1. Introduction
Plant growth and development is highly dependent on a variety of environmental conditions such as temperature, light, water availability, and soil conditions. Among these environmental restrictions, flooding is one of the most severe factors that reduce the productivity of crops. Oxygen deficiency (or hypoxia) is the main stress during soil flooding and leads to a decrease in cellular energy charge, cytoplasmic pH, and accumulation of toxic end products from anaerobic respiration and reactive oxygen species (ROS) during recovery [
1,
2]. Plants have developed adaptation mechanisms at the morphological, anatomical, and molecular levels to enhance their survivability in low-oxygen conditions.
Low oxygen causes a dramatic change in gene expression and protein synthesis in plants. Oxygen deficiency in plant cells leads to the enhancement of metabolism such as sucrose catabolism, glycolysis, and fermentation pathways, which are important for energy conservation [
3,
4]. Hypoxia-induced genes and proteins include signal transduction components and transcription factors, as well as those involved in nitrogen metabolism, ethylene biosynthesis, and cell wall loosening [
5,
6,
7]. The mechanisms to combat low-oxygen stress have been intensively studied over the past years, and these studies mainly used
A. thaliana and rice as a model system [
2,
8,
9,
10,
11,
12,
13,
14]. Recently, low-oxygen stress responses were also studied in other species such as Rorippa and Rumex, which are known as tolerance species to submergence and flooding [
13,
15,
16]. The comparative analysis of various species with different low-oxygen stress tolerance levels could overcome the narrow model plant-based knowledge and provide a useful basis to understand the conserved and unique mechanisms by which species adapt to low oxygen.
Species closely related to Arabidopsis, such as
Arabis stelleri,
Rorippa islandica,
Thlaspi arvense,
Thellungiella salsuginea, and
Thellungiella parvula in the Brassicaceae (or Cruciferae) family have been studied for adaptive responses to various abiotic stresses [
17,
18,
19,
20,
21,
22,
23,
24]. The morphology and growth rate of species closely related to Arabidopsis look very different from each other (
Supplementary Figure S1A). In the phylogenetic relationship of these species based on internal transcribed spacer (ITS) sequence,
A. thaliana was more closely related to
A. stelleri and
R. islandica than three species of
Thellungiella genus and
T. arvense (
Supplementary Figure S1B).
The perennial herb
A. stelleri (or rock cress) is found in seaside sand or between rocks. Recent studies have shown that
A. stelleri var
japonica is stronger in osmotic stress tolerance than
A. thaliana [
17].
R. islandica occurs in high mountain sites and inhabits on the banks of mountain streams and shores of lakes [
25]. It has been reported that Rorippa species such as
Rorippa sylvestris and
Rorippa amphibia could tolerate deep submergence [
15].
T. arvense is a winter annual cruciferous weed, commonly known as stinkweed, which can survive extreme winter temperatures in Canadian Prairies [
26]. With such features in growth habitats,
T. arvense has recently been recognized as a model plant for freezing tolerance [
18].
T. salsuginea (salt cress; also called
T. halophila) is native to harsh environments. The Shandong and Yukon ecotypes of
T. salsuginea have been proposed as new model plants for research on abiotic stress tolerance [
19,
20,
21,
22]. The Shandong ecotype grows in high-salinity coastal areas in eastern China, and has been proposed to be an appropriate relative of Arabidopsis for studies of salinity tolerance mechanisms [
19,
27]. On the other hand, the Yukon ecotype was isolated in the Takhini Salt Flats near Whitehorse in the Yukon Territories, Canada, a subarctic and semiarid region [
28] characterized by multiple simultaneous abiotic stresses, including cold, drought, and high salinity [
21]. Another species,
T. parvula is as tolerant as
T. salsuginea under salt and drought stress, but has distinct morphological features [
23,
29]. Despite numerous abiotic stress-related studies of species closely related to Arabidopsis, the low-oxygen stress responses of these species have yet to be explored, expect for Rorippa species.
Rorippa sylvestris and
Rorippa amphibia were studied for submergence response including plant survival, changes in carbohydrate and metabolite concentrations, and transcriptome [
15].
In this study, we compared the physiological and transcriptional responses to low-oxygen stress with Arabidopsis and its related species to gain comprehensive insights into the low-oxygen responses of the species. We first compared the physiological responses of a total of six species with different accessions, including three ecotypes of A. thaliana (Columbia (Col-0), Landsberg erecta (Ler), and Wassilewskija (Ws), freezing-tolerant Arabidopsis esk-1 mutant (Coumbia ecotype)), A. stelleri, R. islandica, T. arvense, Shandong and Yukon ecotypes of T. salsuginea and T. parvula under low-oxygen stress. Based on these physiological responses, we selected four species and considered A. thaliana (Col-0) as a moderately tolerant species, A. stelleri and R. islandica as highly tolerant species, and Shandong ecotype of T. salsuginea as intolerant species under low-oxygen stress. With these four species, gene expression profiling using an Operon Arabidopsis microarray was carried out at various time points over 72 h, and the gene expression profiles were comparatively analyzed using gene set enrichment analysis (GSEA).
3. Discussion
For the comparison of six species with different accessions of the Brassicaceae family based on both general growth and survival under submergence or low-oxygen treatment, we selected four representative species with different degrees of low-oxygen tolerance: A. stelleri and R. islandica as highly tolerant species, A. thaliana (Col-0) as a moderately tolerant species, and the Shandong ecotype of T. salsuginea as an intolerant species. To investigate the low-oxygen responses of these species at the molecular level, gene expression profiling using an Operon Arabidopsis microarray was carried out at various time points over 72 h under low-oxygen stress. From a comparative analysis of gene expression profiles of the four species, we concluded that the various tolerance levels of these species might be attributed to different reconfigurations of energy metabolisms under low-oxygen stress.
The low-oxygen stress tolerance levels in
A. thaliana and close relatives were investigated to compare and identify candidates suitable for further comparative analysis of gene expression profiles. The Arabidopsis genome sequence and resources provide powerful tools that can initiate comparative genomic studies within Brassicaceae and beyond [
33,
34,
35,
36]. Cross-species hybridization is used in comparison, evolutionary and ecological studies, and for gene expression profiling of many species that lack a representative microarray platform. Recently, Arabidopsis microarrays have been successfully used in close relatives such as Brassica [
37], Thellungiella [
29,
38] and Thlaspi [
18], Rorippa [
15].
For the comparative analysis of low-oxygen stress responses of
A. thaliana and close relatives, we used the GSEA method, which is one of the statistical methods for analyzing gene expression profiles [
39,
40]. When the GSEA method was applied identically to the gene expression data of the four species, the number of enriched gene sets was much lower in close relatives than in
A. thaliana (
Figure 3). These results suggest that the change of gene expression was smaller in close relatives than in
A. thaliana. Alternatively, it might be because the gene set used in GSEA analysis was more appropriate for
A. thaliana gene expression studies. We ruled out the possibility of the low efficiency of cross-hybridization because the averages of hybridization signal intensities were also similar between
A. thaliana and each of the closely related species. The MM gene set database used in GSEA analysis was constructed based on the functional categories of ~30,000 Arabidopsis genes in MapMan ontology, which was determined with information from The Arabidopsis Information Resource (TAIR), the Gene Ontology Consortium (GOC), the functionally categorized Kyoto Encyclopedia of Genes and Genomes (KEGG) database, and manually predefined keywords of TIGR release version 3.0 [
41].
In the comparative analysis of low-oxygen stress responsive gene expression in each species, the enrichment of gene sets associated with energy metabolisms was intensively examined since the balance between the consumption and production of energy is important to survive under low-oxygen stress. Consistent to our observation, an analysis with two populations of
B. rapa with different waterlogging tolerances suggested the importance of carbohydrate supply to roots as a potential parameter for tolerance [
42]. In addition, Zou et al. (2015) suggested that a tolerant variety of
Brassica napus have a greater ability to maintain basic metabolism with lower energy consumption [
43]. Thus, our results and other previous studies indicate that plants must decrease unnecessary metabolic functions to the lowest possible level under low-oxygen condition to minimize unnecessary energy consumption. We also found that each species with a different degree of low-oxygen stress tolerance had distinct strategies for reconfigurations of energy metabolic pathways. The metabolic changes caused by oxygen deprivation were investigated by comparison of changes in expression of genes associated with energy metabolic pathways including glycolysis, fermentation, glyoxylate cycle, gluconeogenesis, TCA/organic acid transformation, mitochondrial electron transport/ATP synthesis, nitrogen metabolism, and photosynthesis (
Figure 6). In general, the mobilization of carbohydrates is caused by oxygen deprivation to promote substrate-level ATP production [
3,
4]. Amylases and sucrose synthase induced by oxygen deprivation promote the conversion of starch and sucrose to glucose. Within minutes of transfer to an oxygen-depleted environment, plant cells relying on external oxygen would limit processes that are highly energy consumptive and alter metabolism to increase the anaerobic generation of ATP by cytosolic glycolysis [
44]. The glycolysis reaction is activated and the excess NADH is recycled through fermentation, regenerating NAD
+ required to maintain glycolytic flux. When pyruvate levels increase by activation of glycolysis, the low K
m of mitochondrial pyruvate dehydrogenase (PDH) and high K
m of pyruvate decarboxylase (PDC) function to limit carbon entering the TCA cycle and promote ethanolic fermentation [
7]. These reconfigurations of energy metabolism were observed in the early response of
A. stelleri and late response of
A. thaliana,
A. stelleri and
R. islandica. Previous studies with waterlogging tolerant and sensitive varieties of rapeseed showed high induction of glycolysis-related genes in both varieties of
Brassica napus under waterlogging [
43], suggesting that the upregulation of glycolysis in energy metabolism is a common response in plants regardless of their tolerance levels to waterlogging. However, our data indicated that regulations of glycolysis under low-oxygen conditions are different in each species with the different tolerance levels. Using nitrate to nitric oxide through the mitochondrial electron transport chain, plant mitochondria can drive ATP synthesis. Nitrite produced by nitrogen metabolism in cytoplasm may serve as an electron acceptor at complexes III (bc1) and IV (
COX). This process can result in proton pumping, and can be linked to an observed ATP synthesis [
45]. Nitric oxide formed in the mitochondrial electron transport process is converted to nitrate in the cytosol by hypoxia-induced hemoglobin (Hb). It is then reduced by nitrate reductase (
NiR) to nitrite, and the cycle is repeated. Under low-oxygen stress, this reaction in nitrogen metabolism was only observed in the early response of
A. thaliana. Fatty acids degraded by beta oxidation and glyoxylate cycle were also used as an energy source in the low-oxygen stress response of
A. thaliana and close relatives. The glyoxylate cycle, a variation of the TCA cycle, is an anabolic metabolic pathway that occurs in the peroxisome of the plants. This cycle allows plants to take in acetate both as a carbon source and as a source of energy. Acetate is converted to acetyl CoA, and some succinate is released during the cycle. The four-carbon succinate molecule can be transformed into a variety of carbohydrates through the TCA cycle. Acetyl CoA can also react with glyoxylate to produce some NADPH from NADP
+, which is used to drive energy synthesis in the form of ATP later in the electron transport chain [
46]. A shunt to control beta oxidation and glyoxylate cycle was observed in the low-oxygen stress response of
A. thaliana and
T. salsuginea, which were relatively sensitive to low-oxygen stress, suggesting that low-oxygen stress-sensitive species are able to use fatty acids as an alternative source of energy in the beta-oxidation process and glyoxylate cycle during a prolonged period of low oxygen. Recently, the glyoxylate cycle was also reported as having unique flexibility in energy metabolism in mycobacteria under oxygen-limiting condition [
47].
At the whole-plant level, complete submergence leads to a dramatic shift in the carbon budget and energy status, potentially resulting in death. Underwater photosynthesis provides some relief for this problem, with the leaves still submerged, [
48], but underwater photosynthesis can be limited by low light and CO
2 availability. When plants are facing low-oxygen stress, photosynthesis is reduced as a result of stomatal closure, decreased activity of carboxylation enzymes and decreased leaf chlorophyll content [
49]. Photosynthesis was inhibited by low-oxygen stress in
A. thaliana,
R. islandica and
T. salsuginea as conjectured, but not in
A. stelleri. In order to further examine the efficiency of photosynthesis of
A. stelleri during low-oxygen stress, the maximum quantum yield of PSII (Fv/Fm) of
A. stelleri was measured after low-oxygen stress (
Supplementary Figure S4). After 72 h of low-oxygen treatment, the Fv/Fm values of the roots and leaves of
A. stelleri decreased by 51% and 25%, while this value for the roots and leaves of
A. thaliana decreased by 100% and 60%, respectively. The results indicated that
A. stelleri was allowed to retain some capacity of PSII photochemistry during low-oxygen stress. In spite of the root samples, the expression of photosynthesis-related genes and the efficiency of photosynthesis were changed by oxygen deprivation. It is not clear what caused the induction of photosynthesis-related genes in
A. stelleri after low-oxygen treatment. Sasidharan et al. (2013) speculated that sugar starvation might be one of the causes of these gene induction [
15]. However, we did not observe the significant differences between an expression of sugar starvation related genes in
A. stelleri and other species. Plants sense ambient light conditions and modulate their developmental processes by utilizing multiple photoreceptors such as phytochromes, cryptochromes, and phototropins. Even roots, which are normally not exposed to light, express photoreceptors and can respond to light by developing chloroplasts [
15,
30]. Roots show various photoresponses, and light influences many aspects of root development including root extension, geosensitivity, and lateral root formation [
24].
In conclusion, this study showed that each species with a different degree of low-oxygen stress tolerance distinctively reconfigures energy metabolic pathways under low-oxygen stress. The comparison of the differences between the responses of the four species to a particular stress helped to explain their ability to withstand stress. In A. thaliana, the dynamical reconfiguration of energy metabolisms in early response was restricted in late response to low-oxygen stress, suggesting that the survival of A. thaliana is seriously affected when exposed to a low-oxygen condition for a prolonged period. Given the fact that photosynthesis genes were enriched in A. stelleri, it is tempting to speculate that in low-oxygen stress conditions, this highly tolerant species sustains some ATP levels through respiration using O2, presumably from high photosynthesis capacity. Therefore, this early response of A. stelleri might lead to a better chance of survival under low-oxygen stress. However, it should be noted that our speculation is still hypothetical and it still requires further study. R. islandica, one of the highly tolerant species, retained some energy via ATP produced by anaerobic energy metabolism during a prolonged period of low oxygen. In T. salsuginea, a relatively intolerant species, there were no significant changes in the expression of genes involved in anaerobic energy metabolisms, suggesting that their ability to cope with energy crises under low-oxygen stress decreased. The comparative analysis of species with different degrees of low-oxygen stress tolerance provides a more comprehensive understanding of the reconfiguration of energy metabolism in the low-oxygen response of plants. Furthermore, the outcome of this study will help to develop flood-resistant or flood-tolerant crop plants.
4. Materials and Methods
4.1. Plant Materials and Growth
Seeds of
A. thaliana (Columbia, Landsberg erecta and Wassilewskija),
T. salsuginea (Shandong and Yukon), and
T. parvula were obtained from the Arabidopsis Biological Resource Center at Ohio State University (Columbus, OH, USA), and the seeds of
T. arvense were obtained from B&T World Seeds (
http://b-and-t-world-seeds.com/). The seeds of Arabidopsis mutant
eskimo 1 (
esk-1) were provided by Dr. John Browse at Washington State University (USA) and the seeds of
R. islandica were provided by Dr. Hong-keun Choi at Ajou University (Korea). The seeds of wild-type
A. stelleri var.
japonica were collected from the seaside near the city of Pohang, Korea.
For submergence treatment, the seeds were sown in Sunshine #5 soil (Sun Gro Horticulture, Canada) in pots. All the plants were grown at 23 °C in long-day conditions (16 h light/8 h dark) with approximately 100 µmol m−2 s−1 light intensity in an environmentally controlled growth chamber. For low-oxygen treatment and gene expression experiments, the seeds were surface-sterilized, imbibed in water in darkness at 4 °C for 3 days for synchronized germination, and grown in growth medium (Murashige and Skoog half-strength solution containing 1% phytoagar).
4.2. Submergence and Low-Oxygen Treatment
For submergence treatment, 13~15 plants were grown in a pot until the developmental stage immediately before bolting in soil. Three-week-old Arabidopsis or four-week-old closely related species were entirely submerged in a tank of water 40 cm deep and held for 7 days at 23 °C in a dark condition.
For low-oxygen treatment, 18~20 plants were grown on an agar plate until the stage of emergence of three to four leaves. Two-week-old Arabidopsis or three-week-old closely related species were transferred to an airtight vacuum chamber in an environmentally controlled growth cabinet. A low oxygen gas mixture (0.1% O2/99.9% N2) was supplied continuously to the airtight vacuum chamber to retain low O2 conditions during treatment. The oxygen concentration in the airtight vacuum chamber was monitored by an XP-3180 oxygen meter (Cosmos, Osaka, Japan). This treatment was carried out at 23 °C for 5 days in a dark condition. In each experiment, plant survival was scored based on the survival of shoot apical meristem after recovery.
Each experiment for submergence and low-oxygen treatment was repeated three times.
4.3. RNA Isolation and Purification
Total RNA was isolated using an RNeasy Plant Mini Kit (Qiagen, Hilden, Germany), following the manufacturer’s instructions. Briefly, the root samples of 18~20 plants on an agar plate were homogenized in the presence of liquid nitrogen, and were lysed in a buffer containing guanidine isothiocyanate (GITC). The lysed samples were placed in the RNeasy columns and washed with an ethanol-containing buffer. Total RNA was eluted with RNase-free water. For ethanol precipitation, 1 mL of 95% ethanol and one-tenth volume of 3 M NaOAc, pH 5.2 were added and the total RNA samples were held at −80 °C for 20 min. After centrifuging at 12,000× g for 15 min at 4 °C, the RNA pellets were washed with 1 mL of 70% ethanol and centrifuged at 12,000× g for 5 min at 4 °C. The pellets were then dissolved in RNase-free water. The concentration and purity of this isolated total RNA were determined by measuring the absorbances at 260 nm and 280 nm.
4.4. Microarray Experiment, Image Acquisition, Data Acquisition and Normalization
For the microarray experiment, plants were grown in an environmentally controlled growth chamber at 23 °C under constant light (approximately 100 µmol m
−2s
−1). Total RNA (5 μg) from root samples of
A. thaliana (Col-0),
A. stelleri,
R. islandica, and
T. salsuginea (Shandong) exposed to low oxygen for 0, 1, 3, 8, 24, or 72 h was reverse transcribed using reverse transcription (RT) primer tagged with either Cy3-3DNA or Cy5-3DNA capture sequence of Array 900 MPX Expression Array Detection Kits (Genisphere, Hatfield, PA, USA). The synthesized cDNA including capture sequence were fluorescently labeled by Cy3-3DNA or Cy5-3DNA based on the sequence complementary to the 3DNA capture reagent, which contained an average of 900 fluorescent dyes. The labeled cDNA was hybridized on an Operon Arabidopsis Version 3.0 microarray consisting of 26,173 probes spotted with synthetic 70-mer oligonucleotides on aminosilane-coated slides by the David Galbraith lab, University of Arizona. The hybridization and washing procedures were performed according to Genisphere technical protocol. After washing, the slides were immediately scanned using an ArrayWoRx (Applied Precision, Issaquah, WA, USA). To maximize the camera’s dynamic range without saturation and to normalize the two channels for signal intensity, the exposure setting was adjusted so that the intensity level of the brightest spot on a slide was 80 to 90%. Experiments were performed a replicated dye swap for each microarray (
Supplementary Figure S3).
Intensity values were quantified from the pairs of TIFF image files from each channel using version 5.6 ImaGene software (BioDiscovery, Los Angeles, CA, USA). Analyses were done using the version 4.1 GeneSight software package (BioDiscovery, Los Angeles, CA, USA). For each slide, the local background was subtracted from the signal intensity, and the minimum intensity was raised to 20 by using the “floor” function. The mean intensity for each element was normalized by the locally weighted scatterplot smoothing (LOWESS) method and expression values (log
2) were calculated by comparing intensities from (1) each three species and Arabidopsis before treatment (LT00) (“comparison each three species versus
A. thaliana as reference” in
Supplementary Table S2) or (2) each species at each time points (LT01, 03, 08, 24, and 72) and each species before treatment (“comparison low-oxygen versus control” in
Supplementary Table S2) using the GeneSight software. A two-sided t-test was performed using the R Statistical Package [
50], to determine which genes were significantly differentially expressed between the low-oxygen treated and control groups, and Benjamini–Hochberg false discovery rate (FDR) multiple testing correction (
https://cran.r-project.org/web/packages/ simulator/vignettes/fdr.html) and alpha level of 0.05 was applied [
51]. The FDR method was used for testing and adjustment of p-values.
The significantly differentially expressed genes were identified using the following parameters as a confidence threshold: adjusted p value < 0.05 and log2 fold change ≤ −1.0 or ≥ 1.0.
4.5. GSEA Analysis Using MapMan-Based Gene Set Database
A MapMan-based gene set (MM gene set) database was constructed based on functional categories of MapMan ontology to analyze microarray data of Brassicaceae species at the gene set level using Gene Set Enrichment Analysis (GSEA) software. MapMan ontology (file name: Map files/Ath_AGI_TAIR9.txt) was downloaded from the MapMan website (
http://mapman.gabipd.org/web/guest/mapman-store), where each of the genes was assigned to the corresponding categories of a tree-like hierarchical structure classified according to the functional categories. Each gene set was defined by the group of genes in each of the hierarchically functional categories assigned by a BIN code of MapMan ontology. “MM gene set_type I” was constructed by genes grouped in level 1 of the BIN code of hierarchically functional categories. However, the number of genes included in a few categories of level 1 was too small or too large to calculate the statistical significance using GSEA software. Hence, “MM gene set_type II” was additionally constructed by genes grouped in levels 2 or 3 of the BIN code of these categories. Each type of functional gene set database was separately carried out on GSEA software but all GSEA results obtained by “MM gene set_type I” and “MM gene set_type II” were used for analysis. A graphical tree map of MapMan-based gene sets was drawn based on the hierarchical structure of functional categories in MapMan ontology using version 1.1 Scalable Vector Graphics (SVG) software (
http://www.w3.org/Graphics/SVG/) (
Figure 3B).
The GSEA was performed using version 2.0.7 of the GSEA-P software downloaded from the GSEA website (
http://www.broadinstitute.org/gsea/) [
40]. To calculate the significance of the enrichment score (ES), class labels were randomly permuted and ES were recalculated 1000 times. In this study, the cutoff for significance of ES was defined as the score according to a
P value of 0.05 and an FDR value of 0.25. A statistically significant value for the gene sets represented by less than 10 genes was defined to be a
P value of 0.1 since a small population size has a negative influence on statistical significance. GSEA evaluated a query microarray data set by using the MapMan-based gene set (MM gene set) database.
4.6. Infrared Thermography
A. thaliana (Col-0), A. stelleri, R. islandica, and T. salsuginea (Shandong) were grown in soil under normal conditions for 5 weeks to take infrared images of its root surface. To investigate the root temperature changes by low-oxygen stress, plants were exposed to a low oxygen gas mixture (0.1% O2/99.9% N2) for 12 h and the root temperature was measured immediately and 5 min after low-oxygen stress. The soil was quickly removed from roots of untreated or low-oxygen-treated plants before measuring the temperature of roots. The root temperature was measured from infrared images captured using ThermaCAM T355 infrared camera (FLIR Systems, Wilsonville, OR, USA), which are devices capable of sensing this radiation in the form of infrared light. The images were analyzed using ThermaCAM Quikplot software (FLIR Systems, Wilsonville, OR, USA). Root temperature measured from infrared images was compared between the control and the low-oxygen treated plants (ΔT = Ttest − Tcont). Infrared thermography measured root temperature with three different biological samples.
4.7. ITS PCR Amplification and Analysis
Genomic DNA was isolated from 1 g of fresh leaves using a genomic DNA extraction kit (RBC; Real Biotech Corporation, Taiwan). The genomic DNA was then used as a template for the PCR amplification of ITS using the two universal ITS primers described by White et al. (1990) [
52]. The ITS5 primer, 5′-TCCTCCGCTTAT TGATATGC-3′, was derived from the 3′ end of 18S rRNA, and the ITS4 primer, 5′-GGAAGTAAAAGTCGTAACAAGG-3′, was derived from the 5′ end of 25S rRNA. The ITS fragments of each species were amplified in a 20-μL reaction volume containing 2 μL of 10× PCR buffer, 2.5 mM each dNTPs, about 50 ng of genomic DNA, one unit of
Taq DNA polymerase (Takara EX, Shiga, Japan), 8 pmole primers in 30 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 30 s. After the PCR reaction, 5 μL of samples were analyzed on 1.4% agarose gel. The amplified ITS fragments were sequenced (Cosmogenetech co, Korea). The ITS DNA sequences were compared using the cluster algorithm of AliBee (
http://www.genebee.msu.su/services/malign_reduced.html).
4.8. Semi-Quantitative RT-PCR
Total RNA (2 μg) was extracted from the roots of 18~20 plants exposed to low oxygen for 0, 1, 3, 8, 24, or 72 h. Residual DNA was removed by treatment with DNase I (1U/ mg RNA)(Promega, Madison, WI, USA) for 30 min at 37 °C, and then DNase I was inactivated by incubation for 5 min at 72 °C. The RNAs were reverse-transcribed with Superscript
®II Reverse transcriptase (Invitrogen, Carlsbad, CA, USA). The target cDNA was PCR-amplified in a 20 μL reaction volume containing 2 μL of 10× PCR buffer, 25 mM of each dNTP, 1 μL of cDNA, and one unit of
Taq DNA polymerase (Takara, Shiga, Japan), plus 8 pmol of the appropriate primer sets (see
Supplementary Table S5) in 25 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 30 s. The amplified PCR products of the target gene were evaluated by 1.5% agarose gel electrophoresis in TBE buffer stained with ethidium bromide (Sigma-Aldrich, St. Louis, MO, USA). Semi-quantitative RT-PCR was repeated with three different biological samples and the representative results were presented. Gene expression level of RT-PCR result was measured from signal intensities on photograph of an agarose gel and normalized using by an internal control such as TUB2 and Actin2. The Pearson correlation coefficients (R
2) were calculated between the expression level of target genes in RT-PCR and microarray data.
4.9. Measurement of Photosynthetic Activity in Root
The maximum quantum yield of PSII (Fv/Fm) was measured in the roots of each species for comparison of photosynthetic activity in A. thaliana (Col-0) and A. stelleri. For measurement of Fv/Fm under normal growth conditions, A. thaliana (Col-0) and A. stelleri were grown on agar media in long-day conditions (16 h light/8 h dark) or darkness for 1, 2, or 3 weeks. For comparison of the effect of low-oxygen stress on Fv/Fm, two-week-old A. thaliana (Col-0) and three-week-old A. stelleri grown on agar plates were exposed to low oxygen (0.1% O2/99.9% N2) for 72 h. Before Fv/Fm measurement, all plants were dark adapted for 30 min. The Fv/Fm was measured using a hand-held fluorometer FluorPen FP100 (Photon Systems Instruments, Czech Republic). Each measurement was repeated three times with ten replications per experiment.