Tamarix hispida seeds were collected from Turpan Desert Botanical Garden of Chinese Academy Sciences (Xinjiang province, China) and were planted in a mixture of sand and turf peat (1:2, v/v) with the conditions of 14 h light/10 h darkness photocycle and 70–75% relative humidity under the temperature of 24°C. The seedlings (approximately 5–6 cm in height, 2-month old) with good growth and uniform size were used for stress treatment. The seedlings were irrigated every 24 h on roots with the solution of 20% (w/v) PEG₆₀₀₀, 0.4 M NaCl, 100 μM ABA, 50 μM GA3 or 100 μM JA solution and were harvested at 6, 12, 24, 48, and 72 h. Meanwhile, the T. hispida plants irrigated with fresh water were served as the controls. After these treatments, roots or leaves from about 20 seedlings were pooled together at each time point, and stored at -80°C.

Seven transcriptome libraries had been constructed with T. hispida plants with 2-month-old, including 4 root transcriptomes treated respectively with NaHCO₃ for 0, 12, 24, and 48 h, and 3 leaf transcriptomes treated with NaHCO₃ for 0, 12, and 24 h (Wang et al., 2014). The unigenes were analyzed using BLASTX program for searching the Swiss-Prot and NR databases. The phrase “MYB transcription factor (TFs)” was searched against the unigenes with functional annotation to identify MYB family genes. Then, the ThMYB genes with complete open reading frames (ORFs) were selected by ORF finder¹. The Mol.Wt and theoretical pI of the proteins encoded by ThMYBs were studied by ProtParam².

Arabidopsis MYB proteins were retrieved from TAIR database³. The MEGA5.0 software (Tamura et al., 2011) was used to systematically analyze the MYB proteins of Arabidopsis and T. hispida. Neighbour-joining (NJ) was employed to predict the phylogenetic tree, which was subjected to a Bootstrap method with 1000 replications (Tamura et al., 2011). Multiple sequence alignments was performed by Clustal X to search the conserved regions of the ThMYBs with the gap extension penalties and a gap open of 10 and 0.1, respectively (Thompson et al., 1997). ExPASy-PROSITE⁴ was used to predict the conserved domains of ThMYBs protein sequences.

Total RNA was extracted from T. hispida samples using an RNA extraction kit (BioTeke Corporation, Beijing, China). About 1 μg of total RNA was transcribed into cDNA. The synthesized cDNA was diluted to 10-fold with sterile water to act as the template for qRT-PCR. In addition, actin (FJ618517), α-tubulin (FJ618518), and β-tubulin (FJ618519) were used as the internal reference genes in T. hispida to normalize the templates in the PCR reactions. All primer sequences used were shown as Supplementary Table S1. qRT-PCR was conducted on an Opticon 2 system (Bio-Rad). The PCR reaction with a volume of 20 μl contains the followings: 2 μl of template cDNA, 0.5 μM of each reverse and forward primer, 10 μl of SYBR Green Real-time PCR Master Mix (Toyobo). PCR program was set as following conditions: 95°C for 30 s, followed by 45 cycles of 95°C for 15 s, 58°C for 15 s, 72°C for 30 s and 1 s at 80°C for reading plate. All the reactions were performed in triplicate. Expression levels were determined according to the 2^-ΔΔCt (Livak and Schmittgen, 2001).

To verify the salt tolerance function of ThMYB genes, the overexpression and RNAi vectors of ThMYB13 were constructed. The CDS of ThMYB13 was inserted into pROKII driving by 35S CaMV promoter to overexpress ThMYB13 (35S:MYB). An inverted repeat truncated CDS of ThMYB13 with the length of 246 bp was inserted into pFGC5941 (Kerschen et al., 2004) at the two sides of the CHSA intron (pFGC:MYB) for silencing ThMYB13 expression. All primer used are shown in Supplementary Table S2. The genetic transformation of T. hispida plants was carried out following Ji et al. (2014). In brief, 40 day tissue cultures of T. hispida seedlings with similar sizes were used for genetic transformation. The seedlings were incubated in a solution for transformation [1/2 MS + 150 μM acetosyringone + 3% (w/v) sucrose + 0.01% (w/v) Tween 20 + 0.6 OD₆₀₀ Agrobacterium tumefaciens, pH 5.6] shaking with 120 rpm at 25°C for 6 h. The seedlings were washed twice using distilled water and planted vertically on 1/2 MS solid medium. In total, 3 types of transiently transformed plantlets were cultured, including plants transiently transformed with 35S:MYB to overexpress ThMYB13 (OE), with pFGC:MYB for silencing ThMYB13 (RNAi) expression, or control (Con) plants transformed with the empty pROKII plasmid.

After culturing for 36 h on 1/2 MS agar medium, the seedlings from transient transgenic T. hispida (OE, Con and RNAi) were transferred to new 1/2 MS medium (as control) or 1/2 MS with 150 mM NaCl stress 2 h for biochemical staining, and then these seedlings were incubated with NBT, DAB or Evans blue solutions. The procedures for NBT and DAB staining were according to Zhang et al. (2011). Evans blue staining was performed as described by Yang et al. (2014). H₂O₂ was measured following the protocol of Dal Santo et al. (2012).

In addition, after culturing for 36 h on 1/2 MS solid medium, the plants were transferred to 1/2 MS medium with 150 mM NaCl for stress for 12, 24, or 36 h. The plants were harvested for qRT-PCR and physiological analyses. The expression of ThMYB13 in the transgenic plants were measured with qRT-PCR as described above. MDA measurements were conducted descried by Wang et al. (2010), while the electrolyte leakage measurement was conducted as described by Ji et al. (2014). The determination of the Na⁺ and K⁺ content was described in Chen et al. (2005) and Ma et al. (2011). All experiments were conducted with triplicate.

The data were analyzed with the Statistical Software Package for Social Science (SPSS) version 17.0. Using Student’s t-test to compare the data, if P < 0.05, the data were considered significantly different. ^∗Indicates significant difference (^∗P < 0.05).

A total of 238 unigenes were found to be the MYB family genes from the seven transcriptomes. BLASTX analysis was performed, and 14 unigenes of the ThMYBs with full ORF were finally obtained and searched in the NR protein database. In addition, these 14 ThMYB genes were named ThMYB1 to ThMYB14. The proteins of these 14 ThMYBs ranged from 284 to 554 amino acids residues, their pI were between 5.05 and 9. 66 and the Mol.Wt ranging from 30.6 to 61.8 kDa (Table 1).

The conserved domains of 14 ThMYB protein sequences were analyzed using ExPASy-PROSITE. The results showed that the structural domain of each ThMYB protein had at least one MYB conserved domain. To further analyze the similarities and characterization of the protein sequences of 14 ThMYB, their deduced amino acid sequences of were sequenced by Clustal X software. As shown in Supplementary Figure S1, 14 ThMYB protein sequences have a conserved MYB domain, and the conserved domains have high similarity. The N-terminus of ThMYB2, 3, 4, 8, 9, and 13 protein sequences contain two conserved MYB domains. The conserved domain of ThMYB12 exists at the end of the peptide chain. The conserved domains of ThMYB5, 6, 7, 10, and 11 were located in the middle of the peptide chain. ThMYB1 and ThMYB14 contain a conserved domain of MYB-type in the N-terminus.

Further analysis of the sequences of these conserved domains revealed that ThMYB2, 3, 4, 8, 9, and 13 all contained the R2R3 conserved domain (Supplementary Figure S2). It was suggested that these 6 ThMYB genes belong to R2R3-MYB family, in which the first tryptophan in the R3 domain of ThMYB8 is replaced by leucine (L) and the first tryptophan in the R3 domain of the other five proteins is replaced by phenylalanine (F). In addition, the peptide chains of ThMYB5, 6, 7, 10, and 11 contain a highly conserved domain of MYB-type HTH DNA in the middle (Supplementary Figure S3).

A NJ phylogenetic tree of 14 T. hispida ThMYB proteins and 125 Arabidopsis MYB proteins were constructed (Figure 1). The phylogenetic distribution indicated that MYB proteins can be classified into 13 groups, while the 14 ThMYB TFs were divided into 5 subgroups: class II, IV, V, VI, and X. Among them, class X was the largest subgroup and contained 8 ThMYB proteins, ThMYB1, 5, 6, 7, 10, 11, 12, and 14. All of these proteins contained only one conserved domain. The other ThMYB proteins (including ThMYB 2, 3, 4, 8, 9, and 13) that contained two conserved MYB domains were clustered into four individual clades.

To study the expression patterns of these 14 ThMYB genes in T. hispida leaves and roots without stress, their expressions were measured by qRT-PCR. The lowest transcript level genes (i.e., the highest delta Ct value) were set as 1 to normalize the transcript levels of the other ThMYBs (Figure 2). The relative expression levels of the 14 ThMYBs exhibited marked differences in leaves and roots. Interestingly, ThMYB9 was the least abundant in both the leaves and roots. However, ThMYB6 was the most abundant gene in leaves, and ThMYB12 was the most abundant gene in roots. The abundance of ThMYB6 was 2939 times higher than the lowest abundance (ThMYB9) in leaves, and the abundance of ThMYB12 was 2391 times greater than ThMYB9 in the roots, indicating that ThMYB6 in the leaves and ThMYB12 in the roots might play important functions than the other ThMYBs.

To analyze the stress response function of ThMYBs, the expression profiles of 14 ThMYBs under different stress, including high salt, osmotic stress, and hormones treatments (ABA, GA3, JA) were analyzed by qRT-PCR.

Under NaCl stress, most of ThMYB genes had upregulated expression in the leaves. Especially, ThMYB1, 3, 4, 11, 13 and 14 had all induced expression at all stress points. The expression of ThMYB9 was significantly upregulated at most stress times (6, 24, and 48 h). The relative expression levels of ThMYB5 and ThMYB12 in leaves were upregulated at 12–72 h but downregulated rapidly at 6 h after salt stress. In contrast, ThMYB6 and ThMYB10 were induced at 6 h and inhibited with the other studied stress times. In the roots, the expression patterns of ThMYB1, 3, 4, 11, 13, and 14 were the same as those in leaves, and they were upregulated at all stress points. However, ThMYB6, 10 and ThMYB12 were significantly different from those in the leaves. The expressions of ThMYB6 and 10 were mainly upregulated at all stress points, and ThMYB12 was downregulated at all studied time points. In addition, the transcript of ThMYB8 was significantly downregulated at 6 h, and the lowest expression was only 1.65% of the control (Figure 3A).

Under PEG stress, the expressions of ThMYB1, 3, 13, and 14 were significantly upregulated at all stress points in the leaves. In addition, the most induced gene was ThMYB9, in which the peak expression level was 14,563-fold (6 h) that of the control. The relative expressions of ThMYB5 and ThMYB12 in leaves were also upregulated after stress at 12 h. In contrast, some ThMYB genes showed downregulated expression after PEG stress. Especially, the expression of ThMYB2 was significantly downregulated at most stress times (except 48 h), which was only 1.24% of the control at PEG stress at 6 h. Similarly, ThMYB6 and ThMYB10 were downregulated with most of the studied stress times (except 6 h). In the roots, almost all of the genes were significantly upregulated at all stress points, except for ThMYB5, 7, 8, and 12. Especially, with the prolonged stress time, the relative expression level of ThMYB1 gradually increased. The expressions of ThMYB5, 8, and 12 were significantly downregulated at 6 h. However, at 12–72 h, the relative expressions of ThMYB5 and 8 were upregulated, while the expression of ThMYB12 did not change significantly (Figure 3B).

Under ABA treatment, the transcripts of ThMYB11 were not significantly differentially regulated at all stress points in the leaves. However, most the ThMYB genes expressed were significantly different after ABA stress. Especially, ThMYB1, 3, and 14 showed significantly upregulated expression at all stress points, while ThMYB2, 7, and 8 were downregulated at all studied stress times. The relative expression of ThMYB5 and ThMYB12 in leaves was upregulated at 12-72 h but downregulated rapidly at 6 h after ABA stress. In contrast, ThMYB6 and ThMYB10 were induced at 6 h and inhibited with the other studied stress times. In the roots, the transcripts of all ThMYB were classified into three groups. One group was included by ThMYB1, 3, 4, 6, 9, 10, 13, and 14 and was clearly upregulated at all stress points. The second ThMYB group composed of ThMYB2, 5, 7, 8, and 11 was upregulated at 12–72 h but downregulated at 6 h after treatment of ABA. The expression patterns of ThMYB12 were different from those of the two groups, which was down regulated at all stress time points (Figure 4A).

Under GA3 treatment, most of ThMYB genes had downregulated expression after GA3 stress in the leaves. Especially, ThMYB2, 4, 11, and 13 were downregulated at all stress points. In addition, the relative expression of ThMYB6, 9 and 10 was significantly downregulated at most stress times (12–72 h). In contrast, ThMYB3 and ThMYB12 were mainly upregulated at most of the studied stress times. In the roots, most of the ThMYB genes expressed were significantly inhibited at 6 h and induced at the other studied stress time points. ThMYB10 was significantly different than the other ThMYBs, which was significantly upregulated at the early stress times (6–48 h) and downregulated rapidly at 72 h after GA3 treatment (Figure 4B).

Under JA stress, except ThMYB1, 9, and 10, almost all of the ThMYB genes transcribed showed downregulation at all stress points in the leaves. In addition, ThMYB9 was highly upregulated at all of the studied stress times. The relative expression reached its peak level at 72 h (17.8-fold higher than the control). Interestingly, most of the ThMYBs expressed (except ThMYB9 and 14) showed distinct expression patterns between the leaves and roots. In the roots, except for ThMYB1, ThMYB10 and ThMYB14, the other genes showed significant upregulation at all stress points. However, ThMYB10 was significantly downregulated at all stress points (Figure 4C).

In general, the expression of 14 ThMYBs in leaf and root tissues were changed at least at one stress time point, which indicated that these ThMYBs might be one or several stress response genes, which might be involved in the stress tolerance of T. hispida. Especially, ThMYB2, 6, and 7 were mainly downregulated in leaves and upregulated in roots under five stress conditions. However, ThMYB1, 3, 13, and 14 genes were mainly upregulated in both leaves and roots in response to NaCl, ABA and PEG treatment conditions, suggesting that they might play an important role in stress tolerance. Therefore, the ThMYB13 gene was selected for further study of stress resistance.

To determine whether transient overexpression and suppression of the ThMYB13 gene in T. hispida were successful, the expression of ThMYB13 in Con (transformed with empty pROKII), OE and RNAi plants were investigated by qRT-PCR. The transcripts of ThMYB13 in OE or RNAi plants was normalized by the expression of Con plants at 0 h. The results suggested that the transcript of ThMYB13 in the OE plants was the highest among the three types of transiently transformed seedlings. At 36 h, the expression of ThMYB13 in OE was 33.94 times that of Con. While the expression of ThMYB13 in RNAi plants was significantly lower than those in Con, it was only 5.37% of control at 36 h (Figure 5). These results indicated that we successfully obtained transient overexpression and inhibition of ThMYB13 plants.

To preliminarily identify the function of ThMYB13 gene, we analyzed and compared the biochemical staining and related physiological indexes of three kinds of transient transformed T. hispida. DAB and NBT in situ staining were carried out to study H₂O₂ and O^2- accumulation, respectively. In the Con, OE and RNAi plants. Oxygen ions released by H₂O₂ in cells can oxidize DAB to form brown precipitates, and the amount of H₂O₂ released from cells can be displayed according to the depth of staining. Similarly, NBT can be oxidized by superoxide ion O^2- to blue formazan, which can reflect the content of O^2- through the blue shade of formazan. The results of both NBT and DAB staining showed that the OE plants showed greatly reduced levels while RNAi plants accumulated high levels compared with Con plants (Figures 6A,B). The content of H₂O₂ was further determined. The results showed that the content of H₂O₂ in the three types of plants was significantly different after 24 h of salt stress, and the level of H₂O₂ in RNAi plants was the highest, which was 1.22 times that of Con plants. While H₂O₂ content in the OE plants was the lowest, it was only 82.41% of the content of Con plants (Figure 6D). These results indicated that in the OE plants, ROS was highly reduced, whereas in RNAi plants, ROS was accumulated highly under salt conditions. Increasing the expression of ThMYB13 can improve ROS clearance.

To monitor the level of cell death, electrolyte leakage and evans blue staining were analyzed. Evans blue in situ staining indicated that in OE plants, cell death was reduced, but it was increased in that of RNAi plants when compared with Con plants exposed to salt stress (Figure 6C). Electrolyte leakage results confirmed these results. The relative electrical conductivities of RNAi plants were the highest at 24 h, 1.26 times that of Con plants, meanwhile those of OE were 0.78 folds those of Con plants (Figure 6E). At the same time, there was no difference among the three types of plants in MDA level under normal conditions. Under salt conditions, OE plants showed significantly lower MDA level than the Con plants, and the content of MDA in RNAi plants was 1.32 folds higher than MDA in the Con plants at 24 h after salt stress (Figure 6F). These results indicated that membrane lipid peroxidation was highly increased in RNAi plants, but was decreased in OE plants.

We further determined the contents of K⁺ and Na⁺ in transformed plant roots and leaves. Under normal conditions, the sodium and potassium ion contents in the leaves and roots of three types of plants were not significantly different. Under salt conditions, all three types of plants showed higher Na⁺ content in roots than leaves. However, the RNAi plants accumulated higher Na⁺ content than the OE and Con plants, and the OE plants showed significantly lower accumulation of Na⁺ than Con plants in leaves and roots (Figures 7A,B). Conversely, all three types of plants showed lower K⁺ content in roots than leaves under salt stress conditions, the RNAi plants still had the lowest K⁺ content, whereas the OE plants showed the highest K⁺ contents in the leaves or roots, followed by that in the Con plants (Figures 7C,D). Consistently, all three types of plants displayed a higher K⁺/Na⁺ ratio in leaves than roots. At the same time, the OE plants showed the highest K⁺/Na⁺ ratio in both roots and leaves, followed by the Con plants, with the RNAi plants having the lowest K⁺/Na⁺ ratio. After salt stress for 24 h, the ratio of K⁺/Na⁺ in OE plants in leaves and roots was 1.48 and 2.13 times higher than that in Con plants, while the K⁺/Na⁺ ratio of RNAi plants was only 45.67 and 41.81% of Con plants (Figures 7E,F).