathway for its biosynthesis in plants was revealed (Wheeler et al., 1998). Two additional routes contributing to the formation of AsA, the L-gulose and D-galacturonate pathways, were proposed in 2003 (Agius et al., 2003; Wolucka and Montagu 2003) and an additional route that uses myo-inositol as a precursor was proposed by Lorence et al (2004) (Figure 1).
Ascorbate is the most abundant antioxidant in plants present at concentrations orders of magnitude higher than others. Ascorbate scavenges reactive oxygen species (ROS) and protects plants from abiotic and biotic stresses by maintaining ROS homeostasis (Foyer and Noctor 2011). Plants with elevated AsA content have been developed by over-expressing biosynthetic genes and master regulators (reviewed in Yactayo-Chang et al., 2018). Plants with lower vitamin C (vtc mutants) are sensitive to high light, heat, osmotic, and oxidative stresses (Conklin et al., 2013; Cho et al., 2016; Tóth et al., 2011), but are resistant to Pseudomonas syringae infection due to increased expression of pathogen-related (PR) genes via reduced glutathione accumulation (Pavet et al., 2006). Contrastingly, enhanced resistance of low ascorbate A. thaliana to P. syringae is due to higher ROS priming with increased salicylic acid (SA) and enhanced expression of SA induced PR proteins (Mukherjee et al., 2010). Plants with elevated AsA on the other hand are tolerant to salt, cold, heat, osmotic, oxidative, high light and water limitation stresses (reviewed in Yactayo-Chang et al., 2018).
Ascorbate acts as cofactor for the thioglucosidase degradation enzyme, myrosinase that produces toxic chemicals against herbivores and pathogens (Burmeister et al., 2000), and there are reports indicating that AsA can also acts as an antioxidant agent and protect caterpillars from ROS generated from the oxidation of ingested polyphenols (Barbehenn et al., 2002). Other work reported that when the peach potato aphid (Myzus persicae) was allowed to feed on potato leaves with normal AsA content, the aphid grew twice the size of its counterpart growing on leaves with low AsA content, indicating that elevated AsA may aid in aphid growth (Kerchev et al., 2013). When Arabidopsis roots were infected with the nematode Heterodera schachtii, MIOX played a significant role in the development of syncytium and female nematodes. The study of metabolites in roots containing syncytia showed that, a decrease in the level of myo-inositol was responsible for syncytium development (Siddique et al., 2009, Siddique et al., 2014). These examples highlight the species-specific and context-dependent versatile role of AsA during biotic stress.
Arabidopsis plants constitutively expressing enzymes in the myo-inositol pathway: myo-inositol oxygenase, glucuronate reductase, or L-gulono-1,4-lactone oxidase (GLOase, a.k.a. GulLO) have 1.5-3 fold elevated AsA content compared to wild type (WT) control (Radzio et al., 2003; Lorence et al., 2004; Tóth et al., 2011; Yactayo-Chang 2011). Using manual phenotyping we and others have showed that Arabidopsis MIOX4 and GulLO over-expressers also display enhanced biomass accumulation of both aerial and underground tissues, and increased tolerance to cold, heat, salt, and pyrene, an environmental pollutant (Lisko et al., 2013; Tóth et al., 2011). Using high throughput phenotyping approaches we have quantified the enhanced growth rate and biomass accumulation of the high AsA line (Lisko et al., 2013; Lisko et al., 2014; Yactayo-Chang et al., 2018). In this study we report our findings on the diverse mechanisms underlying the enhanced growth rate, biomass accumulation, and broad abiotic stress tolerance of the high AsA line using a combination of transcriptomics (RNA-seq), RT-qPCR, liquid chromatography-tandem mass spectroscopy (LC-MS/MS), cell biology, confocal microscopy, and physiological measurements (Figure 2).
Materials and methods
Plants and growth conditions
In this study, we used Arabidopsis thaliana (A. thaliana) wild type accession Columbia (ABRC stock CS-60000), single-insertion homozygous line over-expressing AtMIOX4 (AtMIOX4 L21, Yactayo-Chang 2011), a vitamin-C defective mutant (vtc1-1) (Conklin et al., 1999) and the auxin sensor line (R2D2) (Liao et al., 2015). Seeds were sterilized with 70% (v/v) ethanol for 5 min followed by 50% (v/v) sodium hypochlorite with 0.05% (v/v) Tween-20 for 10 min. Then, seeds were washed 5-6 times with sterile distilled water. Plants were grown in Murashige and Skoog (MS) media pH 5.6 for 12 days after stratification for 3 days at 4°C. Seedlings from MS plates were transferred to soil (Promix PGX; Premier Horticulture) and grown in a growth chamber at 23°C, 65% humidity, 165-210 μmol/m2/s light intensity and 10:14 h photoperiod.
Southern blot
Genomic DNA of the AtMIOX4 OE line was isolated using the CTAB method (Murray and Thompson 1980). Briefly, 4-5 g of fresh tissue was grinded in a mortar and pestle cooled with liquid nitrogen. Then, 20 ml of CTAB buffer (1 M Tris-HCl pH 8.0, 5 M NaCl, 0.5 M EDTA, 2% CTAB, and 1% (v/v) β-mercaptoethanol) were added, vortexed and incubated at 65°C for 20 min. After incubation, 20 ml of chloroform: isoamyl alcohol 24:1 (v/v) were added. The lysates were gently mixed and centrifuged at 4,000xg at 4°C for 20 min. The upper aqueous phase was transferred to new tube and, total DNA was extracted and precipitated with 3M NaAc pH 5.2. The precipitated DNA was dissolved in TE buffer. Five micrograms of extracted DNA were digested with AtMIOX4:pBIB-Kan with SacI or KpnI. Digested DNA was separated on 0.8% (w/v) agarose 1X TAE gel and electro blotted onto a positively charged nylon membrane (Roche), using 1X TBE as transfer buffer. The UV-cross-linked membrane was then pre-hybridized at 65°C for 15 min in Church buffer (sodium phosphate buffer containing, 0.5 M Na2HPO4 pH 7.2, 20% (w/v) SDS, 1 mM EDTA, and 1% BSA), and hybridized overnight at 65°C with 50 ng of DNA probe each labeled with 50 μCi[32P]-dCTP. The probes were prepared using Prime-It® Rm T Random Primer Labeling kit (Stratagene). The probe for AtMIOX4:pBIB-Kan was amplified with npt-II forward (5’-AGA GGC TAT TCG GCT ATG AC-3’) and npt-II reverse (5’-AGC TCT TCA GCA ATA TCA CG-3’) primers at 55°C annealing temperature. After hybridization, the blots were washed with phosphate buffer (0.5 M Na2HPO4, 20% (w/v) SDS, 1 mM EDTA, and 33 mM NaCl pH 7.2 at 65 °C. Photographs were taken by exposing membranes to film for 2-3 days at −80°C (Biomax XAR, Kodak).
RNA extraction and sequencing
Three biological replicates of leaf samples were collected when plants completed the vegetative growth stage (stage 5.0 according to Boyes et al., 2001). Total RNA was purified using the RNeasy kit (Qiagen) following the manufacturer’s instructions. RNA quality and quantity were assessed with an Experion system using the Experion™ RNA HighSens chips (Bio-Rad). RNA samples were submitted to the Core Facility at Michigan State University for sequencing. Six libraries were prepared and sequenced using manufacturer recommendations with a Hi-Seq Illumina 2500 system. Libraries were prepared using the Illumina TruSeq Stranded mRNA Library Preparation Kit. After QC and quantitation the libraries were combined into a single pool for multiplexed sequencing. The pool was loaded on two lanes of an Illumina HiSeq 2500 Rapid Run flow cell (v2). Sequencing was performed in a 1×50bp single read format. Base calling was done by Illumina Real Time Analysis (RTA) v1.18.64 and output of RTA was de-multiplexed and converted to FastQ format with Illumina Bcl2fastq v1.8.4. Sequencing data used in this study (#SRA accession:PRJNA515221) will be available upon publication.
Bioinformatics analysis of raw RNA-Seq data
Quality analysis of raw data was performed using FastQC v0.11.7 (Andrews 2014). Trim Galore v0.4.3.1 was used for trimming the Illumina universal adapter sequences and removal of reads less than 40 bp in length (Krueger 2015). TopHat v 2.1.1 was used for aligning the reads with the whole genome sequence of A. thaliana genome (TAIR10) (https://www.arabidopsis.org/) using maximum edit distance 2.0, library type unstranded, maximum intron length 10,000 bp, minimum intron length 50 bp and default parameters (Kim et al., 2013). Cufflinks (v2.2.1) and Cuffmerge (v3.0) algorithms were used next for transcript compilation and gene identification using A. thaliana gene annotation TAIR10_gff3/TAIR10_GFF3_genes_transposons.gff (https://www.arabidopsis.org/). Differential expression analysis was performed using the CuffDiff (v7.0) algorithm. The CummeRbund algorithm R package (v2.22.0) was used for data visualization (Trapnell et al., 2012). Differentially expressed (DE) transcripts were selected based on an absolute value of log2 ratio ≥1.5 with FDR adjusted p-value ≤ 0.05). False discovery rate correction (FDR) was performed using the Benjamini Hochberg method. Blast2GO (Conesa and Gőtz, 2008) and DAVID v6.8 (Huang et al., 2009) were used for gene ontology and functional enrichment analysis with a false discovery rate correction using the Benjamini-Hochberg method with FDR adjusted p-value ≤ 0.05. MapMan 3.5.1R2 (Usadel et al., 2012Vitamin-C deficient) was employed for visualizing up-and down-regulated genes affecting metabolism in plants mapping with Ath_AGI_LOCUS_TAIR10_Aug2012 (https://www.arabidopsis.org/).
Real-time quantitative PCR (RT-qPCR) for gene expression analysis
Real-time quantitative PCR (RT-qPCR) was used for biological validation of the RNA-seq data using reference genes previously published and following MIQE guideline (Czechowski et al., 2005; Bustin et al., 2008). Three biological replicates of leaf samples were collected when plants completed the vegetative growth stage. Total RNA was purified using the Purelink™ RNA mini kit (Ambion, Life Technologies) following the manufacturer’s instructions. The RNA quality and quantity were assessed with an Experion system using the Experion™ RNA HighSens chips (Bio-Rad). Residual DNA in the sample was removed using DNA-free™ DNA removal kit (Invitrogen). Complimentary DNA (cDNA) were prepared by following manufacturer’s instruction (iScript select cDNA synthesis kit, Bio-Rad). Validated primers were obtained from AtRTPrimer database (Han and Kim 2006). The primers for photosystem II (PSII) related center transcripts and glutathione reductase (GRX480) were validated using the guidelines by Bustin et al., 2008. RT-qPCR was performed following manufacturer’s instruction (SsoFast™ EvaGreen® supermixes, Bio-Rad) in a CFX384TM real-time system (Bio-Rad). Normalization of transcripts counts was obtained using ACTIN2 and UBQ10 as internal controls. The relative expression of the genes was calculated using 2−δδt method (Livak and Schmittgen 2001). Three biological replicates and two technical replicates were used for RT-qPCR. Two-way student’s t-test was performed at 0.05 significance level for normalized expression value.
Di-nitro salicylic acid (DNS) assay for reducing sugars estimation
Six biological replicates of leaf samples were collected when plants completed the vegetative growth stage. Reducing sugars were estimated using di-nitro salicylic assay (DNS). Plant samples were ground and extracted twice in 1 ml of 0.5 mM sodium acetate buffer (pH 5.0) using Precellys 24 lyzer (Bertin Technologies) and 100 μl of DNS reagent (Sigma) was added and incubated in boiling water-bath (95°C) for 20 min. The reaction was stopped by adding 300 μl ddH2O. The 125 μl of reaction mixture was transferred into a 96-well clear plastic reading plate with 2 technical replicates and absorbance was measured at 540 nm in a 96 well plate reader (BioTek). Two-way student’s t-test was performed at a 0.05 significance level. The experiment was replicated twice with similar results.
Glucose oxidase-glucose peroxidase (GOD-POD) assay for estimation of glucose
Five biological replicates of leaf samples were collected when plants completed the vegetative growth stage. Plant samples were ground and extracted twice in 1 ml of 0.5 mM sodium acetate buffer pH 4.5 using Precellys 24 lyzer (Bertin Technologies). Sixty microliters of sample, 54 μl of GOD-POD enzyme mix (Sigma)) and 6 μl of O-dianisidine were mixed and incubated at 37°C for 30 min. The reaction was stopped by adding 60 μl of 10 M HCl. Absorbance was recorded at 540 nm in a 96-well clear plastic reading plate with two technical replicates in a 96 well plate reader (BioTek). Two-way student’s t-test was performed at 0.05 significance level. The experiments were replicated twice with similar results.
Photosynthetic efficiency measurements
Plants were grown on MS media in 23°C, 65% humidity, 165-210 μmol/m2/s light intensity and 10:14 h photoperiod after stratification for 3 days. Twelve days after germination plants were transferred to Arabidopsis soil (Promix PGX; Premier Horticulture LTEE). MultispeQ v1.0 (Kuhlgert et al., 2016) was used to measure the photosystem II efficiency (ΦII) and proton motive force (vH+) 21, 23, 26 and 28 days after germination. Fifteen biological replicates for each genotype were measured. Experiment was replicated twice with similar results. Two-way student’s t-test was performed at a 0.05 significance level for each day of measurement.
Hormones and metabolites quantification using LC-MS/MS
Leaf samples from 10 biological replicates for each genotype were collected and lyophilized in liquid nitrogen when plants completed the vegetative growth stage. Hormones extraction and analysis were performed at the University of Nebraska, Lincoln with modifications of Lan et al., 2014 and Westfall et al., 2016. Briefly, lyophilized samples were ground using a Tissuelyzer (Qiagen). Samples were spiked with deuterium–labelled internal standards and extracted in ice-cold methanol:acetonitrile (MeOH:CAN,1:1,v/v) for 2 min at 15 Hzs−1 then centrifuged at 16,000xg for 5 min at 4°C. Supernatants were pooled and evaporated in a Labconco Speedvac. The pellets after evaporation were redissolved in 200 μl of 30% MeOH. For LC separation, a ZORBAX Eclipse Plus C18 column (2.1 mm × 100 mm, Agilent) was used flowing at 0.45 ml/min. The gradient of the mobile phases A (0.1% acetic acid) and B (0.1% acetic acid/90% acetonitrile) was as follow: 5% B for 1 min, to 60% B in 4 min, to 100% B in 2 min, hold at 100% B for 3 min, to 5% B in 0.5 min. The LC system was interfaced with a Sciex QTRAP 6500+ mass spectrometer equipped with a TurboIonSpray (TIS) electrospray ion source. Analyst software (version 1.6.3) was used to control sample acquisition and data analysis. The QTRAP 6500+ mass spectrometer was tuned and calibrated according to the manufacturer’s recommendations. All hormones were detected using MRM transitions that were previously optimized using standards. The instrument was set-up to acquire in positive and negative ion switching. For quantification, an external standard curve was prepared using a series of standard samples containing different concentrations of unlabeled hormones and fixed concentrations of deuterium-labeled standards (D5-IAA, D5-tZ, D5-tZR, D2-JA, D6-ABA, D4-SA). Two-way student’s t-test was performed at 0.05 significance level.
Semi-quantification of in-vivo auxin
Crosses between AtMIOX4 L21 (♂) and auxin sensor (R2D2) (♀) were used for visualization and in-vivo semi-quantitative analysis of auxin. Plants were grown on MS media in 23°C, 65% humidity, 165-210 μmol/m2/s light intensity and 10:14 h photoperiod after stratification for 3 days. Twelve days after germination plants were transferred to Arabidopsis soil (Promix PGX; Premier Horticulture). AtMIOX4 L21 and R2D2 line were crossed and F1 generation were selected using kanamycin as a selectable marker and PCR using primers for MIOX4 and kanamycin resistance genes. The positive crosses were taken for the in-vivo auxin quantification. The positive crosses were grown on MS media horizontally and vertically for in-vivo auxin quantification of shoots and roots respectively. The plants were grown at 23°C, 65% humidity, 165-210 μmol/m2/s light intensity and 10:14 h photoperiod for 7 days. The seedlings were mounted in propidium iodide (10 μg/ml) for root scanning and in PBS for shoots scanning. The image acquisition and analysis of roots and shots were performed in Cytation™ 5 cell imaging multimode reader and Gen3.03 software (BioTek). Venus was excited at 488 nm and detected at 498-530 nm, ntdTomato was excited at 561 nm and detected at 571-630, propidium iodide was excited at 561 nm and detected at 571-700 nm (Liao et al., 2015).
Root length, hypocotyl length and hypocotyl cell length measurement
Plants were grown on plates positioned vertically on MS media in 23°C, 65% humidity, 165-210 μmol/m2/s light intensity and 10:14 h photoperiod after stratification for 3 days. For primary root length measurement plants were grown on 150-210 μmol/m2/s light intensity whereas for hypocotyl measurement plants were grown on dark condition in same growth chamber. Nine days after germination the primary root length and the hypocotyl length were measured with a caliper. Fifteen biological replicates for each genotype were used in this experiment and a two way student’s t-test was performed at 0.05 significance level.
For cell length measurements, hypocotyl of plants grown vertically on MS media on dark conditions were mounted in 1X PBS. The image acquisition and analysis was performed using Cytation™ 5 cell imaging multimode reader (BioTek) and cell length was measured using the Gen3.03 software (BioTek). Hypocotyl epidermal cell length was the average of 6 biological replicates for each genotype. The length of five cells was measured from each sample. Two-way student’s t-test was performed at a 0.05 significance level (Das et al., 2016).
Ascorbate measurements
Foliar tissue samples (40-60 mg) were collected at developmental stage 5.0 (Boyes et al. 2001) and immediately frozen in liquid nitrogen. The samples were grinded in a cold mortar and pestle using liquid nitrogen and adding 0.75 ml of fresh meta-phosphoric acid (6%) to the sample. The plant extracts were immediately transferred to a 1.5 ml amber tube and centrifuged at 14,000xg for 15 min at 4°C. For the measurement of reduced ascorbate, the supernatants were transferred to 1.5 ml amber tubes and kept on ice. There 0.95 ml of the meta-phosphoric acid buffer was added to a quartz cuvette and 50 μl of supernatant were added and mixed by inverting 2-3 times. The absorbance was recorded at 265 nm. Twenty μl of ascorbate oxidase were added, mixed, and absorbance was recorded at 265 nm. For the measurement of oxidized ascorbate, 0.95 ml of the meta-phosphoric buffer was added into a quartz cuvette and 50 μl of supernatant were added and mixed and absorbance was recorded at 265 nm. The mix was recovered in a 1.5ml amber tube and 1 μl DTT (200 mM) was added mixed and incubated at dark for 20minutes. The final absorbance was recorded at 260 nm. The reduced and oxidized ascorbate content was calculated using the extinction coefficient as [(Total change in absorbance)/14.3]*20*0.75(1/sample weights in grams). The total ascorbate content was calculated as the sum of reduced and oxidized ascorbate pools.
Quantification of H2O2accumulation
Foliar tissue samples (40-60 mg) were collected at developmental stage 5.0 and immediately frozen in liquid nitrogen. H2O2 was quantified using Amplex Red (Molecular Probes, Invitrogen) as previously described (Katano et al. 2017; Suzuki et al. 2015). Briefly, the plant samples were grinded using liquid nitrogen. The Amplex red reagent (50 mM sodium phosphate buffer pH 7.4, 50 μM Amplex red and 0.05 U/ml horse radish peroxidase) was added to grinded tissues and centrifuged at 13,000xg for 12 min at 4°C. After centrifugation, 450 μl of supernatant were transferred to a 1.5ml amber tube and incubated for 30 min in the dark at room temperature. The absorbance was recorded at 560 nm. The concentration of H2O2 was calculated using a standard curve obtained from 0, 0.5, 1, 3, 6 and 9 μM H2O2.
Methione-derived aliphatic glucosinolate HPLC analysis
When plants completed the vegetative growth stage, leaf samples (∼100 mg of DW) from ten biological replicates per genotype were collected and frozen in liquid nitrogen. Glucosinolates extraction was carried out with 80% MeOH and aryl sulfatase (Sigma). Recovered desulfoglucosinolates were lyophilized in Speedvac and redissolved in 500 μl of ultrapure water. Samples were analyzed in a reverse phase HPLC system (Dionex; Ultimate 3000) equipped with an AcclaimTM 300 C18 column (4.6 × 150 mm, 3 μm particle size, 300A°). The mobile phase was composed of acetonitrile (A) and 0.1% TFA in water (B). The gradient program was: 0–1 min, 1.5 % B; 1–6 min, 1.5–5 % B; 6– 8 min, 5–7 % B; 8–18 min, 7–21 % B; 18–23 min, 21–29 % B; 23–23.1 min, 29–100 % B; 23.1–24 min 100 % B, and 24.1–28 min 1.5 % B; flow 1.0 ml/min. Glucoerucin (Nacalai Tesque), 4-methylthiobutil glucosinolate (4MTB), a methionine-derived aliphatic glucosinolate, was analyzed. Detection was performed at 229 nm using a photodiode array. GlucoerucinUV spectra acquisition and analysis were carried following methods described by Grosser and Dam 2017 and Madsen et al., 2016.
Pathogen infection assay
High ascorbate (AtMIOX4 L21), low ascorbate (vtc2-1) and WT (CS-60000) A. thaliana plants were challenged with the virulent pathogen Pst DC3000 (Pseudomonas syringae pv tomato DC3000). Pst DC3000 was grown on King’s B medium for 2 to 3 days and resuspended on MgCl2 (10 mM) and Silwet L-77 (0.02%) in 105 colony forming unit (cfu)/ml. Two-week old seedlings grown on MS media were dip inoculated with bacteria and covered for 1 h and uncovered slowly. The amount of bacteria present in plants was analyzed 1 h after dipping (Day 0) and 3 d after dipping (Day 3). Three biological replicates were pooled and measured for bacterial growth as cfu for each genotype (Zhu et al., 2013). The experiment was replicated three times with similar results.
Results
After years of work we have detected gene silencing in the AtMIOX4 lines we initially published (Lorence et al., 2004). To continue studying the possible roles of constitutive expression of AtMIOX4 in plants we decided to develop new transgenic lines. The line used in this study was one of the new ones we developed. AtMIOX4 line 21 was chosen for this study because it is homozygous, harbors a single copy of the MIOX4 ORF and recapitulates the phenotype of the lines previously reported (Lorence et al., 2004) that includes elevated levels of AsA (Figure 3C), enhanced biomass and tolerance to multiple abiotic stresses relative to the wild type control (Lisko et al., 2014; Yactayo-Chang, 2011; Yactayo-Chang et al., 2018).
RNA-seq, assembly, and annotation
Foliar tissue from Arabidopsis AtMIOX4 L21 plants and wild type controls was collected at developmental stage 5.0. High quality RNA was used to make six cDNA libraries, three from each genotype. These libraries were sequenced using the Illumina HiSeq 2500 Rapid Run flow cell with 1×50 bp paired end format. A total of 200,736,389 raw reads were obtained. The percentage of adapters in samples was up to 3.1%. The Trim-Galore script was used to remove the reads <40 bp in all samples. After removal of adapters and low quality reads, 185,235,265 bp of high quality reads were obtained. Reads that achieved a high Phred score (>38) indicating good quality sequences were used for the assembly. To assess the sequencing quality, the reads were mapped to the A. thaliana genome (TAIR10) using TopHat (Table 1). The functional gene ontology annotation of transcripts was performed using the Blast algorithm. The quality and mapping of the RNA-Seq data obtained from Illumina sequencing was good for transcriptome analysis of the AtMIOX4 OE line.
MIOX4 is highly expressed in the AtMIOX4 line
Differentially expressed transcripts (DETs) in the AtMIOX4 OE line were identified by comparison with the transcriptome of WT control. Genes with an adjusted p-value FDR <0.05 and a log2 fold change value of 2 were assigned as differentially expressed. As a result, 947 up-regulated transcripts and 357 down-regulated transcripts were differentially expressed in the AtMIOX4 OE line compared to the control (Figure 3A). According to the functional annotation of DETs, most of the differential transcripts have been proven to be correlated to different biological processes related to developmental regulation and response to stresses in plants. The majority of up-regulated transcripts fall under the response to stimulus biological process. These include response to external stimulus, response to organic substance, response to endogenous stimulus, response to auxin stimulus, response to abscisic acid stimulus, response to jasmonic acid stimulus, response to stress, response to water deprivation, response to cold stress, and response to salt stress (Supplementary Table 4). On the other hand, the majority of down-regulated transcripts were grouped under biological processes as response to chemical stimulus, response to stimulus and response to stresses (Supplementary Table 5). We identified transcripts belonging to the precursors of three microRNAs (At3g63375, At2g46685, At2g10606), a transcript related with TSM1 (At1g67990), TRNG.1 (AtCg00100), and snoRNA (At1g32385, At1g20015) involved in cell regulation that were exclusively expressed in the AtMIOX4 OE line. Similarly, transcripts related to stress responses such as lipid transfer proteins related to pathogenesis (At4g22513, At3g51590, At4g28395, At1g66850, At5g01870), an extensin family protein (At5g22430), a protease inhibitor (At4g22513), a guard cell specific transmembrane protein (At3g14452), a defensing-like family member protein (At1g64195, At3g05727), a stay-green like protein (At4g11911), a cytochrome P450 family protein (At5g44620), a β-glucosidase 20 (At1g75940), a NAC domain containing protein (At5g41090), and hypothetical proteins (At3g29633, At5g53740, At1g68526, At1g68875, At4g16240, At4g16215, At3g15534) were among the top 25 up-regulated genes in the AtMIOX OE line (Table 2). On the other hand, development related genes such as a plant self-compatibility protein S1 family (At3g24065), an aspartyl protease family protein (At5g48430), a putative non-LTR retroelement reverse transcriptase (At2g07767; At3g49330), a plant invertase/pectin methylestarase inhibitor family protein (At5g12880; At3g17130), transmembrane proteins (At3g49070; At4g18425; At3g21550), GASA10 (At5g59845), ARGOs (At3g59900), IDL3 (At5g09805), RNS3 (At1g26820), LAC11 (At5g03260), AGP19 (At1g68725) and stress response genes such as PR1 (At2g14610), GATA19 (At4g36620), and MRD1 (At1g53480) were down-regulated in the AtMIOX4 OE line compared to the control (Table 3). Additionally, amino acid metabolism genes lysine-ketoglutarate reductase/saccharopine dehydrogenase functional protein (At4g33150), involved in lysine degradation and induced by jasmonate (JA), and arginase (At4g08870), involved in defense response and induced by methyl jasmonate (MeJA), were abundant in AtMIOX4 OE line compared to WT control (Supplementary Table 1-3; Figure 3A). The MIOX4 (At4g26260) gene was significantly over-expressed in the AtMIOX4 transcriptome compared to the WT control, as shown by RT-qPCR (Figure 3B).
We measured elevated foliar AsA content in the AtMIOX4 OE (1.63 fold) compared to the control (Figure 3C), confirming our previous reports that over-expression of AtMIOX4 leads to higher AsA content in Arabidopsis both in the absence and in response to abiotic stresses (Lorence et al., 2004; Lisko et al., 2013; Yactayo-Chang, 2011). The transcripts only expressed in the AtMIOX4 OE line might have multiple roles in plant physiology, such as development, the oxidation-reduction process, carbohydrate and glucosinolate metabolism, and defense response against fungus, which could be vital for development and the stress tolerance/resistance phenotype observed in the AtMIOX4 OE line. A selected number of the DET from each group were validated through RT-qPCR and those results are presented in the following sections.
Primary metabolism is up-regulated in MIOX4 OE
The DETs were classified into different categories using the MapMan software (Usadel et al., 2009). Based on our results, the DETs were associated to primary metabolism, photosynthesis, biotic and abiotic stress, oxidative stress and secondary metabolism. Among the significantly over-expressed genes in the AtMIOX4 OE line involved in primary metabolism, we identified transcripts related with sugar metabolism such as hexokinase like protein 1 (HKL1: At1g50460), α-amylase-like (AMY1: At4g25000), and cellulose synthase (At4g24000). In addition, transcripts related to sugar and starch metabolism were significantly up-regulated in AtMIOX4 OE line compared to the control (Figure 4A). Interestingly, a starch degrading enzyme, α-amylase (At4g25000), was significantly up-regulated in the AtMIOX4 transcriptome which was validated using RT-qPCR (Figure 4B). α-Amylase catabolizes starch and releases free glucose and dextrin. Therefore we proceeded to measure the intracellular pool of free glucose in the high AsA line using two different methods. As show in Figure 4C we found that the intracellular concentration of reducing sugars was significantly higher in the AtMIOX4 OE line (41.64 ± 7.61 μM/g FW) compared to controls (23.13 ± 2.02 μM/g FW). In support of this, the intracellular concentration of glucose in the AtMIOX4 OE line (31.92 ± 0.95 μM/g FW) was significantly higher compared to the one found in the control (19.61 ± 0.99 μM/g FW), as shown by a glucose specific enzymatic assay (Figure 4D). In summary, intracellular glucose concentration was 1.62 fold higher in AtMIOX4 OE line compared to WT control. This indicates that the AtMIOX4 OE line has elevated expression of α-amylase which leads to higher available intracellular glucose in these plants.
MIOX4 over-expression positively influences photosynthetic efficiency
Transcripts involved in light reactions of photosynthesis, as photosystem II related transcripts (PSBA-ATCG00020, PSBC-ATCG00280, PSBB-ATCG00680) and photosystem I related transcripts (PSAJ-ATCG00630, PSAC-ATCG01060, NDHG-ATCG01080) were significantly up-regulated in the AtMIOX4 OE line (Figure 5A). In support of this, over-expression of photosystem II center related transcript (PSBB-AtCg00680) was validated with RT-qPCR (Figure 5B). To determine if the increased gene expression we observed is physiologically significant, we measured photosynthetic efficiency using a handheld device (Kuhlgert et al., 2016). As illustrated in Figures 5C-D photosystem II efficiency and proton motive force (vH+) were significantly higher in the AtMIOX4 OE line compared to WT control at 21, 23, 26 and 28 days after germination. Taking together these results show that the AtMIOX4 OE line is photosynthetically more efficient than its WT counterpart.
MIOX4 over-expression leads to increase auxin
Next, we investigated the role of phytohormones in the enhanced growth and abiotic stress tolerance phenotype recorded for the AtMIOX4 OE line. Analysis of the transcriptome data indicate that genes involved in auxin biosynthesis (YUCCA6: At5g25620, NIT2: At3g44300), auxin-amino acid conjugate hydrolase (ILL6: At1g44350), auxin transport (auxin efflux carrier family protein), and small auxin up RNAs (SAUR2: At4g34780, SAUR4: At4g34800, SAUR36: At2g45210) were up-regulated in AtMIOX4 OE line compared to WT controls, whereas auxin response small RNA (SAUR53: At1g19840) and auxin regulated organ size transcript (ARGOs: At3g59900) were down-regulated in AtMIOX4 OE line compared to WT (Figure 6A). The differential expression of auxin biosynthesis (YUCCA6, NIT2), auxin-amino acid conjugate hydrolase (ILL6), and transcripts involved in regulation of organ size (ARGOs) was confirmed via RT-qPCR (Figure 6B).
Next we determined the intracellular level of auxin and its conjugated form using LC-MS/MS. The intracellular concentrations of the active auxin (IAA) and inactive auxin (IAA-Ile) were significantly higher in AtMIOX4 OE line compared to WT (Figure 6C). To understand the effect of elevated auxin in AtMIOX4 OE line, next we measured the elongation of the primary root in light conditions and the hypocotyl length/hypocotyl epidermal cell length in dark conditions in the AtMIOX4 and controls. As shown in Figure D-E, the primary root of the high AsA line was longer than the control when plants grew under light conditions, and their hypocotyl was longer when plants grew in the dark. Next, we investigated whether the hypocotyl elongation in the AtMIOX4 OE line was due to increase in cell number or cell length using phase contrast microscopy. We found that hypocotyl epidermal cell length was significantly elongated in the AtMIOX4 OE line compared to WT controls (Figure 6F-G). Subsequently, we took advantage of the auxin sensor line R2D2 developed by Liao et al (2015) to monitor in vivo auxin levels in our high AsA line. For this, crosses between the R2D2 and the AtMIOX4 line (both lines are present in the A. thaliana Col-0 ecotype) were made. Positive crosses were selected using kanamycin as a selectable marker and PCR to confirm the presence of the transgenes of interest. Two positive crosses and controls were used for the in-vivo auxin visualization and semi-quantification. The R2D2 line expresses lower ntdTomato fluorescence (red) in the presence of normal auxin level. When there is a higher level of auxin, there is a higher level of ntdTomato fluorescence, compared to n3 × DII-Venus (green) expression which was prominent in both shoots (Figure 7A-B) and roots (Figure 7D-E). As illustrated in Figure 7 the ratio of ntdTomato fluorescence to n3X DII Venus fluorescence was significantly higher in the crosses compared to R2D2 control in both shoots (Figure 7C) and roots (Figure 7F). We did not find evidence of differentially regulated transcripts involved in cytokinin metabolism in the AtMIOX4 OE line. The level of active cytokinin (cis-zeatin), isomeric and relatively less active cytokinin (trans-zeatin), and cytokinin conjugate (zeatin-riboside) levels were similar in AtMIOX4 OE line compared to WT control as data from LC-MS/MS shows (Figure 8A-C). Taken all together, these results indicate that there are elevated levels of auxin due to the transcriptional up-regulation of auxin metabolism in the AtMIOX4 OE line and that the elevated auxin levels observed in the AtMIOX4 OE line are functionally relevant. Our observations support a major role for auxin in cell elongation leading to the higher mass phenotype of the high AsA line.
MIOX4 over-expressers tolerate abiotic stresses independently of abscisic acid up-regulation
We also examined some of the possible molecular mechanisms mediating the abiotic stress tolerance phenotype in AtMIOX4 OE line. We identified abiotic stress response-related based on the expression profiles of the DETs transcripts using MapMan. Among the statistically significant transcripts we found that heat stress response transcripts 17.8KDa class I heat shock protein (HSP17.8-CI-At1g07400), 17.4 KDa class III heat shock protein (HSP17.4-CIII-At1g54050), cold regulated protein 15a (COR15A-At2g42540) and the drought induced protein (ATDI21-At4g15910) were significantly up-regulated in AtMIOX4 OE line compared to WT control (Figure 8A).
We also looked at whether levels of abscisic acid were altered in the AtMIOX4 OE line compared to controls using LC-MS/MS. Abscisic acid levels were similar in AtMIOX4 OE and the WT line (Figure 8C). We then used Genevestigator to find signature genes that were over-expressed in publically available abiotic stresses experiments. Genevestigator analysis of the top 200 up-regulated (log2fold>1.75) and down-regulated transcripts (log2fold<−1.25) was compared against signature expression transcripts from 50 similar perturbations. We have found that cinnamyl alcohol dehydrogenase (ATCAD8: At4g37990), N-acetyl transferase activity 1 (At2g39030), lactolyl glutathione lyase (GLYI7:At1g80160), bHLH family protein (At1g10585) and NAC domain containing protein 19 (NAC019:At1g52890) were up-regulated in both, the set of abiotic stress perturbation experiments and in the AtMIOX4 OE line (Supplementary Figure 4). Taken together, these results point out to an important role for a conserved module of stress response genes in the abiotic stress tolerance phenotype observed in AtMIOX4 OE line.
High AsA exerts a contrasting regulation on glutathione reductase and ascorbate peroxidase genes
Antioxidants maintain the homeostasis of ROS in plants. We looked for genes involved in the antioxidant cycle network in the high AsA line to identify possible crosstalk within the redox network. We found that superoxide dismutase expression was similar in AtMIOX4 OE line compared to WT control (Figure 9A). Glutathione reductase 480 (GRX480: At1g28480), which regulates the protein redox state, and dehydroascorbate reductase 1 (DHAR1: At1g19570), involved in recycling of reduced ascorbate, were up-regulated while peroxisomal ascorbate peroxidase 5 (pAPX5: At4g35970), involved in scavenging ROS, was down-regulated in the AtMIOX4 OE line compared to WT control. The expression of APX5 and GRX480 was validated using RT-qPCR (Figure 9B). Interestingly, expression of genes involved in multiple pathways of L-ascorbate biosynthesis were similar in the AtMIOX4 OE and the control, while expression of DHAR1, involved in recycling of reduced ascorbate, was up-regulated (Supplementary Figure 1). Down-regulation of peroxisomal APX5 leads to significantly higher reactive oxygen species (H2O2) in AtMIOX4 OE compared to its WT counterpart (Figure 9C). Taking together these results indicate that high AtMIOX4 OE leads to higher ROS concentration due to down-regulation of peroxisomal APX gene. Additionally, high AsA leads to up-regulation of glutathione peroxidase and DHAR1 genes which maintain the ROS homeostasis in AtMIOX4 OE line.
High ascorbate plants seem ready to defend themselves against herbivores
Genes related to cell-wall metabolism, specialized metabolism, and defense hormone biosynthesis are important for the biotic stress response of plants. We found that genes involved in arabinogalactan metabolism and xyloglucan metabolism were down-regulated, whereas cellulose synthesis was up-regulated in AtMIOX4 OE line compared to controls. Three members of the Arabinogalactan gene family, AGP19: At1g68725, AGP3 (At4g40090) and AGP14 (At5g56540) were down-regulated. Similarly, xyloglucan metabolism genes XTH17 (At1g65310), XTH25 (At5g57550) and XTH10 (At2g14620) were down-regulated, whereas genes related to cellulose synthesis, CSLA 10 (At1g24070), CSLA 1 (At4g16590) and CSLG2 (At4g24000) were up-regulated in AtMIOX4 OE line compared to WT control.
Specialized metabolites are also an important line of defense in plants to combat stresses. Genes related to phenolic biosynthesis, O-methyl transferase family 2 protein (At1g76790), were up-regulated, while laccase biosynthesis genes, LAC11 (At5g03260) and LAC17 (At5g60020) were down-regulated in the AtMIOX4 OE line compared to WT. The phenylpropanoid biosynthesis gene, TSM1 (At1g67990) was only expressed in the AtMIOX4 OE line, while A. thaliana CAD-B2 (At4g37990) was up-regulated in AtMIOX4 OE line compared to controls. Flavonoid metabolism related transcription MYB 75 (At1g56650), which is involved in radical scavenging activity and anthocyanin metabolism, 2OG-Fe(II) oxygenase family proteins (At2g38240, At3g55970, At5g05600), and AACT1 (At5g61168) were up-regulated in AtMIOX4 OE line compared to controls. Similarly, terpene biosynthesis genes, GGPS6 (At1g49530), TPS10 (At2g24210), LAS1 (At3g45310), and ATTPS3 (At4g16740) were up-regulated in AtMIOX4 OE line compared to WT.
Defense hormones, such as jasmonic acid (JA), salicylic acid (SA), and ethylene are critical to induce the systemic acquired defense response in plants. Genes related to JA biosynthesis and metabolism, OPR3 (At2g06050) and JR1 (At3g16470) were up-regulated in the AtMIOX4 OE line compared to WT control. Similarly, genes related to SA metabolism, MES9 (At4g37150), involved in conversion of methyl salicylate to active SA, UDP- glucosyl transferase family protein (At1g05680), and BSMT1 (At3g11480) were also up-regulated in the AtMIOX4 OE line. JA response transcripts, VSP2 (At4g11740), VSP1 (At5g24780), and PDF1.4 (At1g19610) were significantly up-regulated in the AtMIOX4 OE line compared to WT. Similarly, we noted that ethylene response factors as RAP2.6 (At1g43160), RAP 2.6L (At5g1330), and RAP2.10 (At4g36900) were significantly up-regulated in the AtMIOX4 OE line compared to WT (Figure 11A; Table 4). The expression of the signature genes that play a significant role in biotic stress response (MYB27: At3g53200, OPR3, VSP2: At4g11740 and RAP2.6: At1g43160) were validated using RT-qPCR (Figure 11B). OPR3 regulates the biosynthesis of JA in plants. The over-expression of the OPR3 leads to higher levels of JA and JA-Ile, while the metabolite intermediate 12-oxophytodieonate (OPDA) was similar in AtMIOX4 OE line compared to WT control (Figure 11C-E). The over-expression of the genes related to specialized metabolites, such as phenylpropanoid, flavonoid, and terpene biosynthesis; up-regulation of genes involved in JA and SA biosynthesis; elevated level of JA and JA-Ile hormones; and expression of JA response genes, such as VSP1, VSP2, and PDF 1.4 suggests that AtMIOX4 OE plants are ready to defend themselves from herbivores attack.
Genes from the glucosinolate biosynthetic pathway are upregulated in MIOX4 over-expressers
Glucosinolates are specialized metabolites present mostly in plants of the Brassicaceae family. It is well known that glucosinolates and their hydrolysis products act as defense compounds against herbivores and pathogens. Glucosinolates are derived from amino acids and are grouped into aliphatic, aromatic, and indolic glucosinolates based on their different side chain structures. We have found that genes involved in the glucosinolate -myrosinase system, flavin monoxygenase glucosinolate S-oxygenase 2 (FMOGS-OX2: At1g62540), alkenyl hydroxalkyl producing 2 (AOP2: At4g03060), MYB domain protein 76 (MYB76: At5g07700), myrosinase binding protein 1 (MBP1: At1g52040), cytochrome P450 CYP83B1 (SUR2: At4g31500), and epithiospecifier protein (ESP: At1g54040) were elevated in the AtMIOX4 OE line compared to WT control (Figure 12A). The CYP83B1 gene is involved in the indol-and benzyl-glucosinolates biosynthesis which has triptophane and phenylalanine as precursors (Sønderby et al., 2010a). The MYB 76 is transcriptional activator of methionine-derived aliphatic glucosinolates in Arabidopsis thaliana (col-0) (Sønderby et al.,2007; Gigolashvili et al., 2008). We found that the level of glucoerucin, a methionine-derived aliphatic glucosinolate, was elevated ∼2.5 fold in AtMIOX4 OE line compared to WT control (Figure 12B). Interestingly, we found that negative regulator of methionine biosynthesis Arabidopsis MTO 1 responding down 1 (At5g53480) was significantly down-regulated in AtMIOX4 OE line (Supplementary Table 2).
Severity of down-regulation of pathogen resistant genes is compensated by increased SA level
Surprisingly, we found that genes related to the pathogen response as pathogenesis related proteins (At2g14610), glycosyl hydrolase superfamily protein (At4g16260), peroxidase family proteins (At5g51890), germin (GLP3: At5g51890), leucine rich repeat lipid family protein (LRR: At5g23400), and disease resistance responsive family protein (At1g6580) were down-regulated in the AtMIOX4 OE line compared to WT controls (Figure 13A). To test whether down-regulation of these genes affected the pathogen defense, we challenged AtMIOX4 OE (high AsA), vtc 1-1 (low AsA), and WT line (moderate AsA) with Pseudomonas syringae. As shown in Figure 13B the AtMIOX4 OE line does not seem to be compromised to defend against P. syringae attack, whereas vtc 1-1 line were resistant to P. syringae challenge compared to WT control when analysed by bacterial count (log cfu/cm2) after 3 days of infection (Figure 13B). We evaluated the level of SA in AtMIOX4 because SA is responsible for protecting the plants against pathogens by provoking the systemic defense response. The foliar level of SA was higher in AtMIOX4 OE line compared to controls, as shown by LC-MS/MS, without the challenge of biotic stress (Figure 13C). Despite the down-regulated pathogen defense response transcripts, the AtMIOX4 line was not compromised by the pathogen challenge. This can be due to the presence of high level of SA, which could compensate for the down-regulation of pathogen response genes by expressing lipid transfer proteins related to pathogenesis (At3g51590, At4g28395, At1g66850, At5g01870) and PROTEASE INHIBITOR (At4g22513) genes only in AtMIOX4 OE line.
Discussion
Ascorbate is fundamental for proper development and response to biotic and abiotic stress in plants (Conklin et al., 1999; Debolt et al., 2007; Dowdle et al., 2007; Ortiz-Espín et al., 2018; Gallie et al., 2013; Gest et al., 2013; Kerchev et al., 2011; Pastori et al., 2003). The D-mannose/L-galactose pathway (Wheeler et al., 1998) is well characterized compared to other ascorbate routes. Other pathways also called as “alternative routes” are also functional in different plants species, tissues and developmental stage specific manner (Lorence et al., 2004; Yactayo-Chang 2011; Yactayo-Chang et al., 2018). The MIOX4 transcript is expressed during reproductive stage in leaf stipules and was found predominantly in cytoplasm and in nucleus (Alford et al., 2012; Endres and Tenhaken 2011; Kanter et al., 2005). Over-expression of enzymes involved in the myo-inositol pathway leads to increased AsA content in Arabidopsis (Radzio et al., 2003; Lorence et al., 2004; Yactayo-Chang 2011; Yactayo et al., 2018). This has also been confirmed by other laboratories (Tóth et al., 2011). Additionally, a study in a T-DNA insertion mutant of the vtc4 enzyme showed that this proteins is involved in myo-inositol synthesis resulting in decreased AsA in kiwifruit (Actinidia deliciosa) in mutants compared to WT plants (Torabinejad et al., 2018). A recent study reports no conversion of myo-inositol into AsA in the AtMIOX4 lines in which we have detected gene silencing (the ones reported in the Lorence et al., 2004 paper) (Kavkova et al., 2018). Because of the gene silencing issue we developed new transgenic lines. The AtMIOX4 line 21 used in this transcriptomic paper is homozygous, contains a single copy of the transgene, and consistently has shown elevated AsA in more than 10 generations we have grown these plants.
The rosettes of the low ascorbate A. thaliana vtc2 mutant are smaller compared to WT due to intracellular structural deformity similar to programmed cell death (Olmos et al., 2006; Pavet et al., 2005). Whereas, plants over-expressing Arabidopsis myo-inositol oxygenase (MIOX4) (Lorence et al., 2004) have significantly higher (1.5 to 2-fold) ascorbate. Higher AsA results in enhanced biomass, and increased tolerance to abiotic stresses including cold, heat, salt, and pyrene, an environmental pollutant (Lisko et al., 2013). In this study, we report the physiological and molecular mechanisms likely underlying the enhanced biomass and abiotic stresses tolerance phenotype in AtMIOX4 OE using a comprehensive approach combining transcriptomic analysis, with LC-MS/MS, cell biology and physiological measurements.
We found that an α-amylase-like (At4g25000) transcript was over-expressed, and resulted in higher intracellular concentration of reducing sugars and glucose in AtMIOX4 OE line compared to WT using RNA-Seq, RT-qPCR, biochemical assay and enzymatic assay (Figure 4A-D). Glucose acts as a primary carbon and energy source in plants and plays a crucial role in plant growth, development, and physiology by coordinating with phytohormone signals. Enhanced production of glucose leads to manipulation of root architecture for maximum development through interactions with auxin metabolism genes (Li et al., 2007; Moore et al., 2003; Peng et al., 2018). Furthermore, increased concentration of UDP-glucose leads to accumulation of higher biomass in plants and helps in vegetative phase change by repressing the miR156A/ miR156C (Christopher et al., 2017; Yang et al., 2013). Additionally, Glucose induces expression of MYB family transcription factors and biosynthetic genes in the aliphatic glucosinolate pathway and in the presence of JA acts synergistically regulating glucosinolates accumulation (Guo et al., 2013: Miao et al., 2016).. Enhanced intracellular glucose accumulation indicates the higher pool of primary carbon and energy source, and signaling molecule for stresses tolerance, which might contribute to higher biomass and abiotic stresses tolerance phenotype in AtMIOX4 OE line compared to WT.
We have shown that photosynthesis light reaction associated transcripts were up-regulated, and that photosynthetic efficiency and proton motive force were higher in AtMIOX4 OE line compared to WT (Figure 5A-D). The AtMIOX4 OE line has higher non-photochemical quenching and energy dependent quenching due to the presence of elevated ascorbate in the lumen. Additionally, in heat stressed leaves of high AsA plants, ascorbate acts as PSII donor by retarding photoinactivation (Tóth et al., 2011).
Photosynthesis light reaction utilizes photosystem II (PSII) to capture the sunlight and splits water molecule into proton, electrons, and oxygen. The free electrons move through the thylakoid membrane and synthesize energy storage molecules. Furthermore, these energy storage molecules are utilized by photosynthesis dark reactions for carbohydrate biosynthesis (Johnson 2016). Higher rate of photosynthesis is associated with higher biomass production and greater yield in plants (Evans 2013; Taylaran 2011). Hence, increase in photosynthesis efficiency in AtMIOX4 OE provides plant with a higher amount of energy storage molecules and carbohydrates.
The growth and development of plants are regulated by the interaction between growth and defense hormones. We have shown that expression of auxin biosynthesis, auxin transport, auxin-aminoacid conjugate hydrolase, and auxin response signaling factors were up-regulated in AtMIOX4 OE line compared to the control (Figure 6A-B). The concentration of intracellular active auxin (IAA) and conjugated auxin (IAA-Asp) were higher in AtMIOX4 OE line compared to WT (Figure 6C). The functional response of elevated auxin in these plants is evidenced by increased primary root length when grown in light conditions, elongated hypocotyl and hypocotyl epidermal cell when grown in dark conditions and both visualization and semi-quantification of auxin levels using confocal microscopy. Auxin concentration was higher in both roots and shoots of AtMIOX4 OE line harboring the auxin sensor (R2D2) relative to the R2D2 control (Figure 7A-F). Auxin tightly regulates growth and development of plants. Tryptophan dependent auxin biosynthesis transcripts were over-expressed in AtMIOX4 OE line as nitrilase family gene converts indole-3-acetonitrile to IAA whereas, YUCCA family proteins converts indole-3-pyruvic acid to IAA which is the rate limiting step for tryptophan dependent auxin biosynthesis. Over-expression of both branches of auxin biosynthesis pathway lead to high IAA and IAA-Asp level (Korasick et al., 2013; Zhao et al., 2001). YUCCA6 over-expression results in high auxin phenotype such as delayed senescence, narrow and downward curled rosette leaves and tolerance to drought stress (Kim et al., 2011; Kim et al., 2013). The over-expression of YUCCA6, YUCCA1 and external application of auxin derivative delayed senescence and drought stress tolerance in plants due to elevated auxin (Kim et al., 2011; Lee et al., 2011; Mueller-Roeber and Balazadeh 2014). We have found YUCCA3 over-expression, delayed senescence and abiotic stress tolerance phenotype in the AtMIOX4 OE line compared to WT control. Additionally, auxin promotes plant organ elongation (hypocotyl, roots and coleoptiles) by enhancing the proton extrusion through the membrane H+ ATPase. Auxin lowers apoplastic pH through activation of membrane H+ ATPase, which induces cell wall loosening proteins facilitating inward movement of K+ and water (Claussen et al., 1997; Philippar et al., 2006; Takahasi et al., 2012; Tyburski et al., 2011). Whereas, cytokinins are involved in cell division and morphogenesis regulation, and abiotic/biotic stresses tolerance (Zalabak et al., 2013). However, constitutive expression of AtMIOX4 OE did not seem to lead to changes in cytokinin (cis-zeatin, trans-zeatin and zeatin-riboside) level in A. thaliana foliar tissue (Figure 8A-C), which can be observed in both RNA-Seq and LC-MS/MS data.
Analysis of our transcriptomics results and published evidence indicate that abundance of intracellular primary carbon source, increased photosynthetic efficiency and elevated auxin contributes to higher biomass phenotype in AtMIOX4 OE line compared to WT control. Additionally, delayed senescence in plants was indicated by the presence of high free auxin which is consistent with our results.
Cell wall metabolism is differentially regulated in AtMIOX4
The cell wall is important to maintain the structural integrity and stresses tolerance of plants. We found that cell wall metabolism genes were differentially modulated in AtMIOX4 OE line compared to WT control (Figure 11A; Supplementary Table 3). Lignin maintains the structure of the secondary cell wall. We found that laccase biosynthesis genes (LAC11 and LAC17) were down-regulated in AtMIOX4 OE line compared to WT control. Knockout of laccase biosynthesis (LAC1 and LAC17) genes in A. thaliana results in reduced plant growth and vascular development, decreased root diameter due to reduced lignification and increased saccharification (Berthet 2011; Zhao et al., 2013) which contrasts with our results. Auxin decreases the expression of AGP3 whereas AGP19 is involved in plant growth, cell division and expansion, leaf development and reproduction (Goda et al., 2004; Yang et al., 2007). Interestingly, A. thaliana arabinogalactan protein (AtAGP17) over-expression results in resistance to infection by Agrobacterium (Gaspar et al., 2004). The expression of the gene longofolia is involved in polar cell expansion by turgor driven approach by controlling XTH17 and XTH24 expression (Lee et al., 2017). In contrast, CSLA/CSLA-like gene family members that are involved in mannan and xyloglucan backbones biosynthesis (Liepman and Cavalier 2012) were up-regulated in AtMIOX4 OE line compared to WT counterpart.
Abiotic stress tolerance of high ascorbate line were independent of increased abscisic acid level
Prior studies have shown that, low ascorbate A. thaliana lines were sensitive to abiotic stresses such as heat, cold, light and oxidative stress (Conklin et al., 2013; Pavet et al., 2005). On the other hand, plants with elevated ascorbate were tolerant to abiotic stresses as heat, cold, drought, salt and oxidative stress (Eltayeb et al., 2012; Eltayeb et al., 2006; Lisko et al., 2013; Tóth et al., 2011; Yactayo-Chang et al., 2018). We found that heat stress response genes (HSP17.8-CI; HSP17.4-CIII), cold stress response genes (COR15A), and drought stress response genes (ATDI21, NAC019) were up-regulated in the AtMIOX4 OE line compared to WT (Figure 9A-B; Supplementary Figure 3). Whereas, intracellular concentration of ABA was similar in AtMIOX4 OE line and WT (Figure 9C). Over-expression of class I and class II HSPs are important for basal thermotolerance by interacting with translation factor (eEF1B), and knockdown of class II HSPs by RNAi result in severe heat sensitivity (McLoughlin et al., 2016). Moreover, over-expression of HSP 17.8 aggregates the A. thaliana citrate synthase enzyme and protect it from heat at 43°C (Liu et al., 2005). Additionally, HSP 17.8 over-expression results in freezing tolerance in maize (Shou et al., 2004). Similarly, over-expression of HSP 17.4-CII results in heat stress tolerance in tomato through co-regulation of heat stress induced heat shock transcription factor A2 (HsfA2) (Port et al., 2004). Constitutive expression of Cor15A enhances freezing tolerance of protoplast and chloroplast in A. thaliana. This is due to decrease incidence of lamellar to hexagonal II phase transitions which occur in the close proximity of chloroplast membrane and plasma membrane due to dehydration caused by freezing (Artus et al., 1996; Steponkus et al., 1998; Thalhammer et al., 2014). Interestingly, over-expression of NAC family transcription factors (ATAF, CUC and NAM) in rice enhances drought resistance and salt tolerance by trans-activating activity (Hu et al., 2006). In addition to this, over-expression of ANAC019 increases heat tolerance and drought tolerance in A. thaliana (Tran et al., 2004; Guan et al., 2004).
Peroxisomal ascorbate peroxidase 5 (pAPX5) is involved in scavenging ROS, which is down-regulated in the AtMIOX4 OE line compared to WT (Figure 9A-B). High amount of ROS are produced in peroxisomes due to peroxisomal specific enzymes as acyl coA oxidase, glycolate oxidase, sulphite oxidase, superoxide dismutase, sarcosine oxidase, and polyamine oxidase in different plant species (Arent et al., 2008; Douce and Neuberger 1999; Goyer et al., 2004; Hansch et al., 2006; del Rio et al., 2006; Kamada et al., 2008; Planas et al., 2013). Reactive oxygen species produced in peroxisomes are scavenged by pAPX or catalase (CAT). Spinach pAPX have high affinity for H2O2 compared to catalase (Ishikawa et al., 1998). Reduction in expression of pAPX results in higher level of H2O2. Enhanced level of H2O2 in peroxisomes results in up-regulated expression of genes involved in oxidative stress response as small heat shock family proteins, DNAJ heat shock family protein, aromatic alcohol: NADPH reductase, and bHLH transcription factor. It was previously observed that silencing of pAPX results in oxidative stress tolerance in rice (del Rio et al., 2006; Sousa et al., 2015).
Based on a combination of the results shown here and published evidence, we propose a model were elevated AsA in A. thaliana reduces the expression of pAPX resulting in high level of peroxisomal ROS. High ROS in peroxisomes may act as signaling molecules and elevate the expression of abiotic stresses tolerance genes as heat response genes (HSP17.8-CI; HSP17.4-CIII), DNAJ heat shock protein family, bHLH transcription factor and aromatic alcohol: NADPH reductase. This will result in plants tolerant to oxidative stress. In addition to this, cold stress response gene (COR15A), drought stress response genes (ATDI21, NAC019) that were up-regulated in AtMIOX4 OE line could lead to cold and drought stress tolerance.
High AsA plants seem ready to defend themselves from stresses
Genes related to specialized metabolism were differentially regulated in AtMIOX4 OE line compared to WT control. Phenolics biosynthesis gene o-methyl transferase family 2 protein was up-regulated. Phenylpropanoid biosynthesis genes (TSM1, CAD-B2), flavonoid biosynthesis (MYB75), the JAO2, JAO3, JAO4, AACT1 genes and terpene biosynthesis genes (GGPS6, LAS1, TPS10) were up-regulated in the high AsA line compared to controls (Figure 11A). Over-expression of MYB75 results in accumulation of anthocyanin and protects A. thaliana from high light intensity (Gonzalez et al., 2008; Li et al., 2016). Glucosinolates-myrosinase system genes such as flavin monoxygenase glucosinolate s-oxygenase 2, alkenyl hydroxalkyl producing 2, MYB domain protein 76, myrosinase binding protein 1, and epithiospecifier protein were up-regulated in AtMIOX4 OE line compared to WT (Figure 12A). MYB76 transcription factor positively regulates the biosynthesis of methionine-derived aliphatic glucosinolates (Sønderby et al, 2007; Gigolashvili et al., 2008). MYB76 is involved in synthesis of methylsulfinylalkyl glucosinolates from methylthioalkyl glucosinolates catalyzed by FMOGSOXS enzyme family (Hansen et al. 2007; Li et al., 2008). Epithiospecifier protein is involved in nitrile formation upon glucosinolates hydrolysis and making Arabidopsis less appealing to specialist herbivores (Burow et al., 2010; Lambrix et al., 2001). Methionine-derived glucosinolates acts as a chemoprotective compounds in plant defense against biotic stress (Gigolashvili et al., 2007). Increased levels of Glucoerucin observed in AtMIOX4 OE might indicate that high AsA plants could be less susceptible to herbivore attack. Furthermore, Glucoerucin, when converted into erucin by myrosinase enzyme, acts as an antioxidant by decomposing hydroperoxides and hydrogenperoxides. Erucin oxidizes into glucoraphanin which is most effective chemoprotective for cancer. Eruin protects from cancer by multiple mechanisms as modulation of phase I enzymes, inducing phase III enzymes, up-regulating phase III detoxification system, and modulation cell proliferation (Barillari et al. 2005; Dinkova-Kostova et al. 2017; Fimognari et al. 2017; Melchini and Traka 2010). Further analysis on other glucosinolates present in AtMIOX4 OE line is now in progress and will help us to understand the effect of glucosinolates in this plant.
Jasmonic acid biosynthesis gene and JA-responsive genes were up-regulated in AtMIOX4 OE line compared to WT. JA biosynthesis intermediate metabolites (OPDA) were similar but JA and JA-Ile level were significantly higher in AtMIOX4 OE line compared to controls (Figure 11A-E). Over-expression of OPR3 results in increased level of JA and JA-Ile in A. thaliana and plants become resistance to insects and fungi (Suza and Staswick 2008) which is similar to our results. Interestingly, JA is also involved in plant development, growth, reproduction, and protects plants against abiotic stresses as salinity, drought, and UV-radiation (Kerchev et al., 2011; Wasterneck and Hause 2013). Salicylic acid metabolism genes and SA level were up-regulated in AtMIOX4 OE line compared to WT control. In contrast, pathogenesis related proteins were down-regulated in AtMIOX4 OE line compared to WT control (Figure 11A, 13A, 13C). The expression of pathogenesis related proteins was significantly down-regulated in AtMIOX4 OE line. Surprisingly, the response of AtMIOX4 OE line was similar to WT control when challenged with P. syringae. This may be due to exclusive expression of pathogen response genes in AtMIOX4 line (Figure 3A; Table 2). Salicylate protects the plants from biotrophic and hemibiotrophic pathogens by interacting with NPR1 (Palmer et al., 2017). Salicylic acid not only protect plants against biotic stress, it is also involved in growth, development of plants and increases tolerance against abiotic stresses as cold, heat, drought, metal by increasing antioxidant capacity (Bartoli et al., 2013; Gimenez et al., 2017; Horváth et al., 2015). Furthermore, ethylene-responsive element binding factor 13 induces the over-expression of RAP transcriptions factor. Over-expression of RAP2.6 resulted in increased resistance to nematode (Heterodera schachttii) in A. thaliana by enhanced callose deposition (Ali et al., 2013).
Crosstalk between high AsA, SA and JA
Glutathione reductase 480 (GRX480) and DHAR1 which is involved in recycling of reduced ascorbate were both up-regulated while peroxisomal pAPX5 was down-regulated in AtMIOX4 OE line compared to WT control (Figure 9A-B). It was reported that, under high light stress, elevated H2O2 is associated with up-regulation of small heat shock proteins and transcription factors related to anthocyanin biosynthesis (Vanderauwera et al., 2005). Similarly, anthocyanins protect plants against photo-protection and photo-oxidative damage. Mutation of thylakoid APX results in elevated H2O2 under high light stress which increases the expression of genes involved in anthocyanin biosynthesis regulation (Maruta et al., 2014). We found that the positive regulator of anthocyanin biosynthesis MYB75 and the anthocyanin biosynthesis 2OG-Fe(II) oxygenase family proteins were up-regulated in AtMIOX4 OE line compared to control (Supplementary Table 3). These results suggest that high AsA A. thaliana are ready to protect themselves from photo-oxidative damage by regulating anthocyanin biosynthesis.
Salicylic acid negatively regulates the JA biosynthesis genes and decreases expression of JA-responsive genes (PDF1.2 and VSP2) (Chehab et al., 2011; Vlot et al., 2008; Leon-Reyes et al., 2010). We found that JA and SA were both elevated and JA-responsive defensin proteins were highly expressed in AtMIOX4 OE line compared to control. GRX480 was inducible by SA and suppresses the JA responsive defensin (PDF1.2) transcription in a NPR1 dependent manner. GRX480 interacts with TGA2 transcription factor and suppresses the ethylene-response factor 59 (ERF59) resulting in suppression of JA response defensin (Ndamukong et al., 2007). Published literature indicates that higher levels of endogenous ethylene diminish the SA and JA antagonism (Leon-Reyes et al., 2009; Leon-Reyes et al., 2010). Ethylene response factors as RAP2.6, RAP2.6L, RAP 2.10 were up-regulated in AtMIOX4 OE line compared to WT control (Figure 11A). The over-expression of the ethylene response factors might contribute to diminish the SA and JA antagonism, hence increased level of both SA and JA in high AsA A. thaliana.
Conclusions and future perspectives
We have here demonstrated that high AsA AtMIOX4 OE contain increased intracellular glucose, and enhanced photosynthetic efficiency. We also found that elevated in vivo auxin levels contributed to increased cell elongation and delayed senescence. The enhanced glucose, auxin and photosynthetic efficiency are likely explanations for the enhanced biomass phenotype of the high AsA line. Furthermore, the up-regulated expression of signature genes and transcription factors for heat stress, cold stress, drought stress and salt stress tolerance over-expression likely lead to abiotic stresses tolerance in this high AsA AtMIOX4 OE line. The fact that abscisic acid was similar in AtMIOX4 and controls indicates that abiotic stresses tolerance in this line is most likely independent of up-regulation of ABA. Interestingly, biotic stress response genes related to JA, SA, specialized metabolism and glucosinolates biosynthesis and metabolism genes were up-regulated in the high AsA line. The defense hormones JA, JA-Ile and SA were elevated in the high AsA line compared to its counterparts. These defense hormones and elevated expression of biotic stress response signature genes might confer the tolerant/resistant phenotype of AtMIOX4 in response to biotic stresses. Furthermore, the up-regulated expression of ethylene response factors in the high AsA line strongly suggests the presence of elevated levels of ethylene. Elevated ethylene might abolish the SA and JA antagonism in high AsA line. These results not only shed light into the molecular mechanisms underlying the enhanced biomass and abiotic stress tolerance phenotype of the high AsA line but also point to a conserved module of biotic stress response genes. Future studies are necessary to test the ability of high AsA lines to respond when challenged with herbivores. Next, steps should also include the expression of AtMIOX4 in crops of commercial value to fully exploit the potential biotechnological value of this knowledge.
Author Contributions
NN performed the experiments, prepared figures, and wrote initial draft of the manuscript, JPYC generated the MIOX4 homozygous line; JPYC and LMAG prepared RNA samples for sequencing, MEGR contributed to HPLC analysis, KMJ and MAV analyzed the RNA-Seq data, AL conceived the experiments, secured funding, and wrote final draft of the manuscript.
Acknowledgments
We thank Dr. Fabricio Medina-Bolivar (A-State) for providing access to the HPLC, Dr. Sophie Alvarez (UNL) for technical assistance in LC-MS/MS measurements, Dr. Jian Hua (Cornell) for assistance with pathogens assays, Krishna Deo Sharma (A-State) for assistance in microscopy, Zachary Campbell (A-State) for technical assistance, and Kimberly Lee (A-State) for plant care. We thank Dr. Fiona Goggin (UAF) for insightful suggestions to the project. NN thanks the Molecular Biosciences PhD Program for providing stipend support. We also thank the Arkansas Biosciences Institute (ABI, A-State) for providing facilities to conduct experiments. This research was funded by the Arkansas Center for Plant Powered Production (Arkansas EPSCoR Track 1, fund EPS 0701890), the Plant Imaging Consortium (PIC, AR EPSCoR Track 2, fund IIA 1430427) funded by the National Science Foundation and a grant to AL from the Arkansas Research Alliance.