**Sodium Azide Priming Enhances Waterlogging Stress Tolerance in Okra (***Abelmoschus esculentus* **L.)**


Received: 6 September 2019; Accepted: 24 October 2019; Published: 25 October 2019

**Abstract:** Waterlogging stress adversely affects crop growth and yield worldwide. Effect of sodium azide priming on waterlogging stress tolerance of okra plants was investigated. The study was conducted as a field experiment using two weeks old plants grown from 0%, 0.02%, and 0.05% sodium azide (NaN3)-treated seeds. The waterlogging conditions applied were categorized into control, one week, and two weeks. Different growth and reproductive parameters were investigated. Activity and expression of antioxidant enzymes, root anatomy, and soil chemical analysis were also studied. Results showed that sodium azide priming inhibited germination. The germination percentages recorded were 92.50, 85.00, and 65.00 for 0%, 0.02%, and 0.05% NaN3-treated seeds, respectively, nine days after planting. Waterlogging conditions depressed plant height ten weeks after planting. Under waterlogging conditions, NaN3 promoted plant height and number of leaves formed. NaN3 also supported the survival of plants and formation of adventitious roots under waterlogging conditions. Waterlogging conditions negatively affected the redox potential, organic C, N, and P concentrations in the soil but enhanced Soil pH, Fe, Mn, Zn, and SO4. Under waterlogging conditions, NaN3 increased the average number of flower buds, flowers, and fruits produced in comparison to control. Moreover, NaN3 highly stimulated the development of aerenchyma which in turn enhanced the survival of okra plants under waterlogging conditions. NaN3 priming also enhanced the activities and gene expression level of antioxidant enzymes (ascorbate peroxidase, APX; catalase, CAT) under waterlogging conditions. In conclusion, this study demonstrated that NaN3 priming could improve waterlogging stress tolerance in okra.

**Keywords:** sodium azide; okra; waterlogging stress; antioxidants; gene expression

#### **1. Introduction**

Okra (*Abelmoschus esculentus* L.) is one of the economically important vegetable crops grown in tropical and sub-tropical regions of the world [1]. Okra originated in Ethiopia and was then reproduced in the Mediterranean area, North Africa, and India [1]. Environmental stresses negatively affect the growth, yield, and biological activities of plants worldwide [2–6]. In particular, waterlogging conditions influence the growth and yield of okra plants through causing hypoxic or anoxic conditions, which in turn affect various physiological processes in roots, including carbohydrate metabolism, gas exchanges, and water relations [7–9]. The oxygen-deficient soil environments may lead to changes in the composition and decomposition activities of microbes. Waterlogging conditions also affect soil factors such as EC, pH, soil structure, hydraulic conductance, porosity, and organics [10,11]. Plants could adapt to waterlogging conditions via activating their self-defense mechanisms and developing adventitious roots and hypertrophied stem bases with lenticels and aerenchyma cells [7,12]. Such aerenchyma cells could enhance organ porosity and root aeration [13,14]. These morphological features help plants to manage the low oxygen tension within the tissues, prevent anoxia, and maintain root functions and plant survival.

Applications of chemicals to plants, either as foliar or seed treatments, may induce their physiological mechanisms, leading to plant growth stimulation and stress tolerance [7,15,16]. For instance, seed pretreatment with salicylic acid enhances plant growth, antioxidant activities, and tolerance to harsh environmental factors such as heavy metal, herbicides, low temperature and salt stress [17,18]. Ethylene is also described as a signaling molecule in plants and has been projected as capable of inducing survival traits and tolerance under waterlogging conditions via up-regulating the activity of antioxidant enzymes and genes linked to aerenchyma formation, leaf senescence, adventitious roots, and epinasty [7,19–21]. However, ethylene application as a proactive measure for ameliorating envisaged waterlogging condition on a wide scale may not be appreciated. Hence, seed priming techniques may be easier to enhance growth and yield. Sodium azide (NaN3) has been successfully used for creating genetic variability and enhancing agronomic traits of crop plants. It affects crops based on the concentration applied. Gnanamurthy et al. [22] and Shagufta et al. [23] reported that NaN3 priming delayed and inhibited the germination of maize and fenugreek, respectively. However, Vwioko and Onobun [24] reported that NaN3 enhanced the germination percentage and height of okra plants. Al-Qurainy [25] and Zuzana et al. [26] also stated that NaN3 stimulated the plant height of *Eruca sativa* and *Diospyros lotus*, respectively. On the other hand, Adamu and Aliyu [27] and Gnanamurthy et al. [22] revealed that NaN3 priming inhibited plant height. NaN3 priming also regulates various physiological and molecular mechanisms in plants and modulates the activities of catalase, peroxidase, and cytochrome oxidase [28]. Molecular changes induced by NaN3 treatments produce mutations by base substitution, leading to changes in amino acid sequences. NaN3 is reckoned to be an efficient reagent that induces a broad and high variation of morphological and yield parameters in cultivated species. However, it is not popularly used to initiate plants tolerance to environmental factors. Environmental stresses such as salinity and water stress [29,30] increase production of free radical in plants, and resistance to the unfavorable conditions often involves stimulation of the antioxidant response. Haq et al. [31], El Kaaby et al. [32], and Kuasha et al. [33] carried out in vitro studies on the ability of NaN3 to confer salt tolerance in plants. Haq et al. [31] stated that one of the three cultivars of sugarcane studied regenerated plantlets that were salt tolerant, while El Kaaby et al. [32] and Kuasha et al. [33] stated that NaN3 depressed the responses of the explants of tomato and sugarcane to salinity stress. Salim et al. [34] also studied the effect of NaN3 on various plant traits, including disease resistance, yield, antioxidant activities, pigmentations, and salinity and drought stress tolerance. However, the role of NaN3 in regulating waterlogging stress responses has not been studied yet. Therefore, the main aim of the present study was to assess the ability of NaN3 to induce waterlogging stress tolerance in okra plants.

#### **2. Materials and Methods**

#### *2.1. Plant Material and Application of Sodium Azide Treatments*

Seeds of okra variety Clemson spineless produced by Technism (Longué-Jumelles, France) were obtained and used in this study. Okra seeds were soaked in sodium azide treatments, i.e., 0%, 0.02%, and 0.05% (*w*/*v*), at room temperature (27 ◦C) for 5 h with a continuous gentle stirring. After 5 h, the seeds were removed and washed 5 times with deionized water to remove all traces of NaN3. NaN3 treatments were classed as mild (0.02%) and severe (0.05%).

#### *2.2. Soil Preparation for Potted Field Experiment*

Top soil (0–15 cm deep) was collected from the Demonstration Farm, Faculty of Agriculture, University of Benin, Nigeria. The soil type is categorized as ultisol. The composite soil sample was air-dried for three weeks and sieved to remove gravel and other particles. Each experimental pot was filled with 5 kg of soil. Thirty-six (36) pots were prepared to make twelve pots for each NaN3 treatment. The undersides of the experimental pots were not perforated so that they could retain water.

#### *2.3. Sowing of Seeds in Nursery Beds, Transplanting into Experimental Pots, and Acclimatization*

Twelve soil nursery beds (measuring 2 feet by 2 feet) were prepared for the sowing seeds. The beds were allocated to the treated seeds, i.e., 0%, 0.02%, and 0.05% NaN3. The seeds were sown at a depth of 2–3 cm. Germination records were collected every day for two weeks. After two weeks in the nursery beds, four plants were transferred into each experimental pot and taken to the open field. The plants were allowed to acclimatize for another two weeks in the field before flooding condition was introduced.

#### *2.4. Application of Flooding or Waterlogging Conditions*

When the plants were four weeks old, flooding of experimental pots with tap water was carried out. Three conditions of flooding or waterlogging were set up; no flooding (NF), one-week flooding (1 WF), and two weeks flooding (2 WF). Flooding of the pots was done up to 2 cm mark above the soil level. The water level was maintained in each pot by topping daily after inspection during the period.

#### *2.5. Growth Parameters Measured*

The field data collected were germination percentage, stem girth, plant height, number of leaves formed, survival percentage of plants, number of adventitious roots formed, number of flower buds formed, number of flowers, and number of fruits produced.

#### *2.6. Soil Chemical Analyses*

Soil chemical factors like pH, electrolyte conductivity (EC), redox potential (Eh), nitrogen, phosphorus, sulphate, organic carbon, iron, manganese, zinc, and total soluble phenolics were determined using standard methods. The soil analysis was carried out for the soil samples collected after plant harvested. pH, EC, and Eh were estimated in a soil-water slurry (ratio 1:3) [35]. Total nitrogen was estimated following Kjeldahl method [36]. Total soluble phenolic analysis was done based on the modified citrate extraction protocol followed by Folin–Ciocalteau colorimetric methodology [37]. The methodologies of Appiah and Ahenkorah [38] and Ben Mussa et al. [39] were used to determine sulphate content. Phosphorus measurement was conducted following the methodology of Bray and Kurtz [40]. Walkley–Black chromic acid wet oxidation methodology [41] was used to estimate the organic carbon. Iron content was determined following the hydroxylamine and 1,10- phenanthroline protocol [42]. Manganese was determined following the permanganate oxidation procedures [42]. The determination of zinc was carried using atomic absorption spectrophotometer (Shimadzu Europa GmbH, Duisburg, Germany).

#### *2.7. Soil Microflora Counts*

Presence of bacteria and fungi in the soil samples was investigated after plant harvest. Serial dilution processes were used in the analysis of soil microflora. Ten grams of the samples were dispensed into sterile beakers and mixed thoroughly with 90 mL sterile distilled water. Each sample was serially diluted from the stock sample and then transferred to the first tube 9 mL of sterilized water

to give 10−1dilution, from which further dilution up to 10−<sup>4</sup> was made. The pour plate method was utilized for inoculation on a sterilized nutrient agar (NA) or potato dextrose agar (PDA), impregnated with antifungal or antibacterial agents for the growth of bacterial or fungal isolates, respectively. Nutrient agar plates were kept for 24–48 hrs at 37 ◦C for bacterial growth. Potato dextrose agar was incubated at room temperature (30 ± 2 ◦C) for 3–5 days. Total viable colonies were then counted for the microbial isolates and represented in terms of colony forming units (cfu/g). Viable counts obtained were recorded with reference to the serial dilution used [43,44].

#### *2.8. Root Anatomy*

Harvested plant roots were washed and used to make microscopic slides to examine internal tissues. Root sections were immersed in paraffin wax and left to solidify. Sections were cut and dewaxed by clamping in the microtome. Aniline blue stain was applied to the sections to show a clear contrast of air spaces (aerenchyma) formed. Excess stains were removed by ethanol before oven-drying. Following oven-drying, slides were viewed and then photographed using the microscope IRMECO model IM-660 T1 (IRMECO GmbH & Co. KG, Geesthacht, Germany) with a camera connected to PC. Observations were done under X10 objective lens.

#### *2.9. Antioxidant Enzyme Assays*

Activities of catalase (CAT) and ascorbate peroxidase (APX) were determined in the leafy tissue of the NF, 1 WF, and 2 WF plants treated with 0%, 0.02%, and 0.05% NaN3 collected at the tenth week after planting following the method of Zhang and Kirkham [45]. In brief, 0.25g of leafy tissue was homogenized in 3 mL of solution, composed of PBS (50 mM), EDTA (0.2 mM), and 1% PVP, and centrifuged. Supernatants were assayed to detect the absorbance at 290 nm (for APX) and 240 nm (for CAT).

#### *2.10. RNA Isolation, cDNA Synthesis, and Quantitative RT-PCR*

Quantitative real-time PCR (qRT-PCR) assay was conducted to evaluate the expression level of antioxidant enzyme-encoding genes (*APX*, *CAT*) in the leafy tissue of the NF, 1 WF, and 2 WF plants treated with 0%, 0.02%, and 0.05% NaN3 collected at the tenth week after planting. Total RNA samples were isolated from the tissue following Qiagen RNeasy Plant Mini kit. DNA removal and cDNA synthesis were performed using Qiagen RNase-Free DNase Set and Qiagen Reverse Transcription kit, respectively. Quantitative RT-PCR was performed following Qiagen QuantiTect SYBR Green PCR kit protocol. PCR conditions, housekeeping gene, and gene-specific primers were used as reported by Vwioko et al. [7]. The primer pair 5 -TGCCCTTCTATTGTGGTTCC-3 and 5 -GATGAGCACACTTTGGAGGA-3 was used for *CAT* amplification, whereas the primer pair 5 -ACCAATTGGCTGGTGTTGTT-3 and 5 -TCACAAACACGTCCCTCAAA-3 was used for *APX* amplification. The primer pair 5 -TTCCTTGATGATGCTTGCTC-3 and 5 -TTGACAGCTCTTGGGTGAAG-3 was used for the housekeeping gene (*UBQ1*) amplification.

#### *2.11. Statistical Analysis*

Mean and standard deviation were measured for the data obtained for the different traits measured. Two-way analysis of variance was conducted using NaN3 treatments and flooding conditions as factors. Tukey's test was conducted to determine the significance of values. Statistical analyses were performed using SPSS ver. 19 (SPSS Inc., Chicago, IL, USA).

#### **3. Results**

#### *3.1. Germination of NaN3-Treated Seeds*

The germination was first recorded for okra seeds given control (0%) treatments 2 days after planting (2 DAP). Germination was recorded for 0.02 and 0.05% NaN3-treated seeds three days after planting (3 DAP). Eight days after planting (8 DAP), the highest and least percentage of germination were recorded for 0% and 0.05% NaN3-treated seeds, respectively (Figure 1). Twenty-four hours delay in germination was recorded for the NaN3-treated seeds.

**Figure 1.** Percentage of germination of NaN3-treated okra seeds sown in nursery. Values = mean ± SD, *n* = 4. Mean values with similar letters at the same day after planting (DAP) are not significantly different at *p* ≤ 0.05.

#### *3.2. Plant Height*

Values obtained for plant height showed that non-waterlogged plants produced the highest values irrespective of the NaN3 treatment given to the seeds ten weeks after planting (10 WAP). For example, mean values obtained for plant height were 31.5, 29.5, and 31.1 cm for 0%, 0.02%, and 0.05%, respectively, under non-waterlogging condition, 10 WAP (Table 1). Under one-week waterlogging condition, the values recorded for 0%, 0.02%, and 0.05% were 15.2, 21.8, and 19.4 cm, respectively, 10 WAP. Similarly, under two weeks waterlogging conditions, the values recorded for 0%, 0.02%, and 0.05% were 16.3, 22.4, and 19.9 cm, respectively, 10 WAP, indicating growth stimulations for plants grown from 0.02% and 0.05% NaN3-treated seeds.


**Table 1.** Height (cm) of okra plants grown from NaN3-treated seeds subjected to different waterlogging conditions four weeks after planting (WAP).

Values = mean ± S.D., *n* = 4, WAP = weeks after planting. Mean values with similar letters as superscript in one column are not significantly different at *p* ≤ 0.05.

#### *3.3. Stem Girth*

The highest stem girth values were obtained for okra plants grown under non-waterlogging conditions (Table 2). Ten weeks after planting, the values recorded for the stem girth of okra plants grown under two-week waterlogging conditions were statistically significant compared to those recorded for plants grown under and non-waterlogging conditions (Table 2).



Values = mean ± S.D., *n* = 4, WAP = weeks after planting. Mean values with similar letters as superscript in one column are not significantly different at *p* ≤ 0.05.

#### *3.4. Number of Leaves Formed, Number of Adventitious Roots Produced, and Percentage of Survival of Plants*

The total number of leaves formed per plant recorded indicated that the plants grown under non-waterlogging condition produced the highest number of leaves 10 WAP. The combination of waterlogging conditions and NaN3 treatments gave higher values for number of leaves formed than when the waterlogging condition is applied only (Table 3). For example, total number of leaves under non-waterlogging conditions were 16, 16.5, and 16.5 for 0%, 0.02%, and 0.05%, respectively. Whereas in one-week waterlogging conditions, values were 13, 14, and 15 for plants grown from 0%, 0.02%, and 0.05% NaN3-treated seeds.


**Table 3.** Number of leaves, average number of adventitious roots produced, and survival percentage of okra plants grown from NaN3-treated seeds under waterlogging conditions 10 WAP.

Values = mean ± S.D., *n* = 4, WAP = weeks after planting. Mean values with similar letters as superscript in one column are not significant different at *p* ≤ 0.05.

Plants did not form adventitious roots under non-waterlogging conditions. However, the production of adventitious roots was observed in plants subjected to waterlogging condition. Plants subjected to two weeks of waterlogging condition initiated higher numbers of adventitious roots (Table 3). Furthermore, plants grown from 0.05% NaN3-treated seeds produced the highest number of adventitious roots recorded. The combination of NaN3 concentration and waterlogging condition supported the greater production of adventitious roots in okra.

Ten weeks after planting, the number of plants that survived the waterlogging conditions is shown in Table 3. Higher percentage of survival was recorded with the combination of sodium azide and waterlogging condition. For example, under two weeks waterlogging condition, the percentage of survival of okra plants were 25, 33.3, and 50 for 0%, 0.02%, and 0.05% NaN3-treated seeds, respectively. Similarly, for one-week waterlogging condition, percentage of survival of okra plants were 33.3, 50, and 50 for 0%, 0.02%, and 0.05% NaN3-treated seeds, respectively.

#### *3.5. Number of Flower Buds, Flowers, and Fruits Produced*

The number of flower buds, flowers, and fruits are shown in Table 4. The waterlogging condition caused a decrease in all the reproductive parameters considered. For example, the average number of flower buds recorded for plants grown from control seeds (0.00% NaN3 treatment) were 5.5, 2.75, and 1.75 for NF, 1 WF, and 2 WF conditions, respectively. Similarly, average number of flowers recorded for the same plants were 5, 2, and 1, respectively. Moreover, the average number of fruits recorded for the same plants were 4.5, 1.25, and 0.5, respectively. The average number of flower buds, flowers, and fruits recorded for plants grown from 0.05% NaN3-treated seeds and subjected to waterlogging conditions were higher than those recorded for non-treated plants.


**Table 4.** Average number of flower buds, flowers, and fruits formed per plant of okra grown from NaN3 treated seeds subjected to waterlogging conditions ten weeks after planting.

Values = mean ± S.D., *n* = 4. Values with similar letters as superscript are not significantly different.

#### *3.6. Soil Microflora Counts*

The average values obtained for bacteria and fungi counts are shown in Table S1. The bacterial counts were higher than fungal counts in all soil samples analyzed. The bacterial count values were higher in soils collected from waterlogging condition, while the fungal count values were higher in soils collected from non-waterlogging condition. Soils collected from two-week waterlogging conditions gave the least fungal counts.

#### *3.7. Soil Chemical Analysis*

There were clear differences in many of the soil chemical parameters analyzed between soil samples collected from non-waterlogging and waterlogging experimental pots (Table S2). The differences in values obtained shows a regular pattern. For example, pH values for NF were 6.0–6.1 while higher values were recorded for 1 WF and 2 WF. Redox potential (Eh) values were consistently higher for NF than 1 WF and 2 WF. Soil Eh ranged from 23.60–24.10 for NF and 7.2–7.4 for 1 WF and 2 WF. The highest values of sulphate ion (SO4) concentrations and electrolyte conductivity (EC) readings were observed in 1 WF soil samples. Mean values for non-treated soil EC were 228, 413, and 125 μS/cm for NF, 1 WF, and 2 WF, respectively. Similarly, mean values for SO4 concentration in non-treated soil were 0.52, 1.13, and 0.80 mg/Kg for NF, 1 WF, and 2 WF, respectively. Organic carbon, total nitrogen and available phosphorus contents in soil followed the same reduction pattern under one- and two-week waterlogging conditions. Approximately, 10-fold reductions in organic carbon and total nitrogen contents were observed under waterlogging conditions. The records for soil metallic factors like Fe, Zn, and Mn showed the same pattern where the values were higher in soil samples collected from one- and two-week waterlogging conditions. Mean values obtained for Fe were 116.3, 242.1, and 243.3 mg/kg for NF, 1 WF, and 2 WF, respectively, for soil samples collected from pots where 0% NaN3 plants were grown. The mean values recorded for Zn in soil samples collected from pots containing 0% NaN3 plants were 14.2, 22.7, and 35.4 mg/kg for NF, 1 WF, and 2 WF, respectively. The mean values of Mn in

the same soil samples were 1.34, 9.68, and 12.9 mg/kg for NF, 1 WF, and 2 WF, respectively. The mean values of total phenol content show low variation.

#### *3.8. Anatomy of Okra Roots*

There were structural differences in the anatomy of okra root sections obtained from non-waterlogged and waterlogged plants (Figures 2–4). The presence of air channels (lacunae) was conspicuously absent in non-waterlogged root sections (Figure 2). The development of aerenchyma in the cortex and stele were very conspicuous in root sections of plants subjected to waterlogging conditions (Figures 3 and 4). Furthermore, the aerenchyma cells observed in root sections of waterlogged plants were larger in plants grown from 0.05% NaN3-treated seeds than those from 0.02% NaN3-treated seeds (Figures 3 and 4). This suggests an explanation for the higher percentage of survival recorded for plants grown from 0.05% NaN3-treated seeds. The walls of the aerenchyma cells are thick to prevent their collapse.

**Figure 2.** Root sections of okra plants grown from different concentrations of NaN3-treated seeds show no aerenchyma cells formed under non-waterlogging conditions. (**A**) 0% NaN3, (**B**) 0.02% NaN3, (**C**) 0.05% NaN3.

(**A**)

(**B**)

**Figure 3.** Root sections of okra plants grown from different concentrations of NaN3-treated seeds show aerenchyma cells formed under one-week waterlogging conditions. (**A**) 0% NaN3, (**B**) 0.02% NaN3, (**C**) 0.05% NaN3.

(**A**)

**Figure 4.** Root sections of okra plants grown from different concentrations of NaN3-treated seeds show aerenchyma cells formed under two-week waterlogging conditions. (**A**) 0% NaN3, (**B**) 0.02% NaN3, (**C**) 0.05% NaN3.

(**C**)

#### *3.9. Antioxidant Enzymes Activity and Gene Expression Analyses*

The effects of the waterlogging condition and NaN3 treatments on the activities and expression levels of antioxidant enzymes (APX, CAT) in the leaf tissues were investigated. The activity and expression level of APX enzyme were significantly enhanced in plants exposed to waterlogging and sodium azide treatments with respect to non-treated (control) plants (Figure 5). Additionally, under waterlogging conditions, the activity and expression level of CAT enzyme were slightly enhanced in plants treated with sodium azide, as compared to non-treated plants (Figure 5).

**Figure 5.** Activity (**A**) and gene expression levels (**B**) of APX and CAT enzymes in okra plants grown from NaN3-treated seeds under waterlogging conditions ten weeks after planting (WAP). Values = mean ± SD, *n* = 4. Mean values with similar letters at the same WAP are not significantly different at *p* ≤ 0.05.

#### **4. Discussion**

Waterlogging stress has adverse impacts on crop development and productivity. Waterlogginginduced oxygen depletion resultsin changesin plantmorphology andmetabolism. Waterlogging conditions also cause inhibition of photosynthesis, leaf chlorophyll degradation, and early leaf senescence [46]. Negative impacts of flooding might be due to the reduced level of gas diffusion in water, which does not allow terrestrial plants to survive for a long period. Plants develop specific traits to improve gas exchange and cope with waterlogging conditions. These traits include formation of adventitious roots and aerenchyma cells, as well as elongation of stem root juncture above the water surface. These efficiently ameliorate the stress-induced hypoxic or anoxic conditions. The presence of aerenchyma cells facilitates exchange of gases between aerial and submerged plant parts [47]. Kawai et al. [48] proposed that the development of aerenchyma in tissues and organs decreases the number of cells requiring oxygen for respiration. However, the development of adaptive traits to waterlogging stress is species-dependent [7,49,50]. Enhanced formation of aerenchyma was observed upon treating rice plants with exogenous ethylene [14].

In the present study, NaN3 treatments enhanced waterlogging stress tolerance and aerenchyma formation in okra. The results also showed that NaN3 treatments affected okra germination. NaN3-caused seed germination inhibition has also been reported in different plant species [22,23,51,52]. However, NaN3 stimulated the germination of okra plants [24]. This germination inhibition was dependent on the concentration of NaN3 used as seed treatment. Three days after planting (3 DAP), germination has been recorded in all NaN3 treatments applied. Under waterlogging conditions, NaN3 promoted okra growth 10 WAP, indicating that plants grown from 0.02% NaN3-seed treatments exhibited better performance than those grown from 0.05% NaN3-seed treatments. These findings were in a harmony with that reported by Al-Qurainy [25] and Zuzana et al. [26] who stated that NaN3 could stimulate the plant growth and height of *Eruca sativa* and *Diospyros lotus*, respectively. Moreover, the difference in the number of leaves formed under waterlogging and non-waterlogging conditions was significant. Plants grown from NaN3-treated seeds formed more leaves than those from non-treated seeds. Additionally, plants that were grown from 0.02% and 0.05% NaN3-treated seeds produced a greater number of adventitious roots under waterlogging conditions. The emergence of adventitious roots is preceded by epidermal cell death at the nodes of submerged rice plants [47]. The activities leading to epidermal cell death for the emergence of adventitious roots occurred more in plants grown from NaN3-treated seeds. Waterlogging conditions negatively affected the reproductive parameters recorded for okra plants in the current study. These findings are in harmony with that reported by Vwioko et al. [7]. Plants grown from 0.05% NaN3-treated seeds formed a higher number of buds than plants produced from 0% NaN3-treated seeds subjected to two-week waterlogging conditions. Plants grown from 0.05% NaN3-treated seeds also produced more fruits than the control plants under two-week waterlogging conditions.

Waterlogging conditions cause depletion of soil oxygen due to microbial respiration. The reduction of soil oxygen urges anaerobic microorganisms to shift to alternative electron acceptors for their metabolic requirements [53]. Bacteria and fungi ratio in soil community are altered whenever there are soil inundations. Soil bacteria and fungi have a critical role in decomposition and nutrient cycling [54]. In the current investigation, microbial count results exhibited an increase in the bacteria populations and reduction in the fungi populations. The decrease in fungi populations has been previously reported [53,55–57]. Therefore, under waterlogging conditions, fungi presence is less prevalent than bacteria. Fungi require aerobic conditions to thrive but are inhibited by the scarcity of oxygen in the flooded soil environments. Fungi germinate from spores under flooding slowly, resulting in a decreased colonization. Unger et al. [53] suggested that some microbial groups may thrive well under flooded conditions. Gram-positive bacteria showed higher levels compared to Gram-negative bacteria under waterlogging conditions. Mentzer et al. [57] reported that flooding exhibited greater effect than nutrient loading on the microbial community and profoundly altered the composition and functional components.

Water copiously influences several physicochemical processes in soil, particularly under flooded conditions. This begins with the cutoff of oxygen supply to soil environments under waterlogging stress. The lack of oxygen promotes anaerobic metabolism by microbes through utilizing a decomposable organic matter. A reduction in soil redox potential and an increase in pH are recorded [58]. The soil Eh data recorded in a soil-water suspension rightly predicts the level of transformations present in the waterlogged soil [59]. Other important chemical changes in flooded soils indicate the prevalence of reduced forms of nitrogen, oxygen, iron, manganese, or sulphur in soil [53]. There are changes in phase or solubility because of redox reactions. For example, nitrate-nitrogen is transformed into gaseous forms (N2, NO2, N2O) and lost, resulting in nitrogen depletion of soil [60]. In the present study, the soil chemical analysis showed that waterlogging conditions increased pH towards neutral, reduced soil Eh, organic carbon, total nitrogen and available phosphorus. These soil factors indicate higher reduction-oxidation reactions in soils under waterlogging conditions. These patterns of chemical environments and transformations are suspected to favor the tolerant bacteria for their higher counts recorded in waterlogged soil samples. The chemical environments attained under waterlogging

soil conditions met the metabolic needs of tolerant bacteria. The decomposition of complex organic compounds is slow under anoxic conditions and in some cases leads to detection of higher amounts of phenolics [53] in waterlogged than in non-waterlogged soils. The present study does not reveal changes in the total phenolics of soil samples, suggesting that either the soil is devoid of complex organics for microbes to degrade under waterlogging conditions, or the microbes utilized readily available forms of carbons that are root exudates. Carbon enters the soil profile via the decomposition of plant residue on the surface or via root exudates in the upper soil horizon [53].

In the current study, root anatomy showed some peculiar features with waterlogged plants. Plants did not develop air-chambers in the cortex and stele regions under non-waterlogging conditions. However, plants subjected to waterlogging conditions formed aerenchyma cells. Further examination of the micrographs showed that plants grown from NaN3-treated seeds produce more aerenchyma cells than those grown from untreated seeds. It was evident that 0.05% NaN3-treated seeds produce plants with the highest aerenchyma development and increased with increasing the duration of waterlogging conditions. The formation of aerenchyma in the root as an adaptive trait contributed to the survival of okra plants exposed to waterlogging conditions. Furthermore, under waterlogging conditions, the activities and expression levels of APX and CAT enzymes were enhanced in plants treated with NaN3 compared to non-treated plants in the present study. The survival of plants in stressed environments might be attributed to the induction of expression levels of antioxidant compounds. Salim et al. [34] reported that NaN3-treated seeds produce mutant plants that showed higher antioxidation capacities than the normal plants. Moreover, Jeng et al. [61] revealed that these mutants induced increased antioxidant capacities through the generation of scavenging metabolics (DPPH, LPI ability, FRAP, and ABTS radical scavenging activities) than the wild type. In addition, the antioxidant enhancements could be linked to the accumulation of phenolics, anthocyanin, and proanthocyanidins at higher levels in the seed coats. These results are in harmony with that reported by Elfeky et al. [62] who stated that *Helianthus annus* plants grown from NaN3-treated seeds initiated and induced higher antioxidant capacities than those grown from untreated seeds via increasing carotenoids, peroxidase activity, and protein content. In conclusion, sodium azide priming could enhance waterlogging stress tolerance in okra plants through enhancing the growth and reproductive parameters, inducing the formation of adventitious roots and aerenchyma cells, and increasing the activities and gene expression level of antioxidant enzymes.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4395/9/11/679/s1, Table S1: Total bacteria and fungi count of soil samples analyzed after plant harvest, Table S2: Values obtained for soil factors in soil samples collected from different experimental pots after plant growth under waterlogging conditions.

**Author Contributions:** M.A.E.-E. and E.D.V. designed and performed the experiments, analyzed the data, and wrote and revised the manuscript. M.E.I., A.A.A.-G., H.M.A., E.A.A., and M.A.A.-D. helped with analysis and revision of the manuscript. M.M.E.-S. revised the manuscript. All the authors approved the final version of the manuscript.

**Funding:** The authors would like to extend their sincere appreciation to the Deanship of Scientific Research at King Saud University for funding this Research group no. RG 1440-054. The authors would also like to thank University of Benin in Nigeria and Tanta University in Egypt for supporting this work.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **E**ff**ect of Heat Stress on Growth and Physiological Traits of Alfalfa (***Medicago sativa* **L.) and a Comprehensive Evaluation for Heat Tolerance**

#### **Misganaw Wassie 1,2, Weihong Zhang 1, Qiang Zhang 3, Kang Ji 1,2 and Liang Chen 1,\***


Received: 28 August 2019; Accepted: 27 September 2019; Published: 28 September 2019

**Abstract:** Alfalfa (*Medicago sativa* L.) is a valuable forage legume, but its production is largely affected by high temperature. In this study, we investigated the effect of heat stress on 15 alfalfa cultivars to identify heat-tolerant and -sensitive cultivars. Seedlings were exposed to 38/35 ◦C day/night temperature for 7 days and various parameters were measured. Heat stress significantly reduced the biomass, relative water content (RWC), chlorophyll content, and increased the electrolyte leakage (EL) and malondialdehyde (MDA) content of heat-sensitive alfalfa cultivars. However, heat-tolerant cultivars showed higher soluble sugar (SS) and soluble protein (SP) content. The heat tolerance of each cultivar was comprehensively evaluated based on membership function value. Cultivars with higher mean membership function value of 0.86 (Bara310SC) and 0.80 (Magna995) were heat tolerant, and Gibraltar and WL712 with lower membership function value (0.24) were heat sensitive. The heat tolerance of the above four cultivars were further evaluated by chlorophyll *a* fluorescence analysis. Heat stress significantly affected the photosynthetic activity of heat-sensitive cultivars. The overall results indicate that Bara310SC and WL712 are heat-tolerant and heat-sensitive cultivars, respectively. This study provides basic information for understanding the effect of heat stress on growth and productivity of alfalfa.

**Keywords:** alfalfa; evaluation; growth; heat stress; physiological traits

#### **1. Introduction**

Heat stress is one of the major abiotic stresses limiting plant growth and development. When plants are exposed to high temperature, several cellular injuries, including cell death, may occur within minutes, which then leads to an appalling failure of cellular organization [1,2]. In addition, the rapid closure of stomata, reduction in cell size, an increase in stomatal, trachomatous densities, and xylem vessels of both root and shoot were reported to occur in response to heat stress [3]. However, different plant species may show different responses to heat stress [4]. Generally, heat stress triggers various morphological, physiological, biochemical, and molecular changes to inhibit plant growth and development. Heat stress inhibits seed germination; causes scorching; twigs and burning of leaves, branches, and stems; leaf senescence and abscission; shoot and root growth inhibition; fruit discoloration and damage; reduced yield; and finally plant death [5,6]. High temperature stress also affects shoot net assimilation and decreases the overall dry weight of the plant [4].

It is well established that heat stress has detrimental impacts on various key physiological, biochemical, and metabolic processes of plants, and disrupts normal cellular homeostasis [2]. It

promotes the overproduction and accumulation of reactive oxygen species (ROS), malondialdehyde (MDA) production due to lipid peroxidation, photoinhibition, protein denaturation, and accumulation of compatible solutes [7–9]. The oxidative stress caused by heat stress further leads to cellular injury, membrane proteins breakage, lipid peroxidation, photosynthetic pigment degradation, and enzymes and nucleic acid denaturation [4,10,11]. Furthermore, heat stress influences plant photosynthesis and respiration processes to curtail the life cycle and reduce plant productivity [12]. The heat stress sensitivity of plants varies with the plant genotype and the stage of plant development, but the effect is highly dependent on genotype and species, as well as with abundant inter- and intraspecific variations [12].

Alfalfa (*Medicago sativa* L.) is one of the most important perennial forage legume species. Due to its outstanding nutritional quality, alfalfa is an excellent source of feed nutrient for animals. High temperature stress is a limiting factor for alfalfa cultivation [13,14]. Previous studies have shown that increasing temperature above optimal level markedly affects alfalfa's morphological, physiological, and proteomic processes, and reduced photosynthetic rate, destroyed plasma membrane structure, and accelerated the process of aging [15–18]. Therefore, it is of a great significance to develop heat stress-tolerant alfalfa cultivar that withstand heat stress-induced growth inhibition and biomass reduction. Studying plants' physiology in response to heat stress could be helpful to further understand the molecular tolerance traits [3] and will provide fundamental knowledge to develop heat-tolerant cultivars. It is well reported that different genotypes of a single plant species demonstrate high degrees of variation for heat tolerance; therefore, the selection of varieties with high thermotolerance potential is crucial to further improve thermotolerance. The genetic variability present in alfalfa could be exploited to evaluate and screen for high-temperature tolerance. Moreover, compared with other abiotic stresses reports in alfalfa, studies about the effect of heat stress on alfalfa growth and physiology are very limited. Thus, the objectives of this study were to investigate the effect of heat stress on the growth and physiological traits of alfalfa, to evaluate for heat tolerance, and to identify heat-tolerant and heat-sensitive alfalfa cultivars.

#### **2. Materials and Methods**

#### *2.1. Plant Materials, Growth Conditions, and Heat Treatment*

In this study, fifteen alfalfa cultivars (*Medicago sativa* L.) were used and the details of the cultivars are presented in Table 1. Ten seeds of each cultivar were sown in each plastic pot filled with clay, sand, and loamy soil (1:1:2, v/v). The pots were then kept in a greenhouse with a temperature of 25 ◦C, relative humidity of approximately 60%, light intensity of 500–550 μmol m−<sup>2</sup> s<sup>−</sup>1, and a photoperiod of 14 h/10 h light/dark [19]. There were five replications for each cultivar and treatment. The seedlings were watered daily to field capacity level and fertilized once a week with a half-strength Hoagland nutrient solution. The four-week-old seedlings were divided into two groups for heat treatment. The control group was kept in a greenhouse at 25 ◦C and the treatment group was transferred into a growth incubator and treated at 38/35 ◦C light/dark and light intensity of 500–550 μmol m−<sup>2</sup> s−<sup>1</sup> for 7 days.


**Table 1.** Alfalfa cultivars used in the study.

#### *2.2. Plant Biomass Measurement*

After seven days of heat stress treatment, five plants for each of the treatment and control group were randomly harvested from each pot (25 plants for each cultivar and treatment). The roots of selected plants were washed with distilled water and separated from the shoot. The fresh weight of roots and shoots were measured separately using analytical balance (precision 0.0001 g). Dry weight was measured after drying the shoot and root in an oven at 80 ◦C for 24 h. Total fresh and dry biomasses were calculated using the following formula:

$$\begin{aligned} \text{TFB} &= \text{SFW} + \text{RFW} \\ \text{TDB} &= \text{SDW} + \text{RDW} \end{aligned} \tag{1}$$

where TFW is the total fresh biomass, SFW is the shoot fresh weight, RFW is the root fresh weight, TDB is the total dry biomass, SDW is the shoot dry weight, and RDW is the root dry weight.

#### *2.3. Physiological Traits Analysis*

Fresh leaves were used to measure relative water content, electrolyte leakage, chlorophyll content, and soluble sugar content, while MDA and soluble protein content were determined from liquid nitrogen dried leaves which were stored at −80 ◦C freezer.

#### *2.4. Relative Water Content (RWC) Measurement*

For RWC determination, fresh weight (FW) was measured from the fully expanded top leaves, and Turgid weight (TW) was recorded after dipping the leaves in distilled water for 12 h. Dry weight (DW) was then measured after oven drying the leaves at 80 ◦C for 24 h and RWC was calculated using the following formula:

$$\text{RWC}\% = ((\text{FW} - \text{DW})/(\text{TW} - \text{DW})) \times 100 \tag{2}$$

where FW is the fresh weight of leaves, DW is the dry weight of leaves, and TW is the turgid weight of leaves.

#### *2.5. Chlorophyll Content Determination*

For chlorophyll content determination, 0.1 g fresh alfalfa leaves were placed into a centrifuge tube containing 5 mL of 95% alcohol. The test tubes were wrapped with aluminum foil and incubated for 48 h at room temperature in the dark. The absorbance of the chlorophyll extract was read at 665 and 649 nm. The chlorophyll content was calculated according to the following formula:

$$\text{Chl a (mg\cdot g}^{-1}\text{ FW}) = (13.95 \times \text{D665} - 6.88 \times \text{D649}) \times 0.005 + \text{W} \tag{3}$$

$$\text{Chl }\text{b }(\text{mg}\cdot\text{g}^{-1}\text{ FW}) = (24.96 \times \text{D649 } - 7.32 \times \text{D665}) \times 0.005 + \text{W} \tag{4}$$

$$\text{Total Chl (mg\cdot g}^{-1}\text{ FW)} = (18.08 \times \text{D649} + 6.63 \times \text{D665}) \times 0.005 + \text{W} \tag{5}$$

where D is the absorbance of the chlorophyll extract and W is the fresh weight leaves (g).

#### *2.6. Electrolyte Linkage (EL) Measurement*

The electrolyte leakage was measured using the method described by Huang et al. [20]. Briefly, 0.5 g of fresh alfalfa leaves were collected and washed with deionized water three times and then transferred into a 50 mL plastic centrifuge tube that was filled with 15 mL of deionized water. The tubes were incubated at room temperature for 12 h on a conical shaker and initial conductivity (EL1) using a conductivity meter (JENCO-3173, Jenco Instruments, Inc., San Diego, CA, USA). To release all electrolytes, the leaves were killed by autoclaving at 121 ◦C for 30 min. The tubes were then cooled at

room temperature and the second conductivity (EL2) was measured. The relative EL was calculated by using the formula:

$$\text{Relative EL (\%)} = \text{(EL1/EL2)} \times 100.\tag{6}$$

#### *2.7. Soluble Sugar Content Determination*

For soluble sugar content determination, 0.1 g fresh leaves were placed into a test tube containing 5 mL distilled water. The tubes were then sealed with plastic film and soaked in the boiling water bath for 30 min. The supernatant was collected and extraction continued for the second time using the residue. The supernatant from the primary and secondary extraction was mixed together, and 0.5 mL extract was taken and added to each tube containing 1.5 mL distilled water. Then 0.5 mL anthrone reagent (1 g anthrone dissolved in 50 mL ethyl acetate) and 5 mL 98% (W/V) H2SO4 (NA) was added to the test tubes and mixed gently, then placed in a boiling water bath (100 ◦C) for 10 min. The tubes were then cooled rapidly under running cold water and the absorbance was measured at 620 nm against the blank reagent. The concentration of total soluble sugar was obtained from the glucose standard curve. Finally, the total soluble sugar content was calculated using the equation:

$$\text{Soluble sugar content} \left( \% \right) = \left( \mathbb{C} \times \mathbb{T}. 5 \right) \left( \mathbb{N} \times 104 \right) \tag{7}$$

where C is soluble sugar concentration from the standard curve (μg), and W is the fresh weight of leaves (g).

#### *2.8. Preparation of Crude Enzyme Extract*

For crude enzyme extraction, about 0.3 g of alfalfa leaves, which were dried with liquid nitrogen, were ground into powder in liquid nitrogen using prechilled mortar and pestle (4 ◦C). Then, 5 mL of 150 mM sodium phosphate buffer (PBS), pH 7.4, was added to the powder and the homogenate was centrifuged at 12,000 rpm for 20 min at 4 ◦C. The supernatant was collected and used as a crude extract to determine malondialdehyde (MDA) and soluble protein content.

#### *2.9. Determination of MDA Content*

The MDA content was determined by the thiobarbituric acid (TBA) method according to a previous report [21]. Briefly, previously prepared 1 mL crude enzyme extract was mixed with 2 mL of reaction mixture containing 20% (v/v) trichloroacetic acid and 0.5% (v/v) thiobarbituric acid. The mixture was then heated for 30 min in a 95 ◦C water bath and directly cooled to room temperature. The mixture was then centrifuged at 12,000 *g* for 10 min at 20 ◦C, and the absorbance of the supernatant was read at 450, 532, and 600 nm with a spectrophotometer (UV2600, UNIC, Shanghai, China). The MDA content was calculated using the following formula:

$$\text{MDA (mmol\cdot g}^{-1}\text{FW)} = [6.425 \times (\text{OD532} - \text{OD600}) - 0.559 \times \text{OD450}] \times \text{Vt} (\text{Vs} \times \text{FW}) \tag{8}$$

where Vt is the volume of extraction liquid (mL), Vs is the volume of extraction solution (mL), and FW is the fresh weight of samples (g).

#### *2.10. Soluble Protein Content Determination*

Soluble protein content was determined following the Bradford assay method [22]. Briefly, 100 μL crude enzyme extract was added to a tube containing 3 mL Bradford working solution. After 10 min, the absorbance was measured at 595 nm using a spectrophotometer (UV2600, UNIC, Shanghai, China). The protein concentration of the crude extract was determined from the standard curve established by the reference solution of bovine serum albumin (BSA) and the OD595 value of the sample.

#### *2.11. A Comprehensive Evaluation of Alfalfa for Heat Tolerance*

The comprehensive evaluation was performed according to the previous report [23] based on membership function value, which were calculated from the heat tolerance coefficients. The correlation of each biomass and physiological trait for all alfalfa cultivars was analyzed using heat tolerance coefficients. The heat tolerance coefficient of all biomass and physiological traits was calculated using the following equation:

$$\text{Heat tolerance coefficient} = \text{(HT/CK)} \times 100\tag{9}$$

where CK is the mean value of a single trait under the control treatment and HT is the mean value of a single trait under heat treatment.

The membership function values of all biomass and physiological traits were calculated from the heat tolerance coefficient using the formula (F1 and F2). Formula F1 was used for the traits that are directly related to heat tolerance, and F2 was used for the traits that are inversely related traits.

$$\text{F1 (\text{\textdegree\textquotesingle}i)} = (\text{\textquotesingle} \text{i } - \text{\textquotesingle} \text{m} \text{min}) / (\text{\textquotesingle} \text{m} \text{ax} - \text{\textquotesingle} \text{m} \text{min}) \tag{10}$$

$$\text{F2 (\text{\textquotedbl{}}i)} = 1 - (\text{\textquotedbl{}i-} \text{\textquotedbl{}min}) / (\text{\textquotedbl{}max} - \text{\textquotedbl{}min}) \tag{11}$$

where X is the *i* heat tolerance coefficient, max is the maximum heat tolerance coefficient value from all cultivars of the *i* trait, and Xmin is the minimum heat tolerance coefficient from all cultivars of the *i* trait.

#### *2.12. Chlorophyll a (Chl a) Fluorescence Analysis*

Based on the comprehensive heat tolerance evaluation, four alfalfa cultivars—two relatively heat tolerant (Bara310SC and Magna995) and two relatively heat sensitive (Gibraltar and WL712)—were selected and further evaluated by Chl *a* fluorescence analysis to select the most heat-tolerant and -sensitive alfalfa. Chl a fluorescence transient was measured using a pulse–amplitude modulation (PAM) fluorometer (PAM2500 Heinz Walz GmbH, Eichenring, Germany) according to the previous report [21]. Briefly, the leaves were kept in the dark for 30 min, and all measurements were taken using a saturating light intensity of 2000 μmol photons m−<sup>2</sup> s−1. Five measurements were taken for each cultivar and treatment. The strong light pulses inducted Chl a fluorescence emission, which was subsequently measured and digitized between 10 μs and 300 ms. The chlorophyll fluorescence induction curve (OJIP) transients were analyzed using the JIP-test.

#### *2.13. Statistical Analysis*

Data were subjected to analysis of variance (ANOVA) using SPSS software version 22.0 (IBM Corporation, Chicago, IL, USA). All values were shown as mean ± SD (Standard Deviation) (*n* = 5). The independent sample *t*-test was employed to compare the control and the treatment groups using the least significant difference (LSD) test. All statistical results were considered significant at *p* ≤ 0.05. All figures were created by Origin 9.0 (Origin Lab, Inc., Hampton, MA, USA).

#### **3. Results**

#### *3.1. E*ff*ect of Heat Stress on Alfalfa Plant Biomass*

Heat stress affected the biomass of all alfalfa cultivars (Table 2). Compared with the control, heat stress significantly reduced the shoot fresh weight of Golden Queen (45.74%), 55V48 (34.07%), WL354HQ (33.47%), Gibraltar (31.55%), and WL712 (28.57%). In addition, significantly higher shoot dry weight reduction was noticed in WL354HQ (55.10%), Golden Queen (45.45%), WL712 (40.43%), and Gibraltar (35.14%) (*p* < 0.05).


**Table 2.** Effect of heat stress on alfalfa plant biomass (g).

The data in the table are mean (*n* = 5). Asterisk indicates statistical significance difference between the control (CK) and heat treatment (HT) at *p* < 0.05, independent sample *t*-test. SFW, shoot fresh weight; SDW, shoot dry weight; RFW, root fresh weight; RDW, root dry weight; TFB, total fresh biomass; TDB, total dry biomass.

Furthermore, heat stress remarkably affected the fresh and dry root weights of most cultivars, with significant reductions recorded in SK3010, Sanditi, and Gibraltar (Table 2). Relative to the control, heat stress significantly decreased the total fresh biomass of Golden Queen, WL354HQ, 55V48, Gibraltar, and WL712 cultivars by 41.43%, 30.06%, 29.75%, 29.45%, and 24.04%, respectively (*p* < 0.05). Similarly, total dry biomass was significantly reduced by 48.48%, 43.55%, 41.54%, 39.74%, and 39.62% in WL354HQ, Golden Queen, WL712, Sanditi, and Gibraltar cultivars, respectively (Table 2). The biomass of Gibraltar, Golden Queen, and WL712 were highly affected by heat stress, which may indicate the sensitivity of the cultivars. By contrast, Bara310SC, Magna995, and WL363HQ cultivars were able to maintain their biomass under the heat stress condition, which may indicate their tolerance.

#### *3.2. Heat Stress Reduced the Relative Water Content (RWC) of Alfalfa*

Heat stress obviously decreased the leaf relative water content of all alfalfa cultivars compared to the control (Figure 1A). Of which, significantly higher decrement was noted in WL354HQ, WL712, Sanditi, WL440HQ, 55V48, and Siriver cultivars by 15.29%, 13.28%, 12.54%, 12.42%, 11.43%, and 10.01%, respectively (*p* < 0.05) (Figure 1A). Despite the reduction of RWC in Gibraltar, Golden Queen, and SK3010 cultivars being small (<10%), it was still significant compared to the control (Figure 1A). The result revealed that heat stress had a higher impact on the RWC of some alfalfa cultivars, especially on WL354HQ and WL712 cultivars, which showed higher water loss under heat stress. Bara310SC and Magna995 cultivars sustained their relative water content under heat stress and could be heat tolerant, but cultivar WL712 could be heat sensitive, as evidenced by higher water loss.

#### *3.3. Heat Stress Increased the Electrolyte Leakage (EL) of Alfalfa*

Heat stress affected the membrane integrity and stability and increased the EL of all alfalfa cultivars, as shown in Figure 1B, and was significant for the majority of the cultivars. Heat stress significantly increased the EL of Golden Queen, 55V48, Siriver, WL712, SK3010, WL354HQ, Nofollow, Gibraltar, WL525HQ, Sanditi, WL440HQ, and WL656HQ by 61.36%, 56.72%, 53.45%, 53.20%, 52.86%, 52.08%, 49.56%, 45.78%, 42.12%, 36.84%, and 36.45%, respectively (*p* < 0.05) (Figure 1B). However, Bara310SC cultivar had a lower increase in EL of 24.07%. The result showed that heat stress had a huge impact on the membrane stability of the majority of alfalfa cultivars, as shown by significantly higher increment in EL.

**Figure 1.** Effects of heat stress on physiological parameters. (**A**) Relative water content. (**B**) Electrolyte leakage. (**C**) MDA content. Each bar represents the mean (*n* = 5) and the error bar indicates the standard deviation. Asterisks indicate statistically significant differences between control and treatment group for each cultivar (*p* < 0.05), independent sample *t*-test. EL is electrolyte leakage, MDA is malondialdehyde.

#### *3.4. E*ff*ect of Heat Stress on Lipid Peroxidation*

As indicated in EL, heat stress damaged the membrane of alfalfa and caused lipid peroxidation, which was manifested by higher MDA content in all cultivars (Figure 1C). However, heat stress had no significant effect on some alfalfa cultivars such as Bara310SC, WL363HQ, WL440HQ, Magna995, WL656HQ, and Nofollow (Figure 1C). Meanwhile, under heat stress, Gibraltar, WL712, Siriver, Sanditi, WL525HQ, and Golden Queen cultivars showed significantly higher MDA content compared to the control (*p* < 0.05). The results showed that different alfalfa cultivars had a different level of sensitivity to heat stress. The higher the MDA content, the higher the lipid peroxidation and the greater the membrane damage.

#### *3.5. Heat Stress Decreased the Chlorophyll Content of Alfalfa*

It is obvious that heat stress affects the chlorophyll content of plant leaves, and the same was true for all alfalfa cultivars (Figure 2). The Chl content of some cultivars was more significantly and highly affected by heat stress than others compared to the control. In particular, heat stress significantly reduced the chlorophyll content of Gibraltar, Golden Queen, SK310, WL354HQ, WL363HQ, Sanditi, WL440HQ, WL525HQ, Siriver, and WL712 compared to control. Meanwhile, higher reduction in Chl a was observed in Gibraltar, WL354HQ, Golden Queen, Siriver, WL712, and Sanditi cultivars by 40.30%, 36.06%, 35.06%, 33.41%, 31.28%, and 31.11%, respectively (Figure 2A). Similar reduction in Chl b content was observed in WL712 (44.14%), WL354HQ (34.44%), WL440HQ, (31.53%), Golden Queen (24.69%), and Gibraltar (22.59%) (Figure 2B). In addition, higher reduction in total Chl content was noted in WL712, Gibraltar, 13, and Golden Queen (36.57%, 33.11%, 31.29%, and 31.05%, respectively)

(Figure 2C). The results suggested that some cultivars might be more sensitive to heat stress, as indicated by higher Chl content reduction under heat stress.

**Figure 2.** Effect of heat stress on the chlorophyll content of alfalfa cultivars. (**A**) Chlorophyll a content. (**B**) Chlorophyll b content. (**C**) Total chlorophyll content. Each bar represents the mean (*n* = 5) and the error bar indicates the standard deviation. Asterisks indicate statistical significant differences between control and treatment group for each cultivar (*p* < 0.05), independent sample *t*-test.

#### *3.6. E*ff*ect of Heat Stress on Soluble Sugar Content*

The soluble sugar content of all alfalfa cultivars showed an increment under heat stress relative to the control (Figure 3A). Heat stress significantly increased the soluble sugar content of Bara310SC, Magna995, WL363HQ, Nofollow, WL525HQ, and WL354HQ by 20.20%, 15.57%, 11.53%, 7.13%, 4.68%, and 2.22%, respectively (Figure 3A). The results showed that these cultivars produce higher soluble sugar to maintain their osmotic potential and organize proteins and cellular structures under heat stress, which increases heat tolerance.

#### *3.7. E*ff*ect of Heat Stress on Soluble Protein Content*

Similar to soluble sugar content, heat stress increased the soluble protein content of all alfalfa cultivars as shown in Figure 3B. Compared to the control, significantly higher soluble protein content was noted in Bara310SC, SK3010, WL363HQ, 55V48, WL440HQ, Magna995, and WL354HQ cultivars under heat stress (*p* < 0.05). Heat stress increased the soluble protein content of these cultivars by 36.18%, 26.73%, 25.52%, 24.91%, 24.44%, 21.01%, 18.91%, and 18.63%, respectively (Figure 3B). These results indicate that soluble protein plays an important role in alfalfa heat stress response.

**Figure 3.** Effect of heat stress on soluble sugar and soluble protein content. (**A**) Soluble sugar content. (**B**) Soluble protein content. Each bar represents the mean (*n* = 5) and the error bar indicates the standard deviation. Asterisks indicate statistical significant differences between control and treatment plant (*p* < 0.05), independent sample *t*-test.

#### *3.8. A Comprehensive Evaluation of the Heat Tolerance of Alfalfa Cultivars*

All biomass and physiological traits were standardized for the comprehensive heat tolerance evaluation using the heat tolerance coefficient. The heat tolerance coefficient of each biomass and physiological traits (indexes) are presented in Table 3. The heat tolerance coefficients were further used to calculate the membership function values of alfalfa cultivars.


**Table 3.** Heat tolerance coefficients of biomass and physiological traits (indexes) of alfalfa cultivars.

SFW, shoot fresh weight; SDW, shoot dry weight; RFW, root fresh weight; RDW, root dry weight; TFB, total fresh biomass; TDB, total dry biomass; RWC, relative water content; EL, electrolyte leakage; MDA, malnodialdehyde content; SS, soluble sugar content; SP, soluble protein content.

Furthermore, correlation analysis was performed to investigate the relationship between traits. The results revealed that shoot fresh weight and shoot dry weight were strongly positively correlated with total fresh biomass and total dry biomass, respectively (*r* = 0.94) (*p* < 0.01) (Table 4). Root fresh weight was strongly positively correlated with Chl b and SS, and root dry weight was strongly positively correlated with SS, RWC, and total dry biomass (Table 4). Total fresh biomass was significantly negatively correlated with EL. Total fresh biomass was significantly positively correlated with total dry biomass, and total dry biomass was strongly positively correlated with Chl b (*r* = 0.80) and total Chl content (*r* = 0.80) (Table 4). These results indicated that electrolyte leakage and MDA content were negatively correlated with the rest of traits, which were significant for Chl b, total Chl, and SS for electrolyte leakage, and Chlb and total Chl for MDA. However, electrolyte leakage was positively correlated with MDA content (*r* = 0.40). Chl a (*r* = 0.79) and Chl b (*r* = 0.94) were strongly positively correlated with total Chl content. In addition, there was a positive correlation between soluble sugar and soluble protein content (Table 4). Hence, all biomass and physiological traits were used to evaluate the heat tolerance of all alfalfa cultivars.

The membership function value calculated from heat tolerance coefficient was used to evaluate the heat tolerance of all alfalfa cultivars. As shown in Table 5, Bara310SC and Magna995 cultivars had higher mean membership function value (0.86 and 0.80), respectively. By contrast, Gibraltar (0.26) and WL712 (0.26) showed lower mean membership function value. Based on this result, Bara310SC and Magna995 were ranked "one" and "two", respectively whereas Gibraltar and WL712 had the same rank, "14" (Table 5). Furthermore, the mean membership function value was used for Euclidean distance cluster analysis. The results showed that Bara310SC and Magna995 cultivars were clustered into one group, and Gibraltar and WL712 were clustered in another group (data are not shown). Finally, the rank and cluster results were combined to evaluate the heat tolerance of the cultivars. Thus, Bara310SC and Magna995 were found to be heat tolerant, whereas Gibraltar and WL712 were found to be heat-sensitive alfalfa cultivars. To screen the most heat-tolerant and heat-sensitive alfalfa, the four cultivars (Bara310SC, Magna995, Gibraltar, and WL712) were further evaluated by chlorophyll a fluorescence analysis.

#### *3.9. Alteration of Chlorophyll a Fluorescence under Heat Stress*

Chlorophyll a fluorescence was measured to further screen the most heat-tolerant and heat-sensitive alfalfa cultivars based on the photosynthesis behavior under heat stress. OJIP curve was constructed from fluorescence transient measurement. The effect of heat stress on basic photosynthetic parameters, specific energy fluxes, quantum yield and efficiency, and performance indexes were investigated.



**Table** 

SP, soluble protein content.


**Table 5.** Membership function value of alfalfa cultivars.

electrolyte leakage; MDA,

malnodialdehyde

 content; SS, soluble sugar content; SP, soluble protein content.

#### 3.9.1. OJIP Transient Curve

OJIP transient curve was constructed based on the fluorescence measurement relative to the time (Figure 4). OJIP transient curves of control groups were higher than those of heat treatment groups for all cultivars. The JIP-test was applied to further investigate the structural alteration, functional parameters, and photosynthetic behaviors under heat stress treatment. Basic fluorescence parameters, specific energy fluxes, quantum yield efficiency, and performance index were extracted and analyzed.

**Figure 4.** Effect of heat stress on the chlorophyll *a* fluorescence transient (OJIP curves) of 4 alfalfa cultivars after seven days of heat treatment. CK represents control treatment and HT represents heat treatment (*n* = 5).

#### 3.9.2. Basic Photosynthetic Parameters (F0, Fj, Fi, Fm, F300 μs, and Fv/Fm)

The basic fluorescence parameters were extracted from the OJIP transient curve (Table 6). Heat stress increased the F0 of Gibraltar, Magna99,5 and WL712 alfalfa cultivars, and was significant for Gibraltar and WL712 cultivars when compared with the control (Table 6). Heat stress also increased the Fm (Maximal fluorescence) of alfalfa. Relative to the control, heat stress decreased the Fv/Fm of Gibraltar, Bara310SC, Magna995, and WL712 by 6.10%, 1.25%, 2.50%, and 6.25%, respectively, and was significant in WL712 cultivar (Table 6). The results indicated that WL712 was highly affected by heat stress, which revealed high heat sensitivity.

#### 3.9.3. Specific Energy Fluxes (TP0/RC, ETO/RC, RE0/RC, and ABS/RC)

Heat stress affected the specific energy fluxes of alfalfa. All heat-treated plants showed higher TPO/RC and were significant for WL712 (Table 6). Cultivars showed different responses in ETO/RC under heat stress. Heat stress significantly decreased the ETO/RC of WL712 but increased in Bara310SC and Magna995 (relatively heat tolerant) (Table 6). Unlike other cultivars, heat stress reduced the RE0/RC of Bara310SC (Table 6). On the other hand, heat-treated Gibraltar, Magna995, and WL712 plants showed higher ABS/RC compared to the control and was significant for WL712 (Table 6).


**Table 6.** Photosynthetic parameters extracted from OJIP fluorescence transients.

F0, minimal fluorescence; Fj, fluorescence intensity at the J-step (2 ms) of OJIP; Fi, fluorescence intensity at the I-step (30 ms) of OJIP; Fm, maximal fluorescence; F300 μs, fluorescence intensity at 300 μs; Fv/Fm, maximum quantum yield of photosystem; ABS/RC, absorption flux (of antenna Chls) per RC; TR0/RC, trapping flux (leading to QA reduction) per RC; ETO/RC, electron transport flux (further than QA−) per RC; REO/RC, electron flux reducing end electron acceptors at the PS I acceptor side, per RC; ϕpo, maximum quantum yield for primary photochemistry, namely FV/FM; ϕEo, quantum yield of the electron transport flux from QA to QB; δRo, efficiency/probability with which an electron from the intersystem electron carriers moves to reduce end electron acceptors at the PSI acceptor side (RE); RC/ABS, QA-reducing RCs per PSII antenna Chl (reciprocal of ABS/RC). The data in the table are mean (*n* = 5), and asterisks indicate statistical significance difference between control (CK) and heat treatment (HT) at *p* < 0.05, independent sample *t*-test.

#### 3.9.4. Quantum Yield and Efficiency (ϕpo, ϕEo, δRo, and RC/ABS)

Heat stress affected all quantum yield efficiency components (Table 6). Heat stress noticeably decreased the ϕpo and δRo of all alfalfa cultivars except Bara310SC. Relative to control, WL712 showed significant reductions in ϕpo, ϕEo, and RC/ABS under heat stress. Meanwhile, heat stress significantly decreased the ϕEo of Magna995 (Table 6).

#### 3.9.5. Performance Indexes (PIABS and PItotal)

Heat stress markedly decreased the photosynthetic performance indexes (PIABS and PItotal) of all alfalfa cultivars (Table 6). Meanwhile, heat stress significantly reduced the PIABS of Gibralter, Magna995, and WL712 by 44.44%, 35.64%, and 66.36%, respectively, compared to control (*p* < 0.05). In addition, WL712 showed significantly lower PItotal under heat stress, indicating its more heat sensitivity than others. However, heat stress had no significant effect on the performance indexes of Bara310SC (Table 6).

#### **4. Discussion**

High-temperature stress changes morphological, biochemical, and physiological processes of plants [4,24]. In this study, heat stress had an obvious negative effect on alfalfa plant biomass and most of the cultivars showed significant reductions in biomass following heat stress treatment. It has been reported that heat stress causes leaf wilting, leaf curling, leaf yellowing, reduction in shoot growth, root growth, root number, root diameter, plant height, and biomass [1,25]. Thus, the decrease in alfalfa plant biomass could be associated with the reduction in plant height, wilting, and falling off of leaves

caused by heat stress. Our results were consistent with previous reports in maize [26], sugarcane [27], and wheat [28]. The biomass of most alfalfa cultivars was significantly affected, indicating that heat stress has a huge impact on alfalfa productivity. Like biomass, heat stress caused a significant reduction in the RWC of most alfalfa cultivars. Our results are in agreement with previous reports in rice [29] and wheat [28]. Similar with the findings of Sita et al. [30] heat-tolerant alfalfa cultivars had higher RWC than heat-sensitive ones. The decrease in leaf water content might affect plant metabolism and decrease plant growth and biomass. The reduction in leaf relative water content could be associated with the reduction in the number, mass, and growth of the roots under heat stress, which ultimately limits the supply of water and nutrients to the above-ground parts of the plant [3]. Taken together, significant reductions in biomass and RWC could be indicators of alfalfa heat stress sensitivity.

It is well documented that the plant membrane is sensitive to various abiotic stresses, and stress condition increased lipid peroxidation and impaired membrane selectivity [31]. In this study, both EL and MDA content, which are indicators of stress sensitivity, were higher in heat-treated alfalfa plants compared to the control. The membrane stability of most alfalfa cultivars was significantly affected by heat stress, which may reveal the heat sensitivity of the cultivars. Our results were in agreement with the findings of Kumar et al. [32] in chickpea, Sita et al. [30] in lentil, and Hu et al. [21] in tall fescue, who reported higher EL under heat stress. The increase in lipid peroxidation might be as a result of the overproduction and accumulation of ROS, which then causes membrane peroxidation, protein degradation, and DNA damage to severely inhibit growth [33–35]. Consequently, our results revealed that heat stress highly damaged the membrane integrity and stability of alfalfa, especially heat-sensitive cultivars. However, some cultivars like Bara310SC and Magna995 had lower EL and MDA content than others, indicating that these cultivars could maintain their membrane integrity and stability under heat stress. It has been reported that the maintenance of membrane integrity and stability under stress conditions is a major component of tolerance [36] and is essential to sustained photosynthetic and respiratory performance [37]. Thus, Bara310SC and Magna995 with lower EL and MDA content after heat stress could be heat tolerant.

It is well established that photosynthetic pigments such as chlorophyll a and b are sensitive to high-temperature stress. Heat stress results in plant leaf pigment loss and significantly damages photosynthetic activities [38]. In the present study, heat treatment decreased the chlorophyll content (Chl a, Chl b, and total Chl) of all alfalfa cultivars, and a more pronounced effect was observed in heat-sensitive cultivars. The decrease in chlorophyll content might be attributed to the chlorophyll degradation or inhibition of chlorophyll biosynthesis [39]. In addition, the effect of high temperature on the pigments and other photosynthetic apparatus is due to the production of toxic oxygen species (oxidative damage) and reduction in antioxidative defense [32]. Thus, significant increases in EL and MDA content and a decrease in chlorophyll content might be interconnected, indicating the heat sensitivity of the cultivars. Our result revealed that cultivars with higher EL and MDA content had lower chlorophyll content, biomass, and RWC, suggesting that heat stress had a greater effect on some cultivars including WL712, Gibraltar, and Golden Queen.

When exposed to heat stress, plants accumulate compatible solutes, such as soluble sugar, to protect the plant from stress-induced damage by maintaining membrane stability and cell water balance, and by buffering the cellular redox potential and homeostasis [40]. The accumulation of compatible solute is an important adaptive mechanism, directly participating in osmotic adjustment [41]. Similarly, soluble proteins, which are induced by stress, play a role in stress tolerance, presumably via hydration of cellular structures [27]. In the current study, heat stress increased the soluble sugar and protein content of alfalfa plants. We found significant and higher soluble sugar content in Bara310SC, Magna995, WL363HQ. These cultivars could adjust their osmotic balance and cellular homeostasis, which is one of the tolerance mechanisms. Similar results were reported in lettuce [42] and moth bean [43]. In this study, we found higher relative water content in heat-tolerant alfalfa, which was in agreement with soluble sugar and protein content, thus entailing great implications for heat tolerance [27]. The result revealed that soluble sugar and protein could play a considerable role for alfalfa heat tolerance by maintaining

the water balance and cellular homeostasis. In this study, cultivars with higher soluble sugar and protein content had lower lipid peroxidation and membrane damage under heat stress, which implies that soluble sugar and protein ameliorates heat-induced damage in those cultivars. Our results are also supported by those of Lang-Mladek et al. [44], who stated that osmolyte production under heat stress is thought to increase protein stability and stabilize the structure of the membrane bilayer. Similarly, Khan et al. [45] and Kumar et al. [46] found significantly higher soluble protein content in wheat under heat stress, and maximum accumulation of soluble protein content was observed in thermotolerant wheat genotypes, which was similar with our result. The overall results indicate that soluble sugar and protein play a remarkable role in alleviating heat-induced damage of alfalfa, while increasing heat tolerance.

The heat tolerance of alfalfa was evaluated comprehensively by membership function value, and the results showed that alfalfa cultivars had different sensitivities to heat stress. Cultivars with higher membership function value were considered as heat tolerant, whereas cultivars with lower membership function value were heat sensitive. As shown, Bara310SC and Magna995 had higher membership function value, 0.86 and 0.80, respectively, and were considered as relatively heat tolerant. By contrast, with lower membership function value (0.24), Gibraltar and WL712 were relatively heat-sensitive alfalfa. Furthermore, chlorophyll a fluorescence analysis was performed on the above four cultivars to identify the most heat sensitive and heat tolerant alfalfa cultivars. Chlorophyll a fluorescence analysis is a powerful and nondestructive method to study the photosynthetic behavior of plants and is widely applied to screen tolerant species [21,47]. In this study, heat stress affected the chlorophyll a fluorescence and altered the OJIP fluorescence transient curve of all alfalfa cultivars. The change in the OJIP fluorescence transient curve might have been caused by the oxidation of electron transport chains, which results from the reduction in the electron donor of PSII reaction centers under high temperature. Basic photosynthetic parameters were extracted from fluorescence transient and the results showed that heat stress increased the F0 and decreased Fv/Fm of all alfalfa cultivars, which were significant for sensitive cultivars (WL712). Higher F0 indicates the elevated damage of chloroplast by heat stress, resulting in blocked energy transfer to the PS II traps and a decrease of the quantum efficiency of PS II [48]. Fv/Fm is commonly used to analyze heat-induced damage to PSII [47], and heat stress decreases Fv/Fm in a range of plant species [49]. For many plant species, the approximate optimal Fv/Fm value is in the range of 0.79 to 0.84, with lowered values indicating plant stress [50]. The heat-tolerant cultivar (Bara310SC) had 0.79 and 0.80 Fv/Fm under heat stress and control condition, respectively.

Heat stress altered the specific energy flux parameters (TP0/RC, ETO/RC, RE0/RC, and ABS/RC) of all alfalfa cultivars. ABS/RC and TR0/RC were higher under heat stress, which indicates the inactivation of absorption and trapping reaction centers. Similar results have been reported by Zushi et al. [51] in tomato leaf and fruit under heat stress. In addition, heat stress markedly altered the quantum yield and efficiency parameters (ϕpo, ϕEo, and δRo). The result revealed that the behaviors of PS II on both the electron donor and acceptor side were blocked due to heat stress and PSI was less damaged than PSII [51]. The energy fluxes such as ϕpo, ϕEo, and RC/ABS of PS II were lower, whereas δRo was higher under heat treatment. Similarly, the decrease in RC/ABS was observed in heat-stressed Spirulina [52]. On the other hand, Stefanov et al. [53] reported an increase of δRo in bean plants immediately after heat treatment. This result suggested a difference in energy flux between PS I and PS II in response to heat. It was reported that PS II is the most temperature-sensitive component of the photosynthetic apparatus [54,55]. Our results also confirmed that PS II is more sensitive to heat stress than PS I, which could be as a result of thylakoid membrane fluidity caused by heat stress.

Alteration of specific energy fluxes and quantum yield efficiency of the photosystem could affect the overall photosynthetic performance of alfalfa. Performance indexes (PIABS and PItotal) were measured to investigate the changes in leaf photosynthetic performance. The performance index (PIABS) is a parameter sensitive to various types of stress. PItotal reveals the changes in intersystem electrons and the energy conservation from exciton to the reduction of PSI end acceptors [56]. Heat stress noticeably decreased the performance indexes of all heat-treated alfalfa cultivars, and a significant decrease was observed in sensitive cultivar (WL712). Another heat sensitive cultivar (Gibraltar) showed a significant reduction in PIABS. A similar result was reported in tall fescue under heat stress [21]. In addition, Fahad et al. [29] reported a significant reduction in the photosynthetic activities of two rice cultivars under high day and night temperatures. The decrease in performance indexes could indicate the lower photochemistry of PSII [57]. However, heat stress had no significant effect on the performance indexes of heat-tolerant cultivars (Bara310SC and Magna995), and Bara310SC showed higher performance indexes under heat stress than others. The result revealed that Bara310SC was more tolerant to heat stress compared to others under heat stress. The overall Chl a fluorescence analysis results suggested that Bara310SC and WL712 are the most heat-tolerant and heat-sensitive alfalfa cultivars, respectively. Further studies will be done to understand the molecular mechanisms of the heat tolerance of alfalfa.

#### **5. Conclusions**

Heat stress affected the biomass and physiological characteristics of alfalfa cultivars. The more pronounced effect was observed in sensitive cultivars such as WL712, Gibraltar, and Golden Queen, as evidenced by a significant decrease in biomass, RWC, chlorophyll content, and photosynthetic performance, and significant increases in EL and MDA under heat stress. Heat-tolerant cultivars showed significantly higher soluble sugar and protein content and performed better under heat stress. Among the fifteen cultivars evaluated, Bara310SC was the most heat tolerant and WL712 was the most heat-sensitive one. These cultivars can be used to explore the molecular mechanisms of heat tolerance in alfalfa plants.

**Author Contributions:** Conceptualization, M.W. and L.C.; methodology, M.W., W.Z. and Q.Z.; software, M.W. and K.J.; validation, M.W.; W.Z. and L.C.; formal analysis, M.W.; investigation, M.W.; resources, L.C.; data curation, M.W., Q.Z. and W.Z.; writing—original draft preparation, M.W.; writing—review and editing, M.W. and L.C.; visualization, M.W.; supervision, L.C. and K.J.; project administration, L.C.; funding acquisition, L.C.

**Funding:** This work was funded by the National Natural Science Foundation of China (NSFC) (Grant Nos. 31672482 and 31401915).

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*

### **Transcriptomic Analysis Reveals the Temporal and Spatial Changes in Physiological Process and Gene Expression in Common Buckwheat (***Fagopyrum esculentum* **Moench) Grown under Drought Stress**

**Zehao Hou 1,**†**, Junliang Yin 1,**†**, Yifei Lu 1, Jinghan Song 1, Shuping Wang 1, Shudong Wei 2, Zhixiong Liu 3, Yingxin Zhang <sup>1</sup> and Zhengwu Fang 1,\***


Received: 17 August 2019; Accepted: 19 September 2019; Published: 20 September 2019

**Abstract:** Common buckwheat is a traditional alternative crop that originated from the northwest of China and is widely cultivated worldwide. However, common buckwheat is highly sensitive to drought stress, especially at the seedling stage, and the molecular mechanisms underlying the response to drought stress still remain elusive. In this study, we analyzed the stress phenotypes of buckwheat seedlings under drought condition. The results showed the wrinkled cotyledon due to the decrease of relative water content (RWC) in response to the increased activity of antioxidant enzymes. Transcriptomic analysis was further performed to analyze the regulation patterns of stress-responding genes in common buckwheat cotyledons and roots under drought stress conditions. Characterizations of the differentially expressed genes (DEGs) revealed differential regulation of genes involved in the photosynthesis and oxidoreductase activity in cotyledon, and that they were highly related to the post-transcriptional modification and metabolic process in root. There were 180 drought-inducible transcription factors identified in both cotyledons and roots of the common buckwheat. Our analysis not only identified the drought responsive DEGs and indicated their possible roles in stress adaption, but also primarily studied the molecular mechanisms regulating the drought stress response in common buckwheat.

**Keywords:** common buckwheat; cotyledon; root; drought stress; transcriptome analysis

#### **1. Introduction**

Among the forms of environmental stress, drought stress has been considered as one of the major constraints in plant growth, survival, and production [1,2]. A lack of water not only disturbs photosynthesis, limits metabolic reactions, and inhibits CO2 exchange, but also results in stress-related damage to chloroplasts [3–5]. In order to adapt to the extreme environments, plants have evolved several mechanisms (e.g., drought escape, avoidance, and tolerance) to ensure high survival rates under drought stress [5,6]. Specifically, plants recruit a variety of responding mechanisms to deal with drought stress [7–9], such as stomatal closure, leaf rolling, and alteration in biosynthetic and antioxidant pathways, which are highly regulated by complex transcriptional networks [10].

Drought stress affects several physiological and biochemical pathways in plants [11]. Previous research has shown that the water deficits not only affect the chlorophyll biosynthesis, but also the level of malondialdehyde (MDA) and the relative water contents (RWC) of the plant, brining detrimental effects to the lipid peroxidation, and membrane constitution [12,13]. Also, the abiotic stresses can further induce the oxidative stress through generating reactive oxygen species (ROS), a prevalently recognized destroyer in cellular metabolism [14–16]. ROS generate the oxidation of photosynthetic pigments, initiate lipid peroxidation, and degrade proteins in plants, and thereby cause damage to cell structures and metabolism, particularly those associated with photosynthesis [17,18]. To counteract the effects of oxidative stress, plants have developed an efficient detoxification defense system consisting of non-enzymatic scavengers and enzymatic components to scavenge free ROS [19,20]. In terms of the enzymatic scavenging, a series of antioxidative enzymes, including peroxidase (POD), catalase (CAT), superoxide dismutase (SOD), and ascorbate peroxidase (APX), have been reported to play a vital role in reducing the damage effects (i.e., water deficiency) caused by drought stress [21]. There is evidence that keeping a high antioxidative enzyme activity level to reduce the damaging effects caused by water deficit stress may be associated with the osmotic stress tolerance of plants [22], which is also found to be positively related to plant drought tolerance [23].

Presently, there are several drought-inducible genes, including stress responses and resistance, which have been identified through transcriptome analyses [1]. These genes can be divided into two groups according their functions. The first group is composed of function proteins that include the late embryogenesis abundant (LEA) proteins, ROS detoxification enzymes, molecular chaperones, heat shock proteins (HSP), and lipid-transfer proteins [24,25]. The second group is involved in regulatory proteins or transcriptional factors (TFs), which correlate with the signal transduction and stress-responsive gene expressions, for example, the phospholipases and dehydration-responsive elements [26,27]. In order to elucidate the biological functions of these genes, several transgenic plants overexpressing various drought-resistant genes have been generated, which have both shown enhanced drought tolerance and growth retardation [28–31], demonstrating that plants may adapt to the drought environment at the expense of normal growth [1].

Common buckwheat (*Fagopyrum esculentum* Moench) is an important dual-purpose alternativecrops originated from Yunnan Province of China [32] and is widely cultivated around the world, especially in China, Japan, and Russia [33]. Because of its abundant nutrients in seeds, common buckwheat is considered as one of the sources of flour, groats, and whole grain foods. However, common buckwheat is highly sensitive to drought, especially at the seedling stage [34,35], and short-term drought occurs frequently in China, posing a threat to domestic food safety [36]. Thus, it would be important and necessary to study the physiological and molecular bases of osmotic stress tolerance in common buckwheat. We have previously identified *FeDREB1L* (GenBank: JN600617.1), a CBF/DREB homologous gene, from common buckwheat, and overexpression of the *FeDREB1L* gene was found to significantly increase the water deficit resistance of transgenic *Arabidopsis* [31]. In order to further understand the drought-resistant mechanism and identify novel water-deficit-related genes in common buckwheat, a transcriptomics analysis was carried out to investigate the variations in common buckwheat growth under short-term drought treatment, and the phenotypes and biochemical traits of seedlings were also analyzed. Our results may provide more information with regard to the transcriptional control of common buckwheat under the abiotic stresses, and help to identify the novel genes that are potentially valuable for future common buckwheat breeding.

#### **2. Materials and Methods**

#### *2.1. Plant Material and Drought Treatments*

Common buckwheats (cv. Xi'nong 9976) were germinated in Petri dishes in an incubator (plant growth incubator JY412L, Shanghai, China) in darkness (25 ◦C) and relative humidity of approximately 60%. After germination for 36 h, when the root length of the seedlings grew to approximately 2 cm, the seedlings were transplanted for hydroponics in an incubator with 12 h photoperiods (25 ◦C/20 ◦C, day/night temperature) and relative humidity of approximately 60%. The 7-day old buckwheat seedlings were treated with 15% polyethylene glycol 6000 (PEG 6000) solution for 1 d, 3 d, and 5 d. After treatment, the cotyledons and roots were collected and quickly frozen in liquid nitrogen and stored at −80 ◦C until used. The seedlings before drought treatment were served as the control. Each treatment was carried out in three biological replicates.

#### *2.2. Physiological Measurement*

Relative water content (RWC) was determined according to the formula described by Pan et al. [36]. The chlorophyll content and chlorophyll a/b ratio was calculated using to the method described by Harper et al. [37]. The changes of malondialdehyde (MDA) concentration were determined using the thiobarbituric acid (TBA) reaction [38], and the activities of POD and CAT were detected according to the description of Harper et al. [37]. The Rubisco activities were assayed with Rubisco assay kits (Beijing Solarbio Science and Technology Co., Ltd., Beijing, China) according to the manufacturer's instructions.

#### *2.3. RNA Isolation and Transcriptome Sequencing*

Total RNA was isolated from the non-treated control and drought-stressed cotyledon and root samples using EASYspin Plus Plant RNA Kit (Aidlab, Wuhan, China). The RNA quality was checked by Agilent bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA), and Nanodrop 2000 r spectrophotometer (Nanodrop Technologies, Wilmington, DE, USA) was used for RNA quantification.

Transcriptome sequencing was performed at Beijing Allwegene Technology Co. Ltd. (Beijing, China), following manufacturer protocols. Briefly, mRNA was enriched from total RNA using Oligo (dT) magnetic beads, and the mRNA was fragmented into small pieces using a fragmentation buffer. Then, these fragments were used as reverse transcription to synthesized the first- and second-strand cDNA, and the second-stand cDNA were purified with AMPure XP Beads Kit, repaired, poly (A) added, and ligated to paired-end adapters. Finally, the cDNA libraries were sequenced on Illumina HiSeqTM 2500 platform. Each sample had three biological replicates. The raw reads were submitted to the National Center for Biotechnology Information (NCBI) Sequence Read Archive with a Bioproject ID: PRJNA555746.

The raw reads in FASTQ format were processed using in-house Perl scripts, and the high-quality clean data were obtained by removing the low-quality data, which included the reads that contained the adapter, and more than 10% of N nucleotides, and the low-quality reads that contained more than 50% of low quality bases (Q-value ≤ 20). In addition, we calculated the Q30, GC content, and sequence duplication levels for the clean data. Cleaned and qualified reads were aligned against the *F. esculentum* reference genome [39] using Tophat2 software [40]. Then, these sequences were subjected to functional annotation and coding sequence (CDS) prediction [41], and the resulting sequences were called genes. Finally, fragments per kilobase of transcript permillionmapped reads (FPKM) method was used to calculate the gene expression unit.

#### *2.4. Identification and Functional Annotation of Di*ff*erent Expressed Genes (DEGs)*

The differential gene expression analysis was carried out using DESeq software, and DEGs were determined by combining a *q* value cutoff of 0.05 and adjusting to |log2 (fold change)| ≥ 1. For DEG functional annotation, Gene Ontology (GO) enrichment analysis was carried out by GOseq software, and Kyoto Encyclopedia of Genes and Genomes (KEGG) was used to perform pathway enrichment analysis of DEGs. In addition, the gene expression profiles at the pathway were display by MapMan software (version 3.6.0) [33].

#### *2.5. Quantitative Real-Time PCR (qRT-PCR) Analysis*

Total RNA from cotyledons and roots of both samples were extracted using EASYspin Plus Plant RNA Kit (Aidlab, Wuhan, China) following the manufacturer's protocols, and the first-strand cDNA for qPCR analysis was synthesized from 500 ng of total RNA using PrimeScript RT Reagent Kit with gDNA Eraser (Takara, Dalian, China) following the manufacturer's instructions, and cDNA was diluted 10-fold and used as the template for qRT-PCR. The primers were designed using Primer Premier 5.0 and beta-actin was used as a reference gene, with the primer information listed in Table S1. qRT-PCR was performed on a CFX 96 real-time PCR system (BioRad, Hercules, CA, USA) using TB Green (TaKaRa), according to the manufacturer's protocols. PCR amplification was conducted in a volume of 20 μL, containing ~100 ng of cDNA template, 0.6 μL of each primer (10 μmol), and 10 μL PCR-mix (2×). The conditions for all reactions were as follows: 30 s at 95 ◦C, followed by 40 cycles of 10 s at 95 ◦C, and 30 s at 55 ◦C, and the melting curve was generated to confirm the PCR specificity. The non-treated control treatment was chosen as the control to standardizing all samples, using the 2−ΔΔCt method to calculate the relative expression levels [42].

#### *2.6. Statistical Analysis*

A one-way ANOVA was carried out by SPSS Statistics 19.0 software (IBM Corp, Armonk, NY, USA), and means were compared using the Duncan test to determine significant differences (*p* < 0.05). The results were presented as mean ± SD (standard deviation).

#### **3. Results**

#### *3.1. Changes in Phenotype of Common Buckwheat Seedlings at Drought Stress*

To investigate the dynamic phenotypic changes of common buckwheat seedlings in response to drought stress treatments, plant height, root length, and relative water content (RWC) were measured under 15% PEG 6000 solution treatments across four time-points (0, 1, 3, and 5 days). There was no significant change in plant height, root length, and RWC of control samples after 1, 3, and 5 days (Table S2). Under drought stress, the buckwheat seedlings showed stress phenotypes of wrinkled cotyledon (Figure 1a), but the plant height and root length did not significantly change during the treatment (Figure 1b,c). The RCW is generally used as an important indicator of plant water status under osmotic conditions, and in this study, the RCW values were clearly decreased in the 3 and 5 day-treated (DPT3d and DPT5d) seedlings, but were not significantly different among the control (CK) and the 1 day-treated (DPT1d) seedlings (Figure 1d).

#### *3.2. Changes in Physiology of Common Buckwheat Seedlings under Drought Conditions*

To investigate the physiological changes under different levels of water deficit conditions, the MDA content and the activities of POD and CAT of cotyledons were measured after drought treatment. Compared with the drought treatment, these physiological traits were not significantly changed during the whole treatment under control conditions, in which the content of chlorophyll a and chlorophyll b were increased in the 1, 3, and 5 day control plants; however, the chlorophyll a/b ratios were not significantly changed during the whole treatment under the control condition (Table S3). Under water deficit condition, the MDA content was greatly increased from 0 to 1 days, and then slightly increased until 5 days (Figure 2a). Meanwhile, the activities of POD and CAT were significantly increased under PEG treatment (Figure 2b,c). The chlorophyll a content was elevated in the 3 and 5 dy-treated plants (Figure 2d), while the content of chlorophyll b content was increased in the 1, 3, and 5 day-treated plants (Figure 2e). In addition, there was no difference between the control and 1 and 3 day-treated seedlings in chlorophyll a/b ratios, but marked decreases were observed in the 5 day-treated plants (Figure 2f). These results indicated that there were significant changes in the physiology of the common buckwheat seedlings in response to osmotic stress.

**Figure 1.** Changes in phenotype of common buckwheat seedlings under drought stress. (**a**) A photograph of common buckwheat seedlings after PEG treatments; CK, non-stressed control; DPT1d, DPT3d, and DPT5d, drought treatment with PEG solution for 1, 3, and 5 days, respectively; bar = 20 cm. The change of (**b**) plant height, (**c**) root length, and (**d**) relative water content of cotyledon during drought stress treatment. Bars represent means of three replicates ± SD (standard deviation). Different letters indicate means that are significantly different at the *p* < 0.05 level among different drought conditions.

**Figure 2.** Changes in physiology in cotyledons of common buckwheat seedlings under different drought stress conditions. (**a**) Changes in the malondialdehyde (MDA) content of cotyledons, (**b**) changes in the peroxidase (POD) activities of cotyledons, (**c**) changes in the catalase (CAT) activities of cotyledons, (**d**) chlorophyll a content in the cotyledons, (**e**) chlorophyll b content in the cotyledons, and (**f**) ratios of chlorophyll a/b.

#### *3.3. Overview of the Common Buckwheat Transcriptome and Identification of DEGs*

To reveal the expression changes in common buckwheat cotyledons and roots at 0, 1, 3, and 5 days after PEG treatments, 24 common buckwheat samples (including 12 cotyledon samples and 12 root samples) were used for RNA-Seq analysis to further investigate the changes at the transcriptional level. A full-scale sequencing analysis from 24 cDNA samples is shown in Table S4, 88.76 gigabytes (Gb) of clean reads were abstained, and the percentages of the Q30 base of these 24 common buckwheat samples were greater than or equal to 95.70%. Furthermore, there was a highly mapped efficiency between the samples and reference genome (79.46%–87.91%), which met the requirements for information analysis. There were 877,111 CDS (coding sequence)-encoded proteins, and the length of CDS is shown in Figure 3a. Of these CDSs, only a minority (3729 CDSs, 0.43%) were more than 1500 nt, and 90.57% of CDSs appeared with a length ranging from 0 to 500 nt. In addition, the principal component analysis (PCA) was performed and the results demonstrated that the control treatment was clearly separated from the drought stress treatment in the cotyledons or roots (Figure 3b), suggesting that the gene expression pattern of common buckwheat was greatly changed under drought condition.

**Figure 3.** Overview of the transcriptomic results and changes in gene expression profiles in cotyledons and roots after drought stress treatment. (**a**) Predicted length distribution map of coding sequence (CDS)-encoded protein nucleotide (nt). (**b**) Principal component (PC) analysis of gene expression at different drought stress conditions. (**c**) Numbers of differently expressed genes (DEGs) in cotyledons of common buckwheat seedlings at different drought stress conditions in pairwise comparisons. (**d**) DEGs in roots of common buckwheat seedlings at different drought stress conditions in pairwise comparisons. (**e**) Heat-map graphics exhibiting the gene expression levels of total DEGs. CKL and CKR are the cotyledon samples and root samples of the non-stressed control, respectively. DPT1L, DPT3L, and DPT5L are the cotyledon samples of drought treatment for 1, 3, and 5 days, respectively. DPT1R, DPT3R, and DPT5R are the root samples of drought treatment for 1, 3, and 5 days, respectively. DPT1L vs. CKL, DPT3L vs. CKL, and DPT5L vs. CKL are the cotyledon samples of drought treatment for 1, 3, and 5 days compared to the non-stressed control, respectively. DPT1R vs. CKR, DPT3R vs. CKR, and DPT5R vs. CKR are the root samples of drought treatment for 1 day compared to the non-stressed control, respectively. Up-regulated means that genes were up-regulated in drought stress conditions compared to the non-stressed control and down-regulated means that genes were down-regulated in the drought stress condition compared to the non-stressed control.

The gene expression levels were calculated as FPKM values via HTSeq software analysis, and the differential gene expression analysis was carried out using DESeq software. There were 2436, 2060, and 6377 DEGs identified in the cotyledons after drought treated for 1, 3, and 5 days, respectively (Figure 3c). In the roots, compared with the control treatment, 7304, 11,774, and 10,447 DEGs were identified after drought treatment for 1, 3, and 5 days, respectively (Figure 3d). Upon drought stress exposure, more DEGs were identified in the roots than in cotyledons, suggesting that there were different drought stress response mechanisms between roots and leaves in common buckwheat. Furthermore, in order to provide a comprehensive understanding of the change in gene expression of common buckwheat under drought conditions, a heat map was developed, as shown in Figure 3e, to exhibit the overall changes of the gene expression under water deficit conditions.

#### *3.4. Comprehensive Sets of DEGs in the Cotyledons of Drought-Treated Common Buckwheat Seedlings*

Venn diagrams showed that the number of genes commonly up-regulated under the DPT1L and DPT5L were greater than the number of genes commonly up-regulated under the DPT3L and DPT5L, and the number of commonly down-regulated genes showed the same trend (Figure 4a,b). Gene Ontology (GO) was used to find the functional significance of the identified DEGs (Figure 4c), and the GO terms related to signaling and DNA modification were detected in the set of genes up-regulated under the drought conditions, while the GO terms related to light harvesting and light reaction were detected in the sets of genes down-regulated under the water deficit conditions. Furthermore, the DEGs related to the light reactions of photosynthesis and the Calvin cycle were visualized through MapMan analysis (Figure 4d), and the expression of most of these DEGs was decreased in the DPT3L and DPT5L, compared with control treatment. RubisCO, as the major photosynthetic enzyme in plants, plays a crucial role in photosynthesis of green plants. In this study, the activities of RubisCO were significantly decreased at 3 days and greatly declined at 5 days under drought treatment (Figure 4e). These results indicate that photosynthesis in the cotyledons of the common buckwheat seedlings decreased under drought stress conditions. The representative genes related to ABA (abscisic acid) metabolism are listed in Table S5 according to their functional description, and most of these (including 6 *NCED* (*9-cis-epoxycarotenoid dioxygenase*), 3 B3 domain-containing protein, and 3 protein phosphatase 2C) were significantly up-regulated in DPT5L, which indicated that common buckwheat seedlings may use the ABA regulatory systems to affect leaf wilting and defense against the water-deficit stress.

To confirm and investigate the transcriptomic data, qRT-PCR was performed to check the expression levels of several genes, and the expression of *LCHb*, a gene encode chlorophyll a/b binding protein, was markedly decreased under DPT3L and DPT5L conditions (Figure 4f). The expression of *NCED*, a gene involved in ABA biosynthesis, was observably increased under drought stress conditions (Figure 4i). Furthermore, the *DREB1L* gene, which encodes a stress tolerance-related protein, was markedly enhanced in its expression level under drought stress treatment, compared with the non-stressed control plants (Figure 4m). The correlation coefficient (*R*2) between RNA-Seq data and qPCR results for the 24 total plots was 0.8743 (Figure S1). These analyses of gene expression confirmed that the transcriptomic datasets were efficacious (Figure 4f–n).

#### *3.5. Comprehensive Sets of DEGs in the Roots of Drought Treated Common Buckwheat Seedlings*

Three comparison groups were constructed to further understanding the universal response in root of common buckwheat to drought stress. As shown in Figure 5a, 1731 genes were both up-regulated in DPT1R, DPT3R, and DPT5R, compared with CKR. Furthermore, 2857 genes were collectively down-regulated in DPT1R, DPT3R, and DPT5R (Figure 5b). All non-overlapped DEGs in the three comparison groups were subjected to GO enrichment analysis, and 430, 606, and 621 GO accessions classified into three categories comprising "molecular function", "biological process", and "cellular component" were identified in DPT1R vs. CKR, DPT3R vs. CKR, and DPT5R vs. CKR, respectively (Table S6). The drought-induced DEGs were mainly involved in the nucleotide binding, ATP binding, macromolecule modification, protein phosphorylation, protein modification, protein metabolic process, and cellular protein metabolic process (Figure 5c). According to Mapman software analysis, there were 92 protein modification and phosphorylation-related DEGs that were up-regulated after drought treatment (Table S7), suggesting that they were candidate genes for protein modification in roots of drought-treated seedlings.

**Figure 4.** Comprehensive expression patterns of DEGs in the cotyledons of common buckwheat seedlings under drought stress conditions. (**a**) Venn diagrams of the numbers of up-regulated [|log2 (Fold Change)| > 1 and q-value < 0.005] genes acquired through the transcriptome analysis. (**b**) Venn diagrams of the numbers of down-regulated genes acquired through the transcriptome analysis. (**c**) Over-represented Gene Ontology (GO) terms estimated using GOseq software. (**d**) Changes in the expression of photosynthesis-related genes. Pathway diagram of light and dark reactions of photosynthesis with superimposed color-coded squares showing DEGs, drawn using MapMan. (**e**) Changes in the RubisCO activities of cotyledons. (**f–n**) Expression profiles of the selected DEGs, *LCHb* (**f**), *PARP* (**g**), *SWEET* (**h**), *NCED* (**i**), *HOMEZ*(**j**), *KCS* (**k**), *PSAN* (**l**), *DREB1L* (**m**), and *XLOC\_246139* (**n**) determined using qRT-PCR analyses.

The expression patterns of several genes related to stress tolerance in the roots were analyzed via qRT-PCR (Figure 5d–k). The expression of *AAO*, a gene encoding L-ascorbate oxidase that plays a crucial role in plant cell growth, was markedly decreased under drought stress conditions (Figure 5h), and the expression of some transcription factors, such as *HOMEZ* and *DREB1L*, were significantly induced by water deficit (Figure 5e,j). In addition, the *R*<sup>2</sup> between the two experiments was 0.9385 (Figure S2). These results confirmed the effectiveness of the transcriptomic datasets and indicated that drought stress strongly affected the expression level of the genes that are related to stress tolerance in the roots.

**Figure 5.** Comprehensive expression patterns of DEGs in the roots of common buckwheat seedlings under drought stress conditions. (**a**) Venn diagrams of the numbers of up-regulated [|log2 (Fold Change)| > 1 and qvalue < 0.005] genes acquired through the transcriptome analysis. (**b**) Venn diagrams of the numbers of down-regulated genes acquired through the transcriptome analysis. (**c**) Classification of DEGs based on metabolism, binding and modification categories. (**d**–**k**) Expression profiles of the selected DEGs, *KCS* (**d**), *DREB1L* (**e**), *CDC* (**f**), *LCHb* (**g**), *AAO* (**h**), *SWEET* (**i**), *HOMEZ* (**j**), *Pcyt-like* (**k**), determined by qRT-PCR.

#### *3.6. Change in the Expression of Transcription Factors (TFs) Associated with Drought-Stress Response in Common Buckwheat Seedlings*

Transcription factors (TFs) are important for regulating plant response to abiotic and biotic stresses. In the roots of the drought treated seedlings, large numbers of TFs were identified as DEGs, compared to the cotyledons (Figure 6a), and there were 180 TFs that were commonly identified in response to drought stress in both cotyledons and roots. Among them, the most differentially expressed TF families were the C2C2 family, follow by MYB, bZIP, HB and AP2/ERF (Figure 6b, Table S8). According to the GO enrichment analysis, 30.0%, 12.2%, and 9.4% of TFs were classified into "biological regulation", "intracellular", and "nucleic acid binding", respectively (Table S8). In addition, in order to reflect the major trends and patterns, 180 TFs were assigned to six clusters on the basis of their expression patterns. Those in cluster 1 and 2 were up-regulated by drought conditions, but the expression levels of the cluster 1 genes were high at cotyledons, while the cluster 2 genes were highly expressed at roots. Meanwhile, the cluster 3 and 4 genes showed the lowest expression level at cotyledons and roots. The cluster 5 genes were down-regulated by the drought condition. In contrast, there were 25 TFs in cluster 6, and these genes were up-regulated by water deficit, and the genes' expression was most high at cotyledons and roots. These results indicate that the expression of TFs was greatly affected by water deficit in the cotyledons and roots, and had different patterns between cotyledon and root tissues.

**Figure 6.** The differentially expressed TFs in cotyledons and roots responsive to drought stress. (**a**) Venn diagrams of TFs between cotyledons and roots. (**b**) Classification of TFs that were commonly identified in both cotyledon and root transcriptome libraries. (**c**) Expression pattern of TFs that were commonly identified in both cotyledons and roots in response to drought stress.

#### **4. Discussion**

Drought stress is one of the most detrimental environmental factors disturbing crop growth and production, and therefore understanding the drought-tolerance mechanism is pivotal for crop breeding [4,43]. Currently, RNA sequencing has been widely used to identify the drought-responsive pathway and genes that are activated during the seedling stage when exposed to abiotic stresses [5]. In this study, the phenotypic and physiological alterations of common buckwheat seedlings during drought stress were analyzed and characterized with the transcriptome analysis. Our results indicate that the common buckwheat seedlings relied on complex biological process to tackle the drought stress.

#### *4.1. Morphological and Physiological Characteristics Related to Drought Stress in Common Buckwheat*

Under drought stress, plant seedlings exhibit certain physiological and morphological variations [2,8], shown by leaf rolling and wilting [44,45], as well as the decreased RCW and wrinkled leaves (Figure 1a,d). Previous studies have demonstrated that drought stress also inhibits the photosynthesis of plants by affecting chlorophyll biosynthesis and facilitating stomatal closure [46,47], leading to the accumulation of MDA and ROS, which is harmful to the chloroplast photosystem II (PSII) [17,48]. As a result, plants have evolved antioxidant enzyme systems, such as superoxide dismutase (SOD), guaiacol peroxidase (GPX), CAT, and POD, to counteract the damage caused by drought stress [49]. MDA content has been considered important to reflect the drought tolerance ability of plants [12]. An active antioxidant capability in scavenging the cytotoxic ROS is preferred by plants in drought stress [50]. In common buckwheat seedlings, the ratio of chlorophyll a/b and RubisCO activities were significantly decreased compared to the control treatment in 5 days of treatment (Figure 2f, Figure 4e), and the expression levels of DEGs involved in the photosynthesis were correspondingly decreased DPT3L and DPT5L (Figure 4d), which may be due to the decrease in photosynthesis of the common buckwheat seedlings. By contrast, the CAT and POD activities were enhanced (Figure 2b,c). These observations suggest that the drought stress induced evident perturbations in the photosynthesis and ROS scavenging enzyme activities.

#### *4.2. Multiple Biological Processes Are Involved in Drought Stress Responses in Common Buckwheat*

Previous studies reported that multiple biological processes, such as oxidoreductase activities, and carbohydrate and protein metabolic processes, could be influenced when the ROS accumulation increased [51,52]. Our results indicated that the oxidoreductase activity, kinase activity, and DNA modification were upregulated by the drought treatment in common buckwheat cotyledons, whereas the genes were classified into "light harvesting" and "light reaction" (Figure 4c), and the expression of *LCHb* (encode chlorophyll a/b binding protein) and *PASN* (encode photosystem I reaction center subunit N protein) were markedly downregulated under drought conditions (Figure 4f,l). Repression of photosynthesis under drought stress also occurred in order to help the plants survive the water deficiency [53]. Cellular water deficit in plants caused by drought stress results in weakened carbon fixation, which may be physiologically ascribed to the stomatal closure and the biochemical inhibition of photosynthetic activities, furtherimpacting the carbohydratemetabolism [54]. In addition, phosphorylation, as one of the reversible post-translational protein modification mechanisms, plays an important role in signaling the plant adaptation to osmotic stress [55,56], and a fine-tuned control of protein activity and function [56,57] has also been found to be altered in the common buckwheat root under drought stress. There were 92 DEGs involved in the protein modification and phosphorylation that were up-regulated (Figure 5c, Table S7), which may provide insights in the future study of the protein phosphorylation and modification events in plant drought stress.

#### *4.3. Genes and Functional Proteins Responsive to Drought Stress*

Previous studies have descried cellular changes that occur upon exposure to drought stress in plants, and the gene responses to drought stress have been studied spaciously in various species [1,9,11,53,58]. Moreover, the plant hormone ABA, as a signal-sensing molecule, can control the expression levels of stress-responsive genes, leading to cellular and physiological changes in response to water deficiency [59–61]. Moreover, previous studies have reported two regulatory systems: ABA-dependent and ABA-independent pathways, that play a major role for plant adaptation to drought stress [62]. Genes upstream and within the ABA pathway can be increased under drought conditions, and the *NCED* gene encoded 9-*cis* epoxycarotenoid dioxygenase has been shown to be induced under dehydration

stress [61]. Furthermore, the changes of several metabolite levels under the water deficit condition were associated with the changes of biosynthetic gene expression, many of which were regulated by the changes of ABA accumulation levels [1,63]. In this study, a large number of DEGs were involved in ABA signaling and regulation (Table S5), and the detection of the up-regulation of the *NCED* genes in this study set identified the candidate genes for further studies of drought resistance in common buckwheat seedlings.

Recently, many drought-inducible genes involved in stress tolerance and stress responses have already been identified in several plant species [1], revealing that the transcription factors (TFs) play a central role in the biotic/abiotic stress responses [64–66]. In our previous study, we isolated and identified a *FeDREB1L* gene encoding a DREB-like transcription factor, which was simultaneously involved in the cold stress, drought stress, and ABA-mediated regulations [31]. The increased expression level of *FeDREB1L* during the earlier stage of drought stress displayed in this study (Figure 4m, Figure 5e) demonstrated that *FeDREB1L* could be a positive factor underpinning the drought stress resistance. Other TFs, including AP2/ERF, MYB, and bZIP families, were also identified to be differentially expressed in this study, for example, 180 DEGs that encoded TFs were identified in response to drought stress in both cotyledons and roots of the common buckwheat seedling, as well as the members of C2H2, MYB, bZIP, and WRKY families (Figure 6a, Table S8). Further studies are thereby required to elucidate the functions and gene-regulatory mechanisms of these TFs in response to plant drought stress.

#### **5. Conclusions**

To summarize, a comprehensive transcriptome profile of common buckwheat seedlings under drought stress was obtained using RNA-Seq technology. Phenotypic and physiological changes were determined, and the differentially expressed genes were analyzed to understand the regulatory mechanism of common buckwheat seedlings in response drought stress. The photosynthesis of the common buckwheat seedlings decreased, and the activities of antioxidant enzymes such as CAT and POD were increased under drought conditions. DEGs derived from important regulatory metabolisms were characterized. The results reflected in this study may provide useful information to better understand the molecular mechanism underlying the drought resistance in common buckwheat.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4395/9/10/569/s1, Table S1: List of RT-qPCR primers; Table S2: Phenotypic changes of plant height, root length, and RWC under control conditions across four time points (0, 1, 3, and 5 days); Table S3: Physiological investigation of MDA content, POD and CAT activity, and chlorophyll content under control conditions across four time points (0, 1, 3, and 5 days); Table S4: Summary of the sequencing data of common buckwheat transcriptome; Table S5: Genes related to ABA metabolism in response to drought stress in common buckwheat cotyledons; Table S6: GO enrichment of differentially expressed genes in root transcriptome of common buckwheat seedlings under drought stress; Table S7: Genes related to protein modification and phosphorylation in response to drought stress in common buckwheat roots; Table S8: Classification of TFs that were commonly identified in both cotyledon and root transcriptome libraries; Figure S1: Confirmation of transcriptome data in cotyledons by qPCR analysis; Figure S2: Confirmation of transcriptome data in roots by qPCR analysis.

**Author Contributions:**Z.F., Z.H., and J.Y. designed the study and wrote the manuscript. Z.F., Y.L., J.S., S.W. (Shuping Wang), Z.L., S.W. (Shudong Wei), Y.Z., Z.H., and J.Y. participated in experiments. Z.H. submitted the raw data to Sequence Read Archive (SRA). Z.F., J.S., Z.L., Z.H., and J.Y. discussed the results and revised the manuscript. All authors have read and approved the final manuscript.

**Funding:** This research was funded by the National Natural Science Foundation of China (grant No. 31671755 and No.31571736) and the Supported Project of Outstanding Doctoral and Master's Degree Dissertation Cultivation Program of Yangtze University (YS2018032).

**Acknowledgments:** Authors acknowledge Xiaoyu Xu for critical reading of the manuscript.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **Changes in Root Anatomy of Peanut (***Arachis hypogaea* **L.) under Di**ff**erent Durations of Early Season Drought**

**Nuengsap Thangthong 1,2, Sanun Jogloy 1,2,\*, Tasanai Punjansing 3, Craig K. Kvien 4, Thawan Kesmala 1,2 and Nimitr Vorasoot 1,2**


Received: 12 March 2019; Accepted: 23 April 2019; Published: 27 April 2019

**Abstract:** Changes in the anatomical structure of peanut roots due to early season drought will likely affect the water acquiring capacity of the root system. Yet, as important as these changes are likely to be in conferring drought resistance, they have not been thoroughly investigated. The objective of this study was to investigate the effects of different durations of drought on the root anatomy of peanut in response to early season drought. Plants of peanut genotype ICGV 98305 were grown in rhizoboxes with an internal dimension of 50 cm in width, 10 cm in thickness and 120 cm in height. Fourteen days after emergence, water was withheld for periods of 0, 7, 14 or 21 days. After these drought periods, the first and second order roots from 0–20 cm below soil surface were sampled for anatomical observation. The mean xylem vessel diameter of first- order lateral roots was higher than that of second- order lateral roots. Under early season drought stress root anatomy changes were more pronounced in the longer drought period treatments. Twenty-one days after imposing water stress, the drought treatment and irrigated treatment were clearly different in diameter, number and area of xylem vessels of first- and second-order lateral roots. Plants under drought conditions had a smaller diameter and area of xylem vessels than did the plants under irrigated control. The ability of plants to change root anatomy likely improves water uptake and transport and this may be an important mechanism for drought tolerance. The information will be useful for the selection of drought durations for evaluation of root anatomy related to drought resistance and the selection of key traits for drought resistance.

**Keywords:** xylem vessel; water stress; root anatomy

#### **1. Introduction**

In many areas of the tropics, peanut production is mostly in rain-fed and semi-arid areas with low and unpredictable rainfall and rain distribution. In these areas, drought stress can occur at any growth stage, resulting in yield loss of 22–53% [1]. Drought stress also increases *Aspergillus flavus* infection and aflatoxin contamination by 2–17% [2]. However, drought stress at a pre-flowering growth stage sometimes actually increases yield [3]. Irrigation, planting date selection and drought resistant varieties can improve yield and reduce aflatoxin contamination during periods of drought. However, management of irrigation requires an available water source and investment in additional equipment. Planting date selection, while less expensive than irrigation is not as effective because rainfall and rain distribution are often unpredictable. The use of drought resistant varieties is a promising and sustainable choice in need of further development. When selecting for drought resistance in peanut, yield and biomass during drought are often used as selection criteria. Yet this selection method is complicated by high genotype by environment interaction. Many physiological and morphological traits have been suggested as surrogate traits for drought resistance to increase selection efficiency, yet measurement for these traits are often quite variable.

Root traits are known to improve drought resistance [4] and are therefore important for plant breeding programs. Improving the water acquiring capacity of crops to extract water from the soil profile during drought is one example. Root traits such as large root systems (root dry weight), root length density and the percentage of root length density that respond to drought have been investigated in peanut [1,3,5–7].

Anatomical parameters, such as xylem vessel number and diameter, have been positively correlated with dry matter production under stress in chili (*Capsicum annum* L.) [8]. Drought resistant varieties of several plant species have been reported to have a higher number of vessel cells and a larger xylem cross-section than susceptible varieties of chili (*Capsicum annum* L.) [8], tomato (*Lycopersicon esculentum*) [9] and grape (*Vitis vinifera* L.) [10]. As in the above studies of other plants, it is likely that studies on the fine root structure of peanut, especially under drought stress, will lead to a better understanding of why some peanut genotypes yield better during a drought than others.

Cell-wall ingrowths or phi-thickening have been reported in loquat (*Eriobotrya japonica* Lindl.) root [11], apple (*Pyrus malus*) [12], geranium (*Pelargonium hortorum*) roots [12] Sibipiruna (*Caesalpinia peltophoroides*) [13] with solute movement (salt stress) [12], water logging [13], and drought stress [11]. Although the effect of early season drought on ingrowths and phi-thickenings has not been investigated in peanut and further investigations are necessary to understand phi-thickenings. The response of phi-thickening might be related to the transport processes in the peanut root.

Root anatomy is interesting, and it might play an important role in plant response to drought. The types of lateral roots during root growth were recognized in peanut [14]. The different types and different structures may be related to different functions. The structure of the first order lateral roots helps determine the efficacy of the axial water transport system, yet the structure within the second order lateral roots helps determine the efficacy of the water uptake process. Unfortunately, this useful information has not been thoroughly investigated in peanut. The objective of this study was to investigate the effects of different durations of drought imposition on the root anatomy of peanut in response to early season drought. The information will be useful for selection of drought durations for evaluation of root anatomy related to drought resistance and selection of key traits for drought resistance.

#### **2. Materials and Methods**

#### *2.1. Experimental Design and Plant Material*

The experiment was conducted under a rainout shelter at the Field Crop Research Station of Khon Kaen University, Khon Kaen, Thailand (16◦28 N, 102◦48 E, 200 m. above sea level). The peanut genotype—ICGV 98305—was subjected to four water treatments (0, 7, 14 or 21 days without irrigation), each beginning 14 days after plant emergence (DAE). The experiment used a completely randomized design with three replications and was conducted for two seasons during July–September 2013 and March–May 2014.

ICGV 98305 is a drought resistant line from ICRISAT known for high root length density in the deep sub-soil during periods of drought [1,15].

#### *2.2. Preparation and Irrigation of Rhizobox Experiment*

The plants were grown in rhizoboxes with internal dimensions of 50 cm in width, 10 cm in thickness and 120 cm in height (Figure 1a). The rhizoboxes were filled with dry soil to obtain bulk density of 1.57 Mg m−<sup>3</sup> and height of 115 cm, and water was added to achieve field capacity. Peanut seeds were planted in the center of rhizobox, 5 cm below the soil surface. At 3 days after emergence (DAE), the seedlings were thinned to obtain 1 plant per rhizobox. The front side of the rhizoboxes was transparent and covered with black sheet, and all sides of rhizoboxes were then covered with aluminum foil to reduce light absorption and temperature increase (Figure 1b).

The root needle-board method [16] was used for the observation of root growth and distribution with a minor modification for size and spacing of needles. The root system of the plant in the box was held in place by needles attached to back board of rhizoboxes and projecting out to the transparent front. The needle spacing was 5 <sup>×</sup> 5 cm2. The needle columns started 2.5 cm from left and right margins and the needle rows were started at 12.5 cm from the top of rhizobox and continued at 5 cm intervals to the bottom of the box (Figure 1c).

Soil moisture contents for field capacity and permanent wilting point were determined to be 11.13% and 3.40%, respectively. Water was supplied to the rhizoboxes through horizontal tubes which were installed at 5, 15, 35, 55, 75, 95 and 115 cm below the soil surface. For each rhizobox, water was first supplied at field capacity and all three drought treatments (7, 14 and 21 days without added water) began 14 DAE. The fourth treatment was kept at field capacity for the entire experimental period. The field capacity was maintained uniformly throughout the soil profile by using the six watering tubes. Drainage holes, 1.5 cm in diameter, were placed at the bottom of the rhizobox. Drained water was replenished at the same amount. Crop evapotranspiration was calculated as the sum of water lost through plant transpiration and soil evaporation, as described by Reference [17];

$$\text{ETGroup} = \text{ETo} \times \text{Kc} \tag{1}$$

where ETcrop is crop water requirement (mm/day), ETo is evapotranspiration of a reference under specified conditions calculated using the pan evaporation method, and Kc is the peanut water requirement coefficient.

**Figure 1.** Diagrammatic representation and dimension of rhizobox with six tubes of irrigation (**a**), cross section showing the different elements of the system (**b**), spacing of needle at backside of rhizoboxes (**c**) and taproot system of a rhizobox-grown peanut (**d**).

#### *2.3. Crop Management*

Phosphorus as triple superphosphate (Ca(H2PO4)2H2O) (Chia tai company limited, Phranakhonsiayutthaya, Thailand) at the rate of 122.3 kg ha−<sup>1</sup> and potassium as potassium chloride (KCl; 60% K2O) (Chia tai company limited, Phranakhonsiayutthaya, Thailand) at the rate of 62.5 kg ha<sup>−</sup><sup>1</sup> were applied to the soil before planting. A water-diluted commercial peat-based inoculum of *Bradyrhizobium* (mixture of strains THA 201 and THA 205; Department of Agriculture, Ministry of Agriculture and Cooperatives, Bangkok, Thailand) was applied 5 cm below the soil surface through the irrigation tubes. Seeds were treated with captan (3a,4,7,7a-tetrahydro-2-[(trichloromethyl) thio]-1H-isoindole-1, 3(2H)-dione, Erawan Agricultural Chemical Co., Ltd., Bangkok, Thailand.) at the rate of 5 g kg−<sup>1</sup> seeds before planting. Carbosulfan [2-3-dihydro-2,2-dimethylbenzofuran-7-yl (dibutylaminothio) methylcar-bamate 20% (*w*/*v*) water soluble concentrate] (FMC AG Ltd., Bangkok, Thailand) at 2.5 L ha−<sup>1</sup> was applied weekly to control thrips, and methomyl [S-methyl-N-((methylcarbamoyl)oxy) thioacetimidateand methomyl [(E,Z)-methyl N-{[(methylamino) carbonyl]oxy}ethanimidothioate] 40% soluble powder (Du Pont Co., Ltd., Bangkok, Thailand) at 1.0 kg ha−<sup>1</sup> was used to control mites. Weeds were controlled by hand weeding.

#### *2.4. Data Collection*

Rainfall, relative humidity, pan evaporation, maximum and minimum temperature and solar radiation were recorded daily from planting to 35 DAE at a weather station located 50 m from the experiment. Soil physical and chemical properties were analyzed before planting. Soil samples for analysis were taken from the mixed pile of soil used for this experiment. The soil's physical properties in the experiment were analyzed for percentage sand, silt and clay. The soil chemical properties were analyzed for pH, organic matter, total N, available P, exchangeable K and exchangeable Ca.

#### *2.5. Soil Moisture Content*

Soil moisture content was determined gravimetrically using a micro auger method at 10, 25, 65, and 85 cm soil depths at 14, 21, 28 and 35 DAE. Soil moisture content for each rhizobox was calculated as;

$$\text{Soil moisture content} \left( \% \right) = \left( \left( \text{wet weight} - \text{dry weight} \right) \left( \text{dry weight} \right) \times 100 \tag{2}$$

#### *2.6. Observation of Root Anatomy*

Roots were collected at 7, 14 and 21 days after water withholding began. At the sampling date, the shoot in each box was cut at the soil surface and the roots were carefully washed with a fine spray of tap water to remove soil. Rhizobox needles helped roots maintained the approximate position they were in the soil profile.

Root samples for anatomical observation were taken from 0–20 cm below soil surface. The first-, and second-order lateral roots (Figure 1d) were taken at approximately 5 cm from the root tips from each treatment. The root sampling strategy (5 cm from the tip, and 20 cm deep) was as suggested from a previous rhizotron study [18] in which peanut root growth rates of drought and well-watered treatments were 12.6 and 21.9 cm per week, respectively Therefore, we took the root samples for anatomical study at 5 cm from the root tips, as roots at this position would be expected to be significantly affected by drought. The samples were fixed in a formaldehyde (Sigma-Aldrich; Bangkok, Thailand, 36.5–38% in H2O)-glacial acetic acid (Fisher Chemical)-40% ethanol-solution (FAA40). Dehydration of the samples was accomplished by adding a series of alcohol concentrations at 10% intervals from 40% to 70%. Free-hand cross sections were stained with Safranin O (Dye content ≥ 85%; Sigma-Aldrich). Anatomical characteristics of the root samples were observed using a Nikon eclipse 50i optical microscope with ocular and stage micrometers. The microscope's digital camera (Nikon DS-Fi1, Shingawa-ku, Tokyo, Japan) was used for photographs. All transverse sections of roots were measured and recorded for diameter and area of the xylem vessels of first-order and second-order lateral roots. Xylem vessel elements consisted of protoxylem and metaxylem. Although the identification of these xylem tissues was difficult, we were able to classify them into two groups by diameter. Smaller xylem vessels were equal to or smaller than the overall mean diameter of xylem vessels and bigger xylem vessels were

larger than the mean diameter of xylem vessels. The cell-wall ingrowths were compared in both the drought and well-watered treatments using the cortical layers of both first-order and second-order lateral roots.

#### *2.7. Data Analysis*

The statistical analysis was performed using the statistix-8 program as a completely randomized design. An analysis of variance and least significance difference (LSD) tests were used to compare differences at *p* ≤ 0.05.

#### **3. Results**

#### *3.1. Meteorological Data and Soil Data*

The meteorological details for the two years were collected (data not shown) and are described in Field Crops Research (2016) [19]. Daily air temperatures ranged from 22.7 to 36.8 ◦C in 2013 and 20.2 to 40.5 ◦C in 2014. Relative humidity (RH) values ranged between 63–88% in 2013 and 47–87% in 2014. The means of evaporation (E0) were 4.5 mm in 2013 and 5.7 mm in 2014. While rain did not directly fall on the experimental plants, as it was conducted in a rainout shelter; it did affect relative humidity and evapotranspiration.

Differences between years were observed for maximum temperature (T-max) and minimum temperature (T-min) as the trial in 2013 was conducted during the cooler rainy season (May–July) than the 2014 trial conducted from March–May.

#### *3.2. Soil Moisture Content and Relative Water Content*

Soil moisture content and relative water content are described in Reference [19]. Soil moisture content measured at field capacity was 11.13% and permanent wilting point was 3.40%. Soil moisture content for non-stress conditions was similar to those at field capacity. However, soil moisture content at field capacity (FC) in the lower soil layers was slightly higher than 11.13% at the initiation of drought stress. Drought and well-irrigated treatments were clearly different at all sampling dates, especially at top soil layers of 10 and 25 cm. The differences between drought and well-irrigated treatments were small in lower soil layers and the treatments became similar at 65 and 85 cm except at 28 and 35 DAE in 2014.

#### *3.3. Observation of Root Anatomy*

Peanut has a dicotyledonous root system with a single taproot and branched first-, second-, and higher order lateral roots (Figure 1d). In this study, the anatomy of first- and second- order lateral roots was observed.

#### 3.3.1. First order Lateral Root

Combined analysis of variance for total vessel numbers, bigger vessel numbers, smaller vessel number, total vessel diameter (μm), bigger vessel diameter (μm), smaller vessel diameter (μm), total vessel area (μm2), bigger vessel area (μm2), smaller vessel area (μm2) of the first order lateral root in 2013 and 2014 are shown in Table 1. Significant differences in total vessel numbers, bigger vessel numbers, total vessel area and bigger vessel area were observed in different durations and seasons. The interactions between duration and treatment (D × T) were also significant for total vessel numbers and smaller vessel area traits.

Central cylinders of first order lateral roots had an almost triarch arrangement of the vascular bundles (Figures 2 and 3). Within these bundles, the xylem vessels showed a wide range in size. For ease of discussion, we classified the vessels into two groups (large and small) based on their diameter. Large vessels had a diameter greater than the mean (16.06 μm) of all vessels, and small vessels had a diameter less than the mean.

**Figure 2.** Freehand cross sections of first order lateral roots of peanut under well-irrigated conditions (**a1**, **b1** and **c1**) and drought stress conditions (**a2**, **b2** and **c2**) at 21, 28 and 35 DAE, respectively. CO, cortex; EN, endodermis; P, pericycle; PH, phloem; XY, xylem; Scale bar = 10 μm; 40×.

**Figure 3.** Freehand cross sections of first-order lateral roots under well-irrigated conditions (**a1**, **b1** and **c1**) and drought stress conditions (**a2**, **b2** and **c2**) at 21, 28 and 35 days after plant emergence (DAE). CO, cortex; EN, endodermis; G, phi-thickening or cell wall ingrowth; P, pericycle; PH, phloem; XY, xylem; Scale bar = 10 μm; 40×.

Total xylem numbers per cross-section of first order lateral roots (Figure 4) in the first and second seasons were not significantly different between drought and well-irrigated treatments at 21, 28 and 35 DAE with one exception at 35 DAE in 2014. At 35 DAE in 2014, the drought treatments had higher vessel numbers, in the small diameter vessels, than did well-irrigated treatments. At 35 DAE in 2013, stress and non-stress treatments were not significantly different for the total number of vessels, yet, like in 2014, stress tended to reduce the number of bigger vessels and increase the number of smaller vessel.


ns, \*, = non-significant and significant at *p* < 0.05 and *p* < 0.01 probability levels, respectively, durations (7, 14 and 21 days without added water), treatments (well-watered and water stress) and seasons (2013 and 2014).

#### *Agronomy* **2019** , *9*, 215

**Table 1.** Mean square from the combined analysis of variance for total vessel numbers, bigger vessel numbers, smaller vessel number, total vessel diameter (μm),

bigger vessel diameter (μm), smaller vessel diameter (μm), total vessel area (μm2), bigger vessel area (μm2), smaller vessel area (μm2) of the first order lateral root in

**Figure 4.** Vessel numbers of first order lateral roots (**a1**, **b1**), bigger vessel number (**a2**, **b2**) and smaller vessel number (**a3**, **b3**) of peanut at 21, 28 and 35 DAE in 2013 (**a**) and in 2014 (**b**); Significant at \* *p* ≤ 0.05, non-stress treatments () and stress treatments (-).

Vessel diameters in first order lateral roots (Figure 5) under non-stress and drought stress treatments varied between 4.03 to 41.09 μm (data not shown, unpublished data). Yet, the total vessel area in smaller vessels increased in both 2013 and 2014 and the total vessel area in the large vessels decreased in 2013 and slightly reduced in 2014 when the stress treatments, were compared to the well-watered control (Figure 6) in both 2013 and 2014. Stress and non- stress treatments were not significantly different for vessel diameter at all durations of drought stress. However, the average vessel diameter of long duration stress at 35 DAE and 21 days after irrigation withholding in each season tended to reduce. Figure 5 showed that the diameter of bigger xylem vessels in each season and the diameter of smaller xylem vessels were not significantly different except for the diameter of smaller xylem vessels at 35 DAE in 2014. The diameters of smaller xylem vessels were smaller in size under long duration stress at 35 DAE and 21 days after irrigation withholding compared to under well-watered treatment in 2014.

A significant reduction was observed in the diameter of the smaller xylem vessels and the diameter of the bigger xylem vessels tended to reduce, ultimately reducing total xylem area per root cross section.

The area of total xylem vessel elements in roots grown under stress conditions was significantly lower than those grown under non-stress conditions and these differences in area increased as the length of stress increased. Non-stress and stress treatments were significantly different for the area of total xylem vessels and the area of bigger vessels at 35 DAE. Stress treatment reduced the area of total vessels in 2013 and to a smaller exert the area tended to reduce in 2014.

**Figure 5.** Average vessel diameter of first order lateral roots (**a1**, **b1**), bigger vessel diameter (**a2**, **b2**) and smaller vessel diameter (**a3**, **b3**) of peanut at 21, 28 and 35 DAE under well-irrigate and drought stress in 2013 (**a**) and in 2014 (**b**); Significant at \* *p* ≤ 0.05, non-stress treatments () and stress treatments (-).

**Figure 6.** Vessel area (**a1**, **b1**), bigger vessel area (**a2**, **b2**) and smaller vessel area (**a3**, **b3**) of first order lateral roots of peanut at 21, 28 and 35 DAE in 2013 (**a**) and in 2014 (**b**); Significant at \* *p* ≤ 0.05, non-stress treatments () and stress treatments (-).

The cell-wall ingrowths in the first order lateral roots were detected in the cortical cells under both well-watered and drought stress treatments (Figure 3). The cell-wall ingrowths were localized at the opposite side of the intercellular spaces adjacent to the endodermis except in under drought at 28 DAE (Figure 3b2). The cell-wall ingrowths were found in two positions which were on the opposite side of the intercellular spaces and cell-cell conjunction. The 1–2 layers of this cell were found and indicated as the peri-endodermal layer.

#### 3.3.2. Second Order Lateral Root

Combined analysis of variance for total vessel numbers, bigger vessel numbers, smaller vessel number, total vessel diameter (μm), bigger vessel diameter (μm), smaller vessel diameter (μm), total vessel area (μm2), bigger vessel area (μm2), smaller vessel area (μm2) of the second order lateral root in 2013 and 2014 are shown in Table 2. Differences in duration (D) and treatment (T) were significant (*p* ≤ 0.01 and *p* ≤ 0.05) for most traits. Season (S) was significant for total vessel numbers and bigger vessel numbers. The interactions between duration × treatment (D × T) and duration × season (D × S) were also significant for some traits.

The structure of second order lateral roots differed from that of the first order lateral roots. First order lateral roots are thicker, and the stele and vascular bundle tissues are more extensive than in the second order lateral roots. Second order lateral roots had an almost diarch and triarch organization of vascular bundles (Figures 7 and 8). Average value of vessel diameter was 14.21 μm (data not shown, unpublished data).

Drought and well-irrigated treatments at all durations were not significantly different for number of total xylem per cross-section of second order lateral roots (Figure 9) in 2013 and 2014. Drought and well-watered treatments were also not significantly different in the number of bigger vessels but the number of bigger vessels tended to reduce at 35 DAE, whereas the number of smaller vessels increased at 35 DAE (21 days after water withholding began).

**Figure 7.** Freehand cross sections of second order lateral roots of peanut under well-irrigated conditions (**a1**, **b1** and **c1**) and drought stress conditions (**a2**, **b2** and **c2**) at 21, 28 and 35 DAE. CO, cortex; EN, endodermis; P, pericycle; PH, phloem; XY, xylem; Scale bar = 10 μm; 40×.

**Figure 8.** Freehand cross sections of second-order lateral roots under well-irrigated conditions (**a1**, **b1** and **c1**) and drought stress conditions (**a2**, **b2** and **c2**) at 21, 28 and 35 DAE. CO, cortex; EN, endodermis; G, phi-thickening or cell wall ingrowth; P, pericycle; PH, phloem; XY, xylem; Scale bar = 10 μm; 40×.


ns, \*, = non-significant and significant at *p* < 0.05 and *p* < 0.01 probability levels, respectively, durations (7, 14 and 21 days without added water), treatments (well-watered and water stress) and seasons (2013 and 2014).

#### *Agronomy* **2019** , *9*, 215

**Table 2.** Mean square from the combined analysis of variance for total vessel numbers, bigger vessel numbers, smaller vessel number, total vessel diameter (μm),

bigger vessel diameter (μm), smaller vessel diameter (μm), total vessel area (μm2), bigger vessel area (μm2), smaller vessel area (μm2) of the second order lateral root

Means for the vessel diameter of second-order lateral roots (Figure 10) of all treatments varied between 4.29 to 38.48 μm (data not shown). Stress treatment significantly reduced the vessel diameter of second-order lateral roots at 35 DAE with drought imposition for 21 days in 2013 and slightly reduced the vessel diameter of second-order lateral roots at 35 DAE with drought imposition for 21 days in 2014. Stress treatment significantly reduced the diameter of bigger xylem vessels in 2014 at 35 DAE with drought imposition for 21 days and stress treatment also reduced the diameter of bigger xylem vessels in 2013, although the reduction was not significant. Stress treatment did not significantly affect the diameter of smaller xylem vessel diameter in 2013 and 2014.

**Figure 9.** Vessel numbers of second order lateral roots (**a1**, **b1**), bigger vessel number (**a2**, **b2**) and smaller vessel number (**a3**, **b3**) of peanut at 21, 28 and 35 DAE in 2013 (**a**) and in 2014 (**b**); Significant at \* *p* ≤ 0.05, non-stress treatments () and stress treatments (-).

**Figure 10.** Vessel diameter of second order lateral roots (**a1**, **b1**), bigger vessel diameter (**a2**, **b2**) and smaller vessel diameter (**a3**, **b3**) of peanut at 21, 28 and 35 DAE in 2013 (**a**) and in 2014 (**b**); Significant at \* *p* ≤ 0.05, non-stress treatments () and stress treatments (-).

Because stress treatment reduced the diameters of the average xylem vessels and bigger xylem vessels, the area of vessels per cross section of each season and the area of bigger vessels in 2014 was reduced at 35 DAE, although the reduction was not significant and the area of bigger vessels area was significantly reduced at 35 DAE in 2013 (Figure 11). The area of smaller xylem vessels per cross section under stress treatment was increased.

**Figure 11.** Vessel area (**a1**, **b1**), bigger vessel area (**a2**, **b2**) and smaller vessel area (**a3**, **b3**) of second order lateral roots of peanut at 21, 28 and 35 DAE in 2013 (**a**) and in 2014 (**b**); Significant at \* *p* ≤ 0.05, non-stress treatments () and stress treatments (-).

Cell-wall ingrowths appeared in the cortical cells of the second order lateral root under both conditions (Figure 8). The 1–2 layers of cell-wall ingrowths were found in the peri-endodermal layer.

#### **4. Discussion**

Weather conditions may be a key factor affecting the root anatomy of peanut. The experiment was conducted for two years. In the rainy season, air temperature and humidity were low, but in the summer to the early rainy season, air temperature and humidity were rather high. Soil moisture in the drought and well-watered treatments were clearly different in the upper soil layers. Soil moisture content for drought stress treatment at 28 and 35 DAE at the 10 cm of soil layer was less than 3.4% (the permanent wilting point). However, soil moisture content for drought stress treatment at 65 cm and 85 cm of soil levels was higher than the permanent wilting point. The rate of water loss in 2013 was slower than in 2014, and soil moisture content at 21 days after irrigation withholding in 2013 was similar to those at 14 days after irrigation withholding in 2014.

The responses of plants to water stress depend on many things including timing and the intensity and duration of the drought. Root anatomy and root growth, like other plant parts, are sensitive to drought [20]. In this study, the long duration of the early season drought changed the root anatomical traits of peanut. Long periods of stress caused a significant increase in the number of xylem vessels in first and second order lateral roots but a significant decrease in the vessel diameter and the area of these first and second order lateral roots.

In both seasons, the mean xylem vessel diameters of first order lateral roots was higher than that of the second order lateral roots. The reduction in vessel diameter of first order lateral roots was higher than that of the second order lateral roots and these results may explain the differential root functions. The reduction in vessel diameter of first order lateral roots will better support the transport system's hydraulic conductivity according to Poiseuille's law [21]. In hot pepper, drought stress significantly reduced the diameters of xylem vessels in all of cultivars [8]. Vessel diameter is closely and positively correlated with volume of water flow and therefore it is correlated with the safety of the conductive system [22,23]. The large vessel size under water deficit resulted in xylem cavitation [24]. The narrower diameter of metaxylem vessels maintain the water column, lowers the risk of cavitation, increases water flow resistance and saves water columns in narrower capillaries from damage [25]. Formation of narrower vessels occurring in drought-tolerant dicotyledons (including short-lived perennials and annuals with secondary structure) will likewise be advantageous when the plants are grown under drought [26,27].

Morphometric measurements on xylem vessels showed that the vessels of water-stressed plants had lower sectional areas. These results suggested that the reduction in vessel sectional area due to a diminished growth in response to water stress was the main factor affecting conductivity. Under a water deficit environment, roots develop to help extract soil moisture which being held at greater surface tension [28]. Deep root growth and large xylem diameter in deep roots may also increase the ability of roots to mine more water in deep soil when water in deep soil is abundant [29]. However, small and fine roots with greater specific root length enable plants to efficiently increase water uptake and maintain plant productivity under drought by increasing surface area and root length in contact with soil water, especially at deeper soil with available water [19,29].

The ability of plant to take up water is highly influenced by the number and size of the water conductive elements [25]. The change in number and size of the vessel xylem could help maintain water uptake under water stress [8].

In Ferna'ndez-Garcı'a, Lo'pez-Berenguer, and Olmos book chapter on the role of phi Cells under abiotic stress the authors noted that phi thickening is not the exception in the root anatomy [30]. They noted that the literature has described 16 different families, covering more than 100 species, which present the phi thickening in the roots. The phi thickening is classified into three types based on their root cell location: Type I, the most frequently found phi cell layer, is located in contact with the endodermis. Type II phi cell layer is located in contact with the epidermis and Type III phi cell layers are located in the inner cortical cells but not in contact with either the epidermis or the endodermis. In this study, cell-wall ingrowths were detected in the cortical cells of all first and second order lateral roots under well-watered and drought stress treatments. The 1–2 layers of these cells were localized at the opposite side of the intercellular spaces adjacent to the endodermis. The cell-wall ingrowths layers were indicated as the peri-endodermal layer and also called phi-thickening [31]. In previous studies, phi-thickening was induced under salt stress [30,31] and drought stress [11]. Phi-thickening of loquat roots grown under drought stress developed dramatically compared to normal conditions and the formation of phi-thickening was thought to be a defense mechanism against water stress. As the functions of these cells are difficult to determine precisely, phi thickening would play a role

in controlling the water and solute rate of transportation through cell walls [32]. In peanut, cell wall ingrowth development in cortical cells might be a drought resistance mechanism for peanut roots as well. In this study, the 1–2 layers of cell-wall ingrowths were detected in both well-watered and drought stress treatments which were not significantly different for number of cell-wall ingrowths layers. However, the cells could be seen at higher magnification and using an electron microscope.

#### **5. Conclusions**

Under early season drought stress, root anatomy changes were more pronounced in the longer drought period treatments. At 21 days after imposing water stress, the drought treatment and irrigated treatment were clearly different in diameter, number and area of xylem vessels of first- and secondorder lateral roots. Plants under drought conditions had smaller diameter and area of xylem vessels than did the plants under irrigated control. The ability of plant to change root anatomy likely improves water uptake and transport, and this may be an important mechanism for drought avoidance.

**Author Contributions:** Conceptualization, N.T., S.J., T.P. and N.V.; methodology, N.T., S.J. and N.V.; validation, N.T., S.J. and N.V.; formal analysis, N.T.; investigation, N.T.; resources, S.J.; data curation, N.T.; writing—original draft preparation, N.T.; writing—review and editing, T.K., C.K.K.; supervision, S.J.; funding acquisition, S.J.

**Funding:** This research was funded by the Royal Golden Jubilee Ph.D. Program (6.A.KK/ 53/ E.1), Peanut and Jerusalem artichoke Improvement Project for the Functional Food Research Group, and the Thailand Research Fund for providing financial support through the Senior Research Scholar Project of Sanun Jogloy (Project no. RTA6180002).

**Acknowledgments:** This study was funded by the Royal Golden Jubilee Ph.D. Program (6.A.KK/ 53/ E.1). Assistance was also received from Peanut and Jerusalem artichoke Improvement Project for the Functional Food Research Group, Plant Breeding Research Center for Sustainable Agriculture and the Thailand Research Fund for providing financial support through the Senior Research Scholar Project of Sanun Jogloy (Project no. RTA6180002). Thailand Research Fund (TRF) (IRG 578003), Khon Kaen University (KKU) and Faculty of Agriculture, KKU are acknowledged for providing financial support for training on manuscript preparation. The manuscript was critical reviewed by Ian Charles Dodd.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

**and Tobacco**
