Peanut “Luhua14” seeds of uniform size were used. Plump seeds were first soaked in tap water for 12 h. The imbibed seeds were sown in round plastic pots filled with clean river sand. After germination, the seeds were irrigated with half-strength Hoagland solution until the two-leaf stage was attained. Seedlings at the two-leaf stage were established in hydroponic culture with Hoagland solution in a plastic bucket of 30 cm diameter. Salt treatment was applied after hydroponic culture for 10 days. The NaCl concentration in the nutrient solution was increased stepwise from 50 to 250 mM at the rate of 50 mM, with each concentration applied for 12 h. Control seedlings were not treated with NaCl (CK). Seedlings were treated with 250 mM NaCl for 4 days (S4), then transferred from salt stress to standard (ST) conditions for 4 days (R3). After recovery treatment for 3 days, physiological parameters were measured (Figure S1). Then the seedlings were cultured at 28 ± 5°C (day/night) with light intensity of 600 μmol m⁻² s⁻¹, a 15 h light/9 h dark photoperiod, and 70% relative humidity.

Fifteen seedlings from each of the three treatments (five seedlings per replicate) were sampled and the seedling fresh weight (FW) was recorded. The sampled seedlings were dried at 105°C for 15 min and at 70°C for 72 h, and then the seedling dry weight (DW) was recorded.

Leaves from the peanut seedlings of 15 seedlings from each of the three treatments (five seedlings per replicate) were sampled and frozen immediately in liquid nitrogen. Lipids were extracted in accordance with the method described by Siegenthaler and Eichenberger (1984), followed by separation using two-dimensional thin-layer chromatography (TLC) (Xu and Siegenthaler, 1997). For quantitative analysis, fatty acid methyl esters were prepared after the separation of lipids by TLC. Fatty acid composition was determined by gas chromatography (GC-9A, Shimadzu, Kyoto, Japan) using the method described by Chen et al. (1994).

Chlorophyll fluorescence was measured after treatment with 250 mM NaCl for 4 days (S4) and after recovery for 13 days (R3). Chlorophyll fluorescence was independently measured on five plants with a portable fluorometer (FMS2, Hansatech, King's Lynn, UK) following the method described by Kooten and Snel (1990), and used in a previous study (Sui, 2015). Leaves were kept in the dark for 1 h before measurement of minimal and maximal fluorescence. Minimal fluorescence (F_o) with all PSII reaction centers open was determined with modulated light that was sufficiently low to prevent significant variable fluorescence yield in the dark-adapted state (F_v). Maximal fluorescence (F_m) with all reaction centers closed was determined by irradiating a dark-adapted leaf for 0.8 s with saturating light of 8,000 μmol m⁻² s⁻¹. Then, the leaf was illuminated with actinic light of 500 μmol m⁻² s⁻¹. Steady-state fluorescence (F_s) was recorded when the leaf attained steady-state photosynthesis. A second pulse of 0.8 s of saturating light of 8,000 μmol m⁻² s⁻¹ was applied to determine the F_m in the light-adapted state (Fm′) (Sui, 2015). The maximal photochemical efficiency (F_v/F_m) of PSII was expressed as F_v/F_m = (F_m − F_o)/F_m. The quantum yield of PSII electron transport was determined using the formula ΦPSII = (Fm′ − F_s)/Fm′. Non-photochemical quenching (NPQ) was calculated as NPQ = F_m/Fm′ − 1. The chlorophyll fluorescence of seedlings that underwent recovery treatment for 3 days was measured in the same manner.

The capacity of PSI was determined by measuring the absorbance at 820 nm (ΔI/I_o) using a Plant Efficiency Analyser (Hansatech, King's Lynn, UK) after the leaves were treated in the dark for 30 min (Schansker et al., 2003; Qin et al., 2011). The first reliable measurement point for fluorescence change was at 20 μs, whereas the first measurement point for transmission change was at 400 μs. The transmission measurement point was 100 μs. Light intensity was 3,000 μmol m⁻² s⁻¹ photon flux density. The far-red source was a QDDH73520 light-emitting diode (LED; Quantum Devices Inc., Barneveld, WI, USA) filtered at 720 ± 5 nm. The modulated (33.3 kHz) far-red measuring light was provided by an OD820 LED (Opto Diode Corp., Newbury Park, CA, USA) filtered at 830 ± 20 nm.

Ascorbate peroxidase (APX) activity was determined based on the decrease in absorbance at 290 nm. The reaction mixture contained 50 mM potassium phosphate buffer (pH 7.0), 0.5 mM ascorbate, 0.2 mM H₂O₂, and a suitable volume of enzyme extract (Jimenez et al., 1997). Superoxide dismutase (SOD) activity was determined in accordance with the method described by Giannopolitis and Ries (1977). The leaves (same position, 0.5 g) were homogenized using a pre-cooled mortar and pestle in 5 mL reaction mixture containing 50 mM potassium phosphate buffer (pH 7.0) and 1% polyvinylpyrrolidone. The homogenate was centrifuged at 12,000 rpm for 10 min to collect the supernatant.

Superoxide anion (O2-•) content was assayed in accordance with the method described by Wang and Luo (1990). Fresh leaves with the midrib excised were ground in 0.05 M phosphate buffer (pH 7.8) in an ice bath. The homogenate was centrifuged (5,000 × g) at 4°C for 10 min to collect the supernatant. The supernatant was mixed with phosphate buffer (pH 7.8) and 10 mM hydroxyl ammonium chloride and incubated at 25°C for 20 min. After addition of 17 mM p-aminobenzene sulfonic acid and 7 mM α-naphthylamine, the mixture was incubated at 25°C for 20 min. Finally, ethyl ether was added to the mixture, followed by centrifugation at 1,500 × g for 5 min. Absorbance at 530 nm was determined in the water phase. The O2-• generation was expressed as the content per gram of fresh leaf mass (Sui et al., 2007).

Hydrogen peroxide (H₂O₂) content was determined using the method described by Sairam and Srivastava (2002). The concentration of H₂O₂ was estimated by measuring the absorbance of the titanium–hydroperoxide complex, followed by mapping to a standard curve plotted with H₂O₂ standards (Sui et al., 2007).

The relative electric conductivity (REC) was assayed in accordance with the method described by Guo et al. (2015), using the two-leaf stage seedlings, which were treated for 4 days by various NaCl concentrations (50, 100, 150, 200, 250, 300, and 400 mM). 30 disks from the first three true leaves were placed into a tube containing 10 ml of distilled water. The tube was then shaken for 12 h at 180 rpm and the initial electric conductivity of the solution (S1) was measured. Then tubes were kept in the boiling water for 10 min and cooled down to room temperature. The final electric conductivity (S2) was then measured. The relative electric conductivity (REC) was calculated as follows: REC (%) = S1/S2 × 100.

Total RNA was extracted from leaf tissue sampled from seedlings in the S4, R3, and CK treatments using the RNA plant Plus reagent (DP437, Tiangen, Beijing, China) in accordance with the manufacturer's instructions (Sui et al., 2015; Yuan et al., 2015; Yang et al., 2017). RNA extracts were quantified using a Nanodrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Wilmington, DE, USA). The extracts were electrophoresed in 1% agarose gel buffered with Tris–acetate–EDTA to determine the integrity of the RNA. RNA libraries were constructed following the High Throughput Illumina Strand-Specific RNA Sequencing Library protocol (Zhong et al., 2011). Briefly, mRNA was purified from 5 μg total RNA using oligo (dT) magnetic beads. The purified mRNA was fragmented into small pieces using a fragmentation buffer. Using these short fragments as templates, first-strand cDNA was synthesized using reverse transcriptase and random hexamer primers. Second-strand cDNA synthesis was followed using DNA polymerase I and RNase H.

To obtain sequences of high similarity, unigene sequences were first aligned to sequences lodged in the protein databases NR (Deng et al., 2006), KEGG (Kanehisa et al., 2004), Swiss-Prot (Apweiler et al., 2004), and COG (Tatusov et al., 2000) (e-value < 0.00001) using the blastx tool, and the nucleotide database NT (e-value < 0.00001) using blastn. The KEGG database enables analysis of the gene product during the metabolism process and the related gene function in the cellular processes. KEGG pathways were assigned to the assembled sequences using the online KEGG web server (http://www.genome.jp/kegg/).

Six tag libraries were constructed for both shoots and roots sampled for each of the CK, S4, and R3 treatments. Reads per kilobase per million (RPKM) values were calculated using an in-house script based on the count table of Cuffdiffs output (http://cole-trapnell-lab.github.io/cufflinks/). The mapping and detection of DEGs were performed as described in a previous study (Sui et al., 2015). The DEGs with a fold change of ≥2 and divergence probability ≥0.8 between different samples were identified using the Noiseq method (Tarazona et al., 2011).

Quantitative real-time RT-PCR (qRT-PCR) analysis was performed to validate the RNA-seq results. Twelve DEGs were randomly selected for qRT-PCR. All primers were designed using the Beacon Designer software (version 7.9) (Additional file: Table S1). The housekeeping gene Tubulin (GenBank accession GO264294) was used as an internal standard (Chi et al., 2012). Total RNA (1 μg) was used in a 20 μl reaction volume for reverse transcription using the ReverTra Ace® qPCR RT Kit (Toyobo, Japan) in accordance with the manufacturer's instructions. Real-time PCR was performed in a 20 μl reaction volume containing 10 μl SYBR® Premix Ex Taq, 0.5 μl primer pairs (tubulin and target gene), and 2 μl (50 ng μl⁻¹) cDNA. Real-time PCR was carried out on a real-time quantitative PCR instrument (LightCycler® 96, Roche Diagnostics, Mannheim, Germany). The reaction conditions were as follows: 95°C for 5 min, followed by 40 cycles of 95°C for 15 s, 55°C for 20 s, and 72°C for 20 s. Three biological replicates were included in each experiment.

The ω-3 fatty acid desaturase protein was extracted using the following steps. Leaf samples (0.2 g) were frozen in liquid nitrogen, then ground in 2 ml extraction buffer (100 mM phosphate buffer, pH 7.2, containing 0.1 mM EDTA and 2 mM ascorbic acid). The homogenate was centrifuged at 10,000 rpm for 15 min, and the supernatant was used to measure enzyme activity. The activity of ω-3 fatty acid desaturase was measured with a ω-3 fatty acid desaturase activity kit using the double antibody sandwich method (Jie Shi Kang Biological Technology Co., Ltd, Qingdao, China).

The software package SPSS 16.0 (SPSS, Chicago, IL, USA) was used for all statistical analyses following the method described in a previous study (Cheng et al., 2014). All data are presented as the mean (±SD) of five replicates (n = 5). Comparisons among multiple groups were performed using Duncan's multiple range test. Probability values P < 0.05 were considered statistically significant.

We analyzed the relative electric conductivity (REC) using the two-leaf stage seedlings, which were treated for 4 days by different NaCl concentrations of 50, 100, 150, 200, 250, 300, and 400 mM. Result showed that there was no significantly difference before 150 mM NaCl treatment. REC increased significantly at 250 mM NaCl treatment. According to these data, 250 mM NaCl for 4 days (S4) were used for further experiments (Figure S2). Treatment with 250 mM NaCl for 4 days significantly inhibited growth of the peanut seedlings (Figure 1A). Furthermore, the seedling fresh and dry weights decreased by 42.9 and 17.6%, respectively, under the S4 condition (Figure 1B). No significant differences in seedling fresh and dry weights were observed between the R3 and S4 conditions (Figure 1A).

Treatment with 250 mM NaCl for 4 days increased the contents of the saturated fatty acids palmitic acid (16:0) and stearic acid (18:0) and the unsaturated fatty acid 18:2, whereas the contents of palmitoleic acid (16:1) and 18:3 decreased (Table 1). Under the S4 condition, compared with CK, the 16:1 content decreased by 38.3% and the 18:2 content increased by 13.2%. In addition, the double bond index (DBI) and 18:3 content after S4 treatment decreased by 30.2 and 42.5%, respectively. Furthermore, the 16:1 content decreased by 8.5% and the 18:2 content increased by 5.7% after R3 treatment compared with those observed after S4 treatment. Moreover, the DBI and 18:3 content increased by 18.1 and 27.6%, respectively, after R3 treatment compared with those observed after S4 treatment.

To evaluate the effect of the salt and recovery treatments on PSII capacity, we measured the F_o, NPQ, ΦPSII and F_v/F_m of peanut leaves in each treatment. F_o reflects the minimum fluorescence intensity after dark adaptation. The direction of F_o depends on the dominant factor for PSII inactivation, or damage and energy dissipation. Under the S4 condition, F_o and NPQ increased by 17.7 and 79.6%, respectively. Although NPQ decreased by 5.5% under the R3 condition, compared with that observed under the S4 condition, the difference was not statistically significant (Figure 1C). Under the S4 condition, ΦPSII and F_v/F_m decreased by 12.7 and 8.5%, respectively, compared with the control. However, ΦPSII and F_v/F_m increased by 8.9 and 3.0%, respectively, under the R3 condition compared with those observed under the S4 condition (Figure 1C).

Treatment with 250 mM NaCl decreased PSI oxidoreductive activity (ΔI/I₀) in the leaf by 49.3% (Figure 1D). However, no significant differences in PSI oxidoreductive activity were observed between the R3 and S4 conditions (Figure 1D). Therefore, PSI oxidoreductive activity was reduced under salt stress but was not reversed by recovery.

Compared with the control, under the S4 and R3 conditions SOD activity decreased by 64.4 and 50.8%, respectively (Figure 1E), and APX activity decreased by 76.3 and 68.5%, respectively (Figure 1E). However, the activities of SOD and APX increased by 38.4 and 33.0%, respectively, under the R3 condition compared with those observed under the S4 condition. The contents of O2-• and H₂O₂ exhibited a negative correlation with APX and SOD activity under the S4 and R3 conditions. The contents of O2-• and H₂O₂ increased by 35.4 and 61.1%, respectively, after treatment with 250 mM NaCl for 4 days (Figure 1E). Under the R3 condition, O2-• and H₂O₂ content decreased by 9.6 and 11.3%, respectively, compared with those observed under the S4 condition. Hence, salt stress inhibited antioxidant enzyme activity and increased the contents of O2-• and H₂O₂, but recovery treatment reversed these changes.

To investigate the molecular mechanisms of salt tolerance in peanut seedlings and the recovery condition, RNA samples were extracted from leaves of the control (without NaCl treatment), salt-stress treatment (S4), and recovery treatment (R3) to prepare libraries for RNA-seq analysis. After stringent quality assessment and data filtering, 263.13 million clean reads were obtained (Table S2). Pearson correlation coefficients with a divergence probability ≥0.8 between RNA-seq samples were evaluated based on the square of the Pearson's correlation coefficient. The R² values were all higher than 0.92, which indicated high repeatability (Figure S4). Compared with the control, 1,742 genes were differentially expressed under the S4 condition. Among these DEGs, 898 genes were up-regulated and 844 genes were down-regulated in the leaf under salt stress. Furthermore, 390 genes were differentially expressed under the R3 condition compared with the S4 condition. Among these DEGs, 323 genes were up-regulated and 67 genes were down-regulated in the leaf (Figure 2). A total of 120 genes differentially expressed between the S4 and CK conditions were mapped to 13 pathways associated with lipid metabolism (Figure 3). In addition, 34 genes differentially expressed between the S4 and R3 conditions were mapped to 11 pathways (Figure 3).

As important components of cell membrane phospholipids, polyunsaturated fatty acids determine the fluidity and deformability of the cell membrane by maintaining the structure and function of the cell membrane. Previous studies have observed that lipids might play important roles in protecting the plant against salt stress (Huflejt et al., 1990; Khamutov et al., 1990; Ritter and Yopp, 1993; Sui et al., 2017). In the present study, 26 DEGs were mapped to five lipid metabolism-related pathways, and 11 DEGs were mapped to the fatty acid metabolism pathway (Table 2). Five genes were categorized as long-chain acyl-CoA synthetase. Among these five genes, four genes were down-regulated and one gene was up-regulated under the S4 condition. Gene CL639.Contig19, which was down-regulated under the S4 condition and up-regulated under the R3 condition, was categorized as oxidoreductase. Three genes that encoded alcohol dehydrogenase were up-regulated under the S4 condition. The DEGs that encoded the protein VITISV_013417 and MFP-a were up-regulated under the S4 condition, but no significant changes were observed under the R3 condition.

Five DEGs were mapped to the biosynthesis of the unsaturated fatty acids pathway. Among these five DEGs, two genes (CL8534.Contig6 and CL8534.Contig7) were categorized as ω-3 fatty acid desaturase. These two genes were involved in the synthesis of 18:3 from 18:2 (Additional file: Figure S3). The expression of these two genes decreased under the S4 condition but increased under the R3 condition (Table 2). The unigene10822 and unigene19394, which encode short-chain-type dehydrogenase and 3-ketoacyl-CoA thiolase, were up-regulated under the S4 condition. The DEG that encoded the elongation of fatty acids protein was down-regulated under the S4 condition.

Five DEGs were mapped to the α-linolenic acid metabolism pathway (Table 2). A DEG that encoded 4-coumarate-CoA ligase was down-regulated under the S4 condition. CL7132.Contig3, which encoded 7-methylxanthosine synthase 1, was up-regulated under the S4 condition. Three DEGs that encoded 12-oxophytodienoate reductase 2 were differentially expressed under the R3 condition. Furthermore, the expression levels of unigene27104 and unigene27102 increased under the R3 condition, whereas the expression level of CL3328.Contig1 decreased (Table 2).

Three DEGs were mapped to the fatty acid biosynthesis pathway. Two DEGs that encoded serine-type endopeptidase were down-regulated under the S4 condition. The expression of unigene10129, which encoded glucose 1-dehydrogenase, was up-regulated under the S4 condition.

Two DEGs were mapped to the linoleic acid metabolism pathway. The unigene11867, which encoded oxidoreductase, was down-regulated under the S4 condition. The expression level of CL1272.Contig5, which encoded retinol dehydrogenase, also decreased under the S4 condition, but increased under the R3 condition (Table 2).

To validate the RNA-seq data, qRT-PCR analysis were performed on 14 randomly selected DEGs. As shown in Figure 4, the high reliability of the RNA-seq data was confirmed by the high correlation value (R² = 0.97).

The expression of CL8534.Contig6 was down-regulated under treatment with different concentrations of NaCl. After treatment with 100, 150, 200, and 250 mM NaCl, the expression level of CL8534.Contig6 decreased by 48.1, 58.9, 71.3, and 82.9%, respectively (Figure 5A). Under treatment with 250 mM NaCl, the expression level of CL8534.Contig6 gradually decreased with prolonged duration of exposure. Under the R3 condition, the expression level of CL8534.Contig6 increased by 111.8% compared with the level measured after treatment with 250 mM NaCl for 4 days (Figure 5B).

Treatment with 250 mM NaCl decreased the activity of ω-3 fatty acid desaturase by 32.1%. Furthermore, the activity of ω-3 fatty acid desaturase increased by 23.0% under the R3 condition, compared with that observed under the S4 condition, but remained lower than that of the CK (Figure 6).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.