Introduction

Salt stress is one of the major environmental stresses. Soils affected by salt stress are nearly 10% of the land surface (950 Mha) and 50% of all irrigated land (230 Mha) in the world (Ruan et al. 2010; Song et al. 2016; Wang et al. 2015a), which affect plant growth and crop production worldwide. The cultivation of salt-tolerant crops has been considered as the most economical and effective way to improve saline-alkali land. Therefore, revealing the mechanism of salt tolerance is necessary to improve the salt tolerance of plants.

Plant hormones regulate almost every aspect the growth and development of plants (Peleg and Blumwald 2011). Additionally, hormones have been reported to play very important roles in salt stress responses and adaptation (Javid et al. 2011a). Among all known hormones, abscisic acid (ABA), jasmonate (JA), ethylene (ET), and cytokinin (CK) have been known to play major roles in mediating plant defense response against abiotic stresses (Peleg and Blumwald 2011; Gamalero and Glick 2012; Nakashima 2013). ABA has been regarded as a key regulator in the activation of plant cellular adaptation to drought and salt stress. The level of ABA increases because of enhancing ABA biosynthesis and inhibiting ABA catabolism, which is one of the fastest responses of plants to abiotic stress. The increased ABA binds to its receptor to initiate signal transduction leading to cellular responses to stresses (Ng et al. 2014; Sah et al. 2016). The induction of downstream genes of ABA can cause stomatal closure, thereby reducing water loss via transpiration and eventually restricting cellular growth (Raghavendra et al. 2010; Wilkinson and Davies 2010). Additionally, pathways closely related to stress response such as proline accumulation and calcium signaling pathways can also be mediated by ABA signaling pathway. These can enable plants to survive severe adverse environmental conditions such as salt and drought stresses. Increased level of endogenous ABA has been observed in response to abiotic stress in many plant species such as sorghum (Kannangara et al. 1983), maize (Wang et al. 2008), wheat (Guóth et al. 2009), and alfalfa (Luo et al. 1992).Other hormones also play direct or indirect roles in the response of plants to salt stress. JA is considered to play a positive role in salt stress responses. It has been reported that both drought and high salinity stresses resulted in increase of jasmonate levels in the leaves and roots and induction of JA biosynthesis genes (Moons et al. 1997; Wang et al. 2001; Tani et al. 2008; Du et al. 2013). Expression profiling and physiological characterization on salinity stress response in barley also revealed the JA-mediated adaptation to salinity stress (Walia et al. 2007). Additionally, exogenous JA application significantly reduces Na+ content in salt-tolerant rice and wheat (Kang et al. 2005; Qiu et al. 2014). Moreover, JA treatment could recover some of the salt-induced defects (Jiyoung et al. 2009; Javid et al. 2011a, b). Several evidences have indicated the involvement of auxin in response to salt stress in plants (Wang et al. 2009; Jung and Park 2011; Lee et al. 2012; Javid et al. 2011b). It has been reported that exogenous indole-3-acetic acid (IAA) application significantly reduced the salinity-induced loss (Husen et al. 2016). CKs are also involved in the stress tolerance regulation of plants. Evidence indicates that CKs play both positive and negative roles on stress tolerance (Ha et al. 2012; Wang et al. 2015b; Zwack and Rashotte 2015). For gibberellins (GAs), it has been shown that the reduction of GA levels and signaling could contribute to plant growth restriction on exposure to several stresses, including cold, salt, and osmotic stress (Colebrook et al. 2014). Brassinosteroid (BR) has also been proven to take part in plant response to salt stress (Krishna 2003).

Sweet sorghum [Sorghum bicolor (L.) Moench] is a C4 crop that has been considered as a good source for ethanol production and forage crop (Almodares and Hadi 2009). Additionally, sweet sorghum has high tolerance to salt stress and can complete its life cycle on saline land. According to our physiological results, we found that endogenous hormone concentrations change significantly in sweet sorghum under salt stress. Although several mechanisms have been proposed to contribute to the salt tolerance of sweet sorghum (Chai et al. 2010; Oliveira et al. 2011; Dai et al. 2014), the role of hormones in the salt tolerance of sweet sorghum has not yet been revealed. In order to investigate the mechanism of hormone regulation under salt stress, a transcriptome analysis was performed on two sweet sorghum inbred lines (salt-tolerant M-81E and salt-sensitive Roma). The results of the present study would provide further insight into the complex regulatory network of hormones underlying the mechanism of high tolerance to salt stress in sweet sorghum.

Materials and Methods

Plant Materials and Growth Conditions

Seeds of two sweet sorghum inbred lines (M-81E and Roma) were selected as the experimental materials. M-81E was regarded as salt-tolerant inbred line and Roma was regarded as salt-sensitive inbred line according to a previous study (Sui et al. 2015).

Plants were grown in pots, as previously described (Sui et al. 2015). When they grew to the four-leaf stage, salt treatment was performed. In a previous study, 150 mM was regarded as the proper concentration of NaCl, which can induce a significant difference in physiological parameters between M-81E and Roma (Sui et al. 2015). The treated plants were irrigated with nutrient solution supplemented with 0 and 150 mM NaCl. The NaCl concentration was increased stepwise towards a final concentration of 50 mM each day.

Measurement of Physiological Parameters

After exposure to salt treatments for 48 h, five replicates of each treatment were sampled to determine the root and leaf lengths. Then, the roots and leaves were separated, and their fresh weight (FW) was directly determined. For dry weight (DW) determination, the samples were weighed after being dried at 150 °C for 15 min and 70 °C for 72 h. Water content was defined as follows: WC (%) = [(FW − DW)/FW × 100].

Determination of Endogenous Hormone Concentration

IAA, free zeatin riboside (ZR), ABA, and GAs (including GA1 and GA3) contents were analyzed using the indirect ELISA technique. In brief, the samples (0.2 g of fresh leaves or roots of each genotype with or without salt treatment) were homogenized in liquid nitrogen and extracted in cold 80% (v/v) methanol with butylated hydroxytoluene (1 mM) overnight at 4 °C. These extracts were collected after centrifugation at 10,000g (4 °C) for 20 min, the extracts were passed through a C 18 Sep-Pak cartridge (Waters, Milford, MA), and the extracts were dried in N2. Then, the residues were dissolved in PBS (0.01 M, pH 7.4) in order to determine the IAA, GA, ABA, and ZR levels according to Yang’s method (Yang et al. 2001). Three replicates were used for each treatment.

Total RNA Extraction

Total RNA was isolated from the leaves and seminal roots of each genotype treated with 0 and 150 mM NaCl for 48 h. Total RNA isolation was performed using a Total Plant RNA Extraction Kit (Karroten, Beijing, China) following the manufacturer’s protocols. The RNA was quantified using a nanodrop-ND-1000 spectrophotometer (Thermo Fisher Scientific, Wilmington, DE, USA). A 1% agarose gel buffered by Tris-acetate-EDTA was also run to determine the integrity of the RNA.

Library Construction and Illumina Sequencing

Libraries were constructed following a high throughput Illumina Strand-Specific RNA Sequencing Library protocol (Zhong et al. 2011). The final cDNA library was sequenced using an Illumina HiSeq TM 2500 (BioMarker Technologies Co Ltd., Beijing). The RNA-seq data of the control and salt-treated samples were obtained from two and three biological replicates, respectively.

Mapping and Detection of Differentially Expressed Genes

Clean reads were mapped to the sorghum genome using TopHat version 2.0.10. Mapping results generated by TopHat were filtered to retain only unique mapped reads before being piped into Cuffdiff (http://cufflinks.cbcb.umd.edu/index.html) to estimate read counts for each gene. Differential expression analysis was performed using the DESeq by the counts. A RPKM threshold value of 0.1 was set to detect the presence of a transcript for a particular gene. Differentially expressed genes (DEGs) were defined using DESeq (Anders and Huber 2010) as fold changes ≥ 2 with a false discovery rate (FDR) adjusted P value ≤ 0.01.

Gene Annotation and Classifications

The optimal assembly results were chosen according to the assembly evaluation. The assembled sequences were compared against the NCBI non-redundant (nr) database, Swiss-Prot, GO, COG, and KEGG database using BLAST as previously described (Sui et al. 2015) with E value ≦ 1e−10 as the cutoff.

Quantitative Real-Time PCR Analysis

Ten DEGs which annotated to pathways related to hormone biosynthesis and signal transduction were selected for quantitative real-time PCR (qRT-PCR) to verify the RNA-Seq results. Primers for these 10 genes were designed using the Beacon Designer software (version 7.0) and are shown in Table S1. S. bicolor’s housekeeping gene β-actin (GenBank ID: X79378) was used as an internal standard. Measurement and calculations were performed as previously described (Sui et al. 2015).

Statistical Analysis

The significance of the data was analyzed by ANOVA followed by Tukey’s posthoc test. All tests were performed using SPSS Version 16.0 for Windows (SPSS, Chicago, IL, USA).

Results

Effects of Salt Stress on Growth Parameters

After treated with 150 mM NaCl, the growth of both M-81E and Roma was inhibited. However, according to the statistical analysis, changes in all growth parameters in M-81E and Roma did not reach a significant level. Under salt stress, the leaf length and root length of M-81E decreased by 6.56 and 11.57%, respectively, while, the leaf length and root length of Roma decreased by 12.23 and 11.14%, respectively (Fig. S1). The FW and DW of leaves decreased in both genotypes after treated with NaCl. FW and DW of leaves in M-81E decreased by 23.52 and 11.2%, respectively. In Roma, the FW and DW of leaves decreased by 28.25 and 17.82%, respectively (Fig. 1). In M-81E, the FW of roots increased by 1.31% under salt stress. In Roma, the FW of roots deceased by 30.15%. The DW of M-81E increased slightly by 0.32% under salt stress. In contrast to M-81E, the DW of Roma had a 19.44% decrease (Fig. 2). There were no significant differences in water content between M-81E and Roma (Figs. 1 and 2).

Fig. 1
figure 1

Fresh weight (a), dry weight (b), and water content (c) in leaves of M-81E and Roma treated with different concentrations of NaCl (0 and 150 mM) for 48 h. Values are means ± SD of five replicates

Full size image
Fig. 2
figure 2

Fresh weight (a), dry weight (b), and water content (c) in roots of M-81E and Roma treated with different concentrations of NaCl (0 and 150 mM) for 48 h. Values are means ± SD of five replicates

Full size image

Effects of Salt Stress on Hormone Concentration

The hormone (ABA, IAA, ZR, GA, BR, and JA) concentrations measured in the leaves and roots of M-81E and Roma under controlled conditions and salt stress are shown in Fig. 3. ABA concentration significantly increased in leaves of M-81E by 49.02%. Although it was not significant, ABA concentrations also slightly increased (12.81%) in leaves of Roma. In roots, the concentration of ABA increased in M-81E by 5.51%, while this increase was not demonstrated in Roma (Fig. 3a). After treated with NaCl, the concentration of JA had no significant changes in both leaves and roots of M-81E. However, the concentration of JA decreased by 20.93% in leaves and increased by 40.71% in roots of Roma (Fig. 3b). ZR concentrations decreased by 11.37 and 10.04% in leaves and roots in M-81E, respectively. However, in Roma, ZR concentrations did not exhibit any significant changes in leaves, but exhibited a 16.05% decrease in roots (Fig. 3c). Moreover, IAA contents were significantly induced by salt stress in leaves of both M-81E by 42.8%. Though not significant, the IAA concentration increases by 23.62% in Roma leaves as well. The increase of IAA content in roots was not as significant as that in the leaves of both genotypes (Fig. 3d). To our surprise, concentrations of BR and GA did not exhibit a significant change in both M-81E and Roma under salt stress (Fig. 3e, f).

Fig. 3
figure 3

Concentration of ABA (a), JA (b), IAA (c), ZR (d), BR (e), and GA (f) in leaves and root of M-81E and Roma treated with different concentrations of NaCl (0 and 150 mM) for 48 h. ML and MR stand for leaves and roots of M-81E, respectively. RL and RR stand for leaves and roots of Roma, respectively. Values are means ± SD of three replicates. Asterisks indicate significant differences between control plants and those subjected to salt stress (∗ P ≤ 0.05)

Full size image

In order to investigate the molecular mechanisms of hormone regulation under salt stress, RNA-seq analysis was conducted, using libraries created from leaves and roots of M-81E and Roma under normal and NaCl treatment, respectively.

Verification of RNA-Seq Data

As shown in Fig. 4, a high correlation (R 2 = 0.91) between RNA-seq and qRT-PCR was observed, confirming the reliability of the RNA-seq data.

Fig. 4
figure 4

Validation of RNA-seq results by RT-qPCR. R 2 represents the correlation coefficient value between the two platforms. The numbers in the scale bar stand for RPKM values in RNA-seq and ΔΔCt in qRT-PCR, which were used to evaluate the correlation (R 2). Primers and annotation of the 10 DEGs are listed in Supplementary Table S1

Full size image

DEGs Related to Hormone Biosynthesis

According to our RNA-seq data, in M-81E, 864 and 2085 genes were differentially expressed between control plants and those subjected to salt in leaves and roots, respectively. In Roma, 930 and 3172 genes were differentially expressed between control plants and those subjected to salt in leaves and roots, respectively. DEGs related to hormones biosynthesis and signal transduction were selected for further analysis.

DEGs Related to ABA Biosynthesis

According to our RNA-seq data, after salt treatment, the expression of Sb01g046730, which encodes zeaxanthin epoxidase (ZEP), was induced by 6.88- and 6.24-fold in the roots of M-81E and Roma, respectively. The expression of the 9-cis-epoxycarotenoid dioxygenase (NCED)-encoding gene Sb02g003230 increased in the roots of both M-81E and Roma under salt stress. These results indicate that ABA biosynthesis may be enhanced by salt stress in the roots of both genotypes. In leaves, Sb04g030640 which encodes carotenoid cleavage dioxygenase 4 was differentially expressed in both genotypes under salt stress. This gene was upregulated in M-81E, but was downregulated in Roma. ABA catabolism is the major route of ABA degradation. ABA 8′-hydroxylase (ABA 8′-OH) degrades active ABA into 8′-hydroxy ABA, and 8′-hydroxy ABA is subsequently converted to phaseic acid by spontaneous isomerization (Kushiro et al. 2004). After treated with NaCl for 48 h, ABA 8′-OH-encoding gene Sb04g030660 was significantly upregulated in leaves and roots of Roma by 2.66- and 8.19-fold, respectively. However, no ABA 8′-OH-encoding gene was differentially expressed in M-81E (Fig. 5, Table S2).

Fig. 5
figure 5

Heat map display of DEGs mapped to KEGG pathways related with hormone biosynthesis. The numbers in the scale bar show the log2 Fold change of each DEG.

Full size image

DEGs Related to JA Biosynthesis

JA and its methyl ester, methyl jasmonates (MeJAs), are derivatives of fatty acid metabolism. After treatment with NaCl for 48 h, Sb05g002750 which encodes phospholipase A1 (PLA1) was induced only in the roots of Roma. Moreover, 4-Cl-like enzyme-encoding gene Sb01g048200 was also upregulated in leaves of Roma. DEGs encoding Acyl-coenzyme A oxidase (ACXs), which participated in fatty acid metabolism, as well as the signaling and generation of polyunsaturated fatty acids, including the precursors of JA, were all downregulated in the leaves and roots of M-81E (Fig. 5, Table S2).

DEGs Related to IAA Biosynthesis

After treated with NaCl for 48 h, many IAA biosynthesis-related genes were differentially expressed in both genotypes. In M-81E, there was one DEG-encoding aldehyde dehydrogenase (ADL) downregulated in leaves and roots (Sb04g033420 and Sb04g029040, respectively). Moreover, indole-3-acetaldehyde oxidase-encoding gene Sb01g005650 was upregulated in leaves of M-81E. For Roma, four DEGs related to IAA biosynthesis were found in the roots. Among these DEGs, two ADL-encoding genes (Sb02g025790 and Sb05g005470) and one NIT4 (Sb04g026940)-encoding gene were downregulated. Sb06g001210 which encodes indole-3-pyruvate monooxygenase was induced by salt stress in the roots of Roma (Fig. 5, Table S2).

DEGs Related to Zeatin, GA, and BR Biosynthesis

CK dehydrogenase (CKX), which can catalyze the irreversible inactivation of cytokinins, is the key enzyme of cytokinin degradation (Frébort et al. 2011). After treated with NaCl for 48 h, DEG-encoding CKX 5 was downregulated in the leaves of M-81E, while, CKX 4-encoding gene Sb03g045410 was upregulated in the roots of M-81E. In the roots of Roma, two DEGs (Sb03g036160 and Sb03g045410) which encode CKXs were induced by salt stress (Fig. 5, Table S2).

According to our RNA-seq data, CYP88A1-encoding gene Sb10g000920 was induced by salt stress in the roots of M-81E. While, there was no DEG related to GA biosynthesis in Roma. Sb06g019600 which encodes CYP 724B1 was downregulated in both leaves and roots of M-81E. Additionally, CYP 90A1-encoding gene Sb05g002580 was downregulated only in the roots of M-81E. In Roma, CYP 90B1-encoding gene Sb01g041900 was downregulated in the roots (Fig. 5, Table S2).

DEGs Related to Hormone Signal Transduction

The KEGG pathway of the plant hormone signal transduction shows the transduction network of different hormones (Fig. S2).

ABA Signal Transduction

Many proteins such as pyrabactin resistance (PYR)/PYR-like (PYL)/regulatory component, protein phosphatase 2C (PP2C), and Snf1-related kinase (SnRK2)-type kinases have been proven to be involved in the ABA signal transduction pathway (Ma and Grill 2009; Park et al. 2009; Santiago et al. 2009a; Santiago et al. 2009b). According to our RNA-seq data, PYL8-encoding gene Sb09g006700 was downregulated only in the roots of Roma. PP2Cs play important roles in ABA signal transduction as well as in other signaling pathways. In the leaves of M-81E, Sb02g022090 which encodes PP2C 68 was downregulated by 3.69-fold. However, DEG (Sb09g022410) in leaves of Roma was upregulated. Two and three DEGs encoding PP2C were identified in the roots of M-81E and Roma, respectively. Among these DEGs, Sb09g029080 was significantly downregulated in both genotypes, particularly in M-81E, by 73.28-fold. In addition, one and two DEGs were upregulated in M-81E and Roma, respectively. It has been reported that all members of the rice SnRK2 family designated SAPKs that could phosphorylate TRAB1 directly to activate transcription in response to ABA (Kobayashi et al. 2005). In M-81E, only one gene (Sb08g019700) encoding SAPKs was differentially expressed in the roots after treated with NaCl. While, in Roma, there were one and three DEGs encoding SAPKs identified in the leaves and roots, respectively (Fig. 6, Table S3). DEGs encoding TRAB1 were only found in the roots. Furthermore, Sb10g007090 was upregulated by 2.86-fold in M-81E and Sb04g034190 was upregulated by 2.49-fold in Roma. Moreover, ABA insensitive 5 (ABI5)-like protein 2-encoding gene Sb09g023920 was downregulated only in the roots of Roma.

Fig. 6
figure 6

Heat map display of DEGs mapped to KEGG pathways related with hormone signal transduction. The numbers in the scale bar show the log2 fold change of each DEG.

Full size image

JA Signal Transduction

Generally, it has been reported that COI1 required for almost all JA-dependent responses (Fonseca et al. 2009). The Jasmonate ZIM (JAZ) domain is a repressor of JA-induced transcriptional activity (Thines 2007). According to our RNA-seq data, four genes encoding TIFYs were differentially expressed under salt stress. In M-81E, Sb01g023431 and Sb02g025720 which encode TIFY 3A and TIFY 10A were both downregulated only in leaves. While, in Roma, TIFY 10A-encoding gene Sb01g027325 was downregulated in leaves and Sb01g023300 which encodes TIFY 3A was significantly upregulated in the roots (Fig. 6, Table S3).

IAA Signal Transduction

A number of DEGs related to IAA signal transduction were determined in both genotypes, particularly in roots. Two and four auxin transporter-like protein-encoding genes exhibited differential expression in the roots of M-81E and Roma, respectively, and all of them were upregulated. Sb10g026970 which encodes vegetative cell wall protein gp1 was downregulated in the leaves of M-81E and upregulated in roots of Roma. Indole-3-acetic acid-induced protein ARG7-encoding gene was downregulated in the leaves and upregulated in the roots of Roma. Additionally, four DEGs encoding indole-3-acetic acid-amido synthetase, which could inactivate auxin, were found only in roots of Roma and all of them were upregulated (Fig. 6, Table S3).

ZR, GA, and BR Signal Transduction

Merely one gene (Sb10g005580) related to ZR transduction differentially expressed in M-81E under salt stress. However, one (Sb04g022780) and four (Sb04g035980, Sb04g038320, Sb08g001140, and Sb02g010680) DEGs related to ZR transduction were obtained in the leaves and roots of Roma, respectively (Fig. 6, Table S3).

Sb03g032990, which encode the receptor of BR signaling SR160, was downregultated in the leaves of M-81E. Sb10g028580 encoding xyloglucan endo-transglucosylase/hydrolase was significantly downregulated in both leaves and roots of M-81E by a fold greater than 11. In the roots of Roma, Sb10g028580 was also downregulated by 3.48-fold. Additionally, serine/threonine-protein kinase-encoding gene Sb01g048160 was upregulated only in the roots of Roma after salt treatment for 48 h (Fig. 6, Table S3).

Furthermore, no DEGs were found to be related to the GA signal transduction pathway in the RNA-seq data obtained in the present study (Fig. 6, Table S3).

Discussion

Phytohormones are the key regulators of plant growth and development. Moreover, phytohormones play important roles in mediating responses to various abiotic stresses in plants (Sreenivasulu et al. 2012).

The gene expression may response to salt treatment quickly, but the phenotypes of plants may appear later. As a result, though the expression levels of genes which related to phytohormones changed significantly, the changes of growth parameters in both M-81E and Roma did not reach a significant level after 48-h salt treatment (Figs. 1 and 2). Additionally, as previously described (Sui et al. 2015), after treated with NaCl for 7 days, significant changes in growth parameters of both M-81E and Roma were obtained. This may be related to early hormonal regulation.

ABA is regarded as a key regulator in the activation of plant cellular adaptation to drought and salt stress. Two pathways have been suggested for ABA biosynthesis. One is the direct route, in which ABA is derived from the C15 compound farnesyl pyrophosphate. The other one is the indirect route, in which ABA is derived from a C40 carotenoid. The indirect pathway has been proven to be the main pathway (Zeevaart and Creelman 1988; Zeevaart 1999; Seo and Koshiba 2002). The first biosynthetic step in ABA biosynthesis is the conversion of zeaxanthin to all-trans-violaxanthin, which is catalyzed by ZEP. Moreover, NCED catalyzes the oxidative cleavage of 9-cis-violaxanthin and/or 9-cis-neoxanthin to produce xanthoxin. This reaction is considered to be the rate-limiting step in ABA biosynthesis (Chernys and Zeevaart 2000; Iuchi et al. 2001; Qin and Zeevaart 2002; Thompson et al. 2007). In the present study, Sb04g030640, which encodes carotenoid cleavage dioxygenase 4, was differentially expressed in leaves. It was upregulated in M-81E and downregulated in Roma. Additionally, the encoding gene of ABA 8′-OH, which participates in ABA degradation, was upregulated in Roma leaves. While, the expression level of it did not change in leaves of M-81E. Compared to the leaves of both genotypes, the encoding genes of ZEP and NCED were significantly enhanced in the roots. ABA 8′-OH-encoding gene Sb04g030660 was significantly increased only in the roots of Roma. These results indicated that in both leaves and roots, ABA biosynthesis was enhanced in M-81E, particularly in the roots. In Roma, the biosynthesis and degradation of ABA were both enhanced by salt stress. To our surprise, in both genotypes, ABA concentration increased only in leaves after treated with NaCl for 48 h, particularly in salt-tolerant genotype M-81E. However, it did not exhibit any change in the roots of both genotypes. Many earlier studies have shown that leaves can accumulate ABA in response to salt stress, particularly in salt-tolerant strains (Kefu et al. 1991; He and Cramer 1996; Zörb et al. 2013). Since the biosynthesis of ABA was more significantly enhanced in the roots and the concentration of ABA was increased only in leaves, we speculated that the redistribution of ABA from roots to leaves may be induced by salt stress in sweet sorghum.

ABA has been proposed to act as an endogenous messenger in regulating the plant’s water status in plant response to a range of stresses. Additionally, many components of the ABA signaling pathway have been identified. According to our RNA-seq data, many DEGs were annotated to the ABA signal transduction pathway after treated with 150 mM NaCl for 48 h. PP2Cs have been proven as negative regulators of ABA signaling. Additional PP2Cs also function in different signaling pathways (Javid et al., 2011a, b). Moreover, PP2Cs are also associated with the expression of ABA-induced genes. Overexpression of PP2Cs inhibits the transcription of ABA-induced genes in maize (Javid et al. 2011a, b; Peleg and Blumwald 2011). In both leaves and roots of Roma, majority of DEGs encoding PP2Cs were upregulated. However, for M-81E, DEGs encoding PP2Cs were mainly downregulated, particularly in the roots, with Sb09g029080 downregulated by 73.3-fold. PYR/PYLs can bind to ABA and inactivate the activity of PP2Cs. According to our RNA-seq data, PYL8-encoding gene Sb09g006700 was only downregulated in the roots of Roma. It has been reported that ABA has an impact on PYL expression level (Wang et al. 2013). Additionally, JA signaling involves PYL4 and MeJA elicitation can modulate NtPYL4 transcript levels (Lackman et al. 2011). In the present study, no change in ABA content was observed, while JA content significantly increased in roots of Roma. The increase in JA content might be related to the differential expression of PYL8. The changes in PYL8 indicated that the inhibition of PP2C activity may be weakened by salt stress in Roma. As a result, Roma may obtain a stronger negative regulation capacity of ABA signal. Additionally, as previously reported, mutants of PYLs exhibited ABA insensitivity, causing the ABA-inducing gene expression to be compromised (Park et al. 2009; Nishimura et al. 2010). The downregulation of PYL 8 in Roma may have effects on ABA-inducing gene expression. SnRK2-type kinases are regarded as positive regulators of ABA signaling and water stress (Kobayashi et al. 2006; Fujita et al. 2009; Nakashima et al. 2009; Nakashima 2013). In rice, 10 members of the SnRK2 family (SAPK) were found. ABA activates these SAPK kinases and phosphorylates TRAB1, which is a rice ABF transcription factor. In the present study, DEGs encoding SAPKs were mainly downregulated, while DEGs encoding TRAB1 were upregulated in the roots of both genotypes. These results were consistent with those of previous studies that expression of homologs of TRAB1 can be induced by abiotic stress (Casaretto and Ho 2003; Xue and Loveridge 2004). All these may contribute to the difference in salt tolerance between M-81E and Roma.

JA and MeJAs have been considered to play an active role in regulating the biotic and abiotic stress tolerance of plants (Küpper et al. 2009; Koo et al. 2009; Wasternack and Hause 2013). Moreover, the content of JA and related compounds changes in response to salt stress (Pedranzani et al. 2003). JAs and MeJA are produced from α-linolinic acid. It has been proven that PLA1 releases α-LeA from the sn1 position of galactolipids (Wasternack 2007). Enzymes of fatty acid β-oxidation, such as ACX, have also been shown to play roles in JA biosynthesis (Castillo and León 2004; Li et al. 2005). Moreover, 4-coumarate-CoA ligase-like (4-Cl-like) enzymes are also able to activate downstream intermediates in the JA biosynthesis (Schneider et al. 2005; Koo et al. 2006). In the present study, JA content in M-81E did not reveal any significant change in the leaves, and decreased slightly in the roots. While in Roma, JA content slightly decreased in the leaves and significantly increased in the roots. These results were consistent with the study conducted by Pedranzani, in which JA content remained almost steady in salt-tolerant tomato cultivar Pera and increased in salt-sensitive HF, particularly in the roots (Pedranzani et al. 2003). DEGs related to JA biosynthesis were all downregulated in both the leaves and roots of M-81E. Sb05g023970 which encodes ACX2 was downregulated by 8.62-fold, which may contribute to the decrease in JA content in the roots of M-81E. However, Sb01g048200 (4-Cl-like-encoding gene) and Sb05g002750 (PLA1-encoding gene), which play roles in JA biosynthesis, were upregulated by 2.54- and 10.19-fold in the leaves and roots of Roma, respectively. The increased expression level of these genes may be related to the increase in JA content in Roma.

JAZs, which belong to the TIFY superfamily, have been characterized as critical regulators of JA signaling. It has been reported that the expression of 10 members of JAZs were quickly and specifically induced by JA treatment (Chini et al. 2007). DEGs encoding TIFYs were downregulated in the leaves of both M-81E and Roma and upregulated in the roots of Roma, which is consistent with the changes in JA content, suggesting that TIFYs may also be induced by endogenous JA. Additionally, the overexpression of TIFYs resulted in increased salt stress tolerance (Ye et al. 2009; Zhu et al. 2012; Hua et al. 2015). The changes in DEGs encoding TIFYs suggested that the JA signal pathway participates in the response of salt stress in sweet sorghum. However, the mechanism of the JA pathway involved in regulating salt tolerance in the leaves and roots may be different.

Several studies have indicated the involvement of auxin in response to salinity stress in plants, but little information is available regarding the mechanisms of salt stress regulation. In the present study, IAA content increased in the leaves of both genotypes. Changes of IAA content in the roots were not as significant as those in the leaves of both genotypes. Aldehyde dehydrogenase can catalyze the formation of indoleacetate in the IAA biosynthesis pathway, which may lead to a reduction of IAA concentration. According to our RNA-seq data, DEGs encoding aldehyde dehydrogenase were downregulated in both the leaves and roots of M-81E (Sb04g033420 and Sb04g029040, respectively). In Roma, two DEGs (Sb05g005470 and Sb02g025790) encoding aldehyde dehydrogenase were downregulated only in roots. These suggested that the IAA biosynthesis was enhanced in the roots of both genotypes and in the leaves only in M-81E. Moreover, two and four auxin transporter-like protein-encoding genes were significantly upregulated in the roots of M-81E and Roma, respectively. These results suggested that salt stress may affect not only the biosynthesis of IAA, but also its distribution. As it is known, IAA mainly plays an important role in regulating plant growth. The changes and redistribution of its concentration may be related to the decrease in FW and DW of leaves in both genotypes. These results were consistent with a previous report, in which the redistribution of IAA maxima formation in plant tissues was correlated with reduced growth (Ryu and Cho 2015).

CKs are produced in the root tips and developing seeds of plants and are translocated from the roots via xylem to the shoots, where they regulate plant development and growth processes (Zahir et al. 2001; Kakimoto 2003). Under abiotic stress, the biosynthesis of CKs in the roots is inhibited, and consequently, the supply of CKs decreased in the shoots, which could elicit appropriate responses to ameliorate the effects of stress (Tran et al. 2010; Nishiyama et al. 2011; Nishiyama et al. 2012). ZR is considered as a naturally occurring CK (Sakakibara 2006). In the present study, ZR content in the leaves and roots decreased slightly in both genotypes after treated with NaCl. Our RNA-seq data also showed that DEGs participating in ZR biosynthesis were all downregulated and DEGs participating in ZR degradation were all upregulated in the roots of both genotypes.

Genetic studies in Arabidopsis have provided evidence that CKs act as negative regulators of plant adaptation to drought (Nishiyama et al. 2011; Ha et al. 2012). Response regulators (ARRs) can mediate CK signaling (To and Kieber 2008). It has been reported that ARR 9 and ARR 12 act as negative regulators and ARR 3 acts as a positive regulator in plant responses to water stress (Wohlbach et al. 2008; Nguyen et al. 2016). In the present study, ARR 12-encoding gene Sb10g005580 was upregulated in the leaves of M-81E. ARR 3-encoding gene Sb04g022780 was downregulated in the leaves of Roma. These suggested that CK signaling was negatively regulated in the leaves of both genotypes under salt stress. However, in the roots, DEGs encoding ARRs were only found in Roma. The downregulation of ARR 9- and ARR 12-encoding genes and the upregulation of ARR 3-encoding gene suggested that CK signaling positively regulated in the roots of Roma under salt stress.

There was no significant change in BR content in both genotypes. Our transcriptomic results revealed that all DEGs related to BR biosynthesis and BR signal transduction were downregulated. These suggested that salt stress inhibits not only the biosynthesis but also the signal transduction of BR in sweet sorghum.

There was no significant change in GA content in both genotypes. DEG (Sb10g000920) related to GA biosynthesis was only found in the roots of M-81E which was upregulated by only 2.51-fold. Additionally, no DEG related to GA signal transduction was determined. These results revealed that GA may have little to do with the salt stress response of sweet sorghum.

In conclusion, we report that hormones are involved in the salt stress response of sweet sorghum. In salt-tolerant inbred line M-81E, ABA may play a key role in salt stress response. While, in salt-sensitive inbred lines, JA may act as the key hormone to salt stress regulation. ABA and JA regulate the growth and development of sweet sorghum through different regulation pathways—ABA plays roles mainly in leaves and JA plays roles mainly in the roots. Additionally, IAA also plays important roles in regulating the growth of sweet sorghum, and its redistribution may be an important mechanism for sweet sorghum in response to salt stress (Fig. 7). These results are important resources for future studies aimed at improving salt tolerance in monocotyledon crops such as sweet sorghum. Further genetic and molecular analysis would be critical to understand the detailed roles of these genes in the salt response of sweet sorghum.

Fig. 7
figure 7

Visualization of DEGs involved in pathways related to hormone biosynthesis and signal transduction. A square block represents a gene assigned to our RNA-seq data. The numbers in the scale bar show the log2 fold change of each DEG. ML and MR represent M-81E leaves and M-81E roots, respectively. RL and RR represent Roma leaves and Roma roots, respectively

Full size image

4-Cl-like, 4-coumarate-CoA ligase; ABA, abscisic acid; ABA8′-OH, ABA 8′-hydroxylase; ACXs, acyl-coenzyme A oxidase; ADL, aldehyde dehydrogenase, ABI5, ABA insensitive 5; ARRs, response regulators; CK, cytokinin; CKX, CK dehydrogenase; DW, dry weight; DEGs, differentially expressed genes; FW, fresh weight; FDR, false discovery rate; IAA, indole-3-acetic acid; JA, jasmonate; JAZ, jasmonate ZIM domain; MeJAs, methyl jasmonates; NCED, 9-cis-epoxycarotenoid dioxygenase; PYR, pyrabactin resistance; PP2C, protein phosphatase 2C; PLA1, phospholipase A1; RNA-seq, RNA sequencing; SnRK2, Snf1-related kinase; WC, water content; ZR, zeatin riboside; ZEP, zeaxanthin epoxidase