Genome-wide transcriptional and metabolic responses of Eschscholzia californica to salt stress

Abstract

Eschscholzia californica, as an abiotic stress-tolerant ornamental plant, has long been studied for phylogenetic analysis and biosynthesis of benzylisoquinoline alkaloids. However, little is known about its environmental adaptation mechanisms. In this study, we performed comprehensive investigations on responses of E. californica to salt stress, one of serious threats to modern agriculture. The effects of 250 mM NaCl solution treatment on plant growth and several stress response related indices were investigated at 0d, 7d, and 10d, representing non-stressed (CK), mildly-stressed (St0), and severely-stressed (St1), respectively. RNA-seq analysis revealed salt-stress responsive and differentially expressed genes (DEGs) enriched in photosynthesis-related pathways, phenylpropanoid biosynthesis, ɑ-Linolenic acid metabolism, isoquinoline alkaloid biosynthesis, benzoxazinoid biosynthesis, etc. With the aggravation of salt stress, pathways of the tryptophan metabolism, plant hormone signal transduction, ɑ-Linolenic acid metabolism, and isoquinoline alkaloid biosynthesis, etc., were more significantly enriched. We also identified DEGs involved in intracellular ion homeostasis, osmotic regulation, oxidative stress and detoxification, plant hormone biosynthesis and signaling, as well as those encoding transcription factors. Furthermore, Comparison of metabolites in CK and St1 showed that the differentially accumulated metabolites under salt stress were mainly alkaloids (23.6%), phenolic acids (16.5%), lipids (15.2%), organic acids (8.1%), amino acids and their derivatives (7.6%), flavonoids (6.5%), etc. Further joint analysis of transcriptome and metabolome data revealed key pathways in responding to salt stress. The jasmonic acid biosynthesis and signaling and isoquinoline alkaloid biosynthesis were particularly noteworthy due to their extremely significant up-regulation by salt stress and unclear or debatable roles in regulating salt stress tolerance. The DEGs in the two pathways, such as AOX, LOX, and JAZs, as well as NCS1, 6OMT, TNMT, and BBEs, etc., are therefore valuable for uncovering the functional roles of two pathways in salt stress tolerance of E. californica, and future breeding salt-tolerant crops.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12870-026-08201-w.

Introduction

The growth and development of plants are affected by various stresses. Salt damage caused by soil salinization is one of the main abiotic stresses affecting the overall growth and yield of crops and has been considered a serious threat to modern agriculture. Salt stress affects the physiological level of the whole plant through osmotic regulation and ion homeostasis. In general, NaCl is the main factor affecting soil salinity [1]. The high concentration of Na⁺ and Cl⁻ in the soil will destroy the ion balance. The competition between Na⁺ and plant essential nutrients hinders the absorption of other nutrients by plants, and ultimately reduces the biomass of seedlings [2]. At the same time, salt stress can destroy the integrity of chloroplast structure, damage its ultrastructure, and lead to the decrease of chlorophyll content. Previous study has found that salt stress causes damage to plant photosystem II (PSII) and inhibits the maximum quantum yield (Fv/Fm) of PSII photochemistry [3].

Plants have formed multiple morphological, physiological and molecular mechanisms to cope with the harmful effects of salinity [4]. The adaptation of plants to salinity mainly maintains the osmotic balance by accumulating high concentrations of osmotic pressure in the cytoplasm and enhances the water retention capacity of cells [5]. Plants can use organic solutes such as amino acids and their derivatives (proline, glutamic acid and glycine betaine), polyols (glycerol, mannitol and inositol), sugar compounds (sucrose, trehalose and raffinose) and polyamines (putrescine) as osmotic regulators to protect the integrity of cell membranes [6]. Plants will also employ enzyme scavengers (superoxide dismutase, catalase, peroxidase, ascorbate peroxidase, and glutathione S-transferase) and non-enzymatic systems (glutathione, ascorbic acid, carotenoids, flavonoids, alkaloids, phenolic compounds, and tocopherols) to reduce oxidative stress brought by high concentrations of salt induced ROS production [7]. In addition, plants can also adapt to salt stress by regulating ion transport process, Ca²⁺ signal transduction pathway [8], MAPK cascade reaction [9], and hormone signal transduction pathways [10].

Comparative transcriptome and metabolome analysis provides a convenient way to uncover the complex regulatory networks between genes and metabolites in responding to stress. Through comprehensive analysis of the transcriptome and metabolome of rice overexpressing (OE) a drought-tolerant gene OsDRAP1, amino acids (proline, valine), organic acids (glycine, phosphoenolpyruvate and ascorbic acid) and many secondary metabolites were found to accumulate to a higher level in the OE lines after salt stress. Finally, the key role of amino acid and carbohydrate metabolic pathways in OsDRAP1-mediated salt tolerance was emphasized [11]. Similar study in Dendrobium officinale showed that salt stress affected various metabolic pathways, such as phenylalanine metabolism, flavonoid biosynthesis and α-linolenic acid metabolism, and significantly up-regulated the mRNA expression levels involved in jasmonic acid (JA) and flavonoid biosynthesis pathways. These results indicated that the biosynthesis of secondary metabolites may play an important role in adaptation to salt stress [12].

Eschscholzia californica is a herbaceous plant of the Papaveraceae family and rich in benzylisoquinoline alkaloids (BIAs), which are limited distributed in higher plants [13]. It is also a basal eudicot plant with a small genome and short growth cycle [14], and has long been well known as a model for studies of the floral organ development [15], phylogenetic development of basic eudicot plants, as well as the BIAs biosynthesis [16]. E. californica grows in different habitats, its structure and life history have undergone great changes, which resulted in rich genetic diversity and amazing adaptability [17, 18]. For example, E. californica is a copper indicator plant, and can adapt to adverse growth environments, including places with severe drought [19–21]. However, the mechanism underlying its environmental adaptability is unclear. There is also no report on its response mechanism to salt stress.

At present, morphological and physiological responses of E. californica to salt stress was analyzed by artificially simulated salt stress. Additionally, the transcriptome and metabolomics methods were employed to analyze the salt induced genes, metabolites, as well as important pathways, which provided a preliminary and global view for the network involved in responding to salt stress.

Materials and methods

Plant growth conditions and salt treatment

Seeds of E. californica were purchased from Jiangsu Huazhiyin Seed Industry (Suqian, Jiangsu, China). The seeds were planted in the pot containing same weight of soil and grown in the green house of Sichuan Agricultural University under condition of 23 ± 2 °C, ~ 60% relative humidity and 16 h/8 h of the light/dark period. The pre-treatments with salt solutions containing different concentrations of NaCl were conducted. 100 mM NaCl has minmal impact on growth of E. californica. 200 mM and 250 mM NaCl have significant impact on the growth (Fig. S1). The latter has a more severe impact. After 41 days of treatment of 250 mM NaCl solution, although the plant growth was severely impaired compared to the control, some plants were still alive. These results demonstrated good salt tolerance of E. californica and the treatment of 250 mM NaCl was then employed for subsequent experiments.

Six-week-old (6–8 leaf stage) E. californica seedlings were treated with 250 mM NaCl solution. The above-ground parts of plants (3–4 leaves from top to bottom) were collected at 0 d (CK), 7 d (designated St0), and 10 d (designated St1) of salt treatment for physiological index determination and transcriptome analysis. The samples collected at 0 d and 7 d were also subjected to metabolome analysis. The collected samples were stored in a refrigerator at −80 °C, and the relative water content (RWC), relative electrical conductivity (REC), chlorophyll content, root-shoot ratio, and antioxidant enzyme activities were measured with fresh samples. Each treatment had three biological replicates, and each replicate contained 10 seedlings.

Phenotyping and physiological analysis

In order to evaluate the effect of salt on the morphology and stress-response-related physiological activity of E. californica, the chlorophyll content, root-shoot ratio, RWC, REC, osmotic adjustment substance content (proline, soluble sugar and soluble protein), malondialdehyde (MDA) content and antioxidant enzyme activities (SOD, POD and CAT) were measured.

RWC and REC were measured according to the method of Esringü et al. (2011) [22]. The chlorophyll content was determined according to the method of Lichtenthaler et al. (2001) [23]. The root-shoot ratio was measured according to Kapoor ‘s experimental protocol [24]. Proline was determined by acid ninhydrin method. Soluble sugar and soluble protein were determined by the method of Irigoyen [25]. MDA content, POD and SOD activity were determined by Shafi ‘s research method [26]. CAT activity, O₂⁻ and H₂O₂ content were measured by using catalase (CAT) detection kit, hydrogen peroxide detection kit, and hydrogen peroxide detection kit, respectively (Nanjing Jiancheng, Nanjing, China).

The absorbance values involved in determination of the above mentioned physiological and biochemical indicators were measured by using an Absorbance Microplate Reader (SpectraMax^® 190, Molecular Devices, USA). The wave length used were 405 nm for activity of CAT and H₂O₂ content, 450 nm, 532 nm and 600 nm for MDA content, 470 nm for activity of CAT, 520 nm for proline content, 530 nm for O₂^- content, 560 nm for activity of SOD, 595 nm for soluble protein content, 620 nm for soluble sugar content, and 649 nm and 665 nm for chlorophyll content, respectively.

RNA extraction and cDNA library construction

Total RNA was extracted from E. californica leaves. The integrity of RNA and the presence of DNA contamination were analyzed by 1% agarose gel electrophoresis and Agilent 2100 biological analyzer. The purity of RNA samples was detected by NanoPhotometer spectrophotometer. Finally, RNA concentration was measured by Qubit 2.0 fluorometer. cDNA library was prepared and according to procedures as described previously [27].

Illumina sequencing and sequence assembly

Based on Sequencing by Synthesis technology, the cDNA libraries of CK and the salt-treated samples were sequenced on the Illumina high-throughput sequencing platform (Metware, Wuhan, China). Fastp software was used to filter the data by removing adapter reads, low-quality reads, and reads with an unknown proportion of base information greater than 10% (Fig. S2). Clean reads were obtained by raw data filtering, sequencing error rate checking, and GC content distribution checking. The Trinity program was used to assemble the transcripts of clean reads, and Corset was used to cluster the assembled transcripts to remove redundancy. After obtaining the final Unigenes, functional annotation and coding sequence prediction were performed.

Annotation of salt stress-related transcripts

DIAMOND software was used to compare Unigene sequences with KEGG, NR, Swiss-Prot, GO, COG/KOG, Trembl databases, and HMMER software was used to compare amino acid sequences with Pfam database to obtain annotation information from seven database.

Differential expression analysis

Differential expression analysis between sample groups was performed using DESeq to obtain a set of differentially expressed genes between CK and the two salt treatments. After the difference analysis, the Benjamini-Hochberg method was used to perform multiple hypothesis test correction on the hypothesis test probability (P value) to obtain the false discovery rate (FDR). The corrected P value and | log2 (fold change) | were used as the threshold for significant differential expression. The screening conditions for differentially expressed genes (DEGs) were | log2 (fold change) | ≥ 1 and FDR < 0.05. GO functional enrichment analysis and KEGG metabolic pathway analysis were further performed on these DEGs to determine the GO entries significantly enriched by DEGs and the main metabolic pathways they participated in.

Real time quantitative RT-PCR (qRT-PCR) confirmation

1 µg total RNA was reverse transcribed into cDNA using a reverse transcription kit (Innovagene, Sweden). Then qRT-PCR was performed using SYBR Green premix (Accurate Biotechnology, Hunan, China) with 2 µl diluted cDNA (1:4). Ten DEGs at St0 or St1 stages under salt stress were randomly selected for real-time fluorescence quantitative PCR analysis on a real-time fluorescence quantitative PCR detection system (Bio-Rad, California, USA). The reaction conditions were: 95 °C 3 min, 95 °C 10 s, 60 °C 30 s, 65 °C 5 s, 95 °C 5 s, 40 cycles. The qRT-PCR primer design was performed using NCBI Primer Blast (Table S1). The data were normalized using the E. californica actin gene. The expression level of the gene was calculated by 2^−ΔΔCt relative quantification method, with three replicates per sample.

Sample pretreatment, metabolite extraction and quantification

The samples were placed in a vacuum freeze-drying machine (Scientz-100 F), and then ground (30 Hz, 1.5 min) to powder using a grinder (MM 400, Retsch). 50 mg powder was dissolved in 1.2 mL 70% methanol extract, and then vortexed and mixed. After centrifugation (12000 rpm, 3 min), the supernatant of the sample was taken and filtered with a microporous membrane (0.22 μm pore size). The Ultra Performance Liquid Chromatography (ExionLC™ AD) and Tandem Mass Spectrometry (Applied Biosystems 4500 QTRAP) (UPLC-MS/MS) were employed for metabolite identification and quantificaation, which were conducted by Metware Biotechnology Co., Ltd. (Wuhan, China).

Screening of differential metabolites

Based on the results of orthogonal partial least squares discriminant analysis (OPLS-DA), the variable importance in projection (VIP) of the obtained multivariate analysis OPLS-DA model can preliminarily screen out different groups of differential metabolites. The metabolites with VIP > 1, p ≤ 0.05, Fold Change ≥ 2 or Fold Change ≤ 0.5 are considered as differentially accumulated metabolites (DAMs).

Statistical analysis

The data of physiological indexes in this study were analyzed by ANOVA (Analysis of variance) analysis. The measured values were expressed as mean ± standard deviation (SD) of three biological replicates, and the asterisk (*) indicated significant difference (single asterisk represented p < 0.05 and double asterisk represented p < 0.01).

Results

Morphological and physiological alterations under salt stress

Under control condition, E. californica leaves were bright green, showing a good state of health. The dehydration symptoms of plants treated with 250 mM NaCl were gradually aggravated. After 7 d of salt stress (St0), the leaves of seedlings drooped, some leaves turned yellow, and the new leaves were gray-green; after ten days (St1), the leaves were obviously yellowed, the plant growth was inhibited, the leaf margin was rolled and burned, and gradually withered and necrotic (Fig. 1a). We measured the root-shoot ratio of E. californica plants under salt stress for 10d. There was no significant change in the root-shoot ratio of E. californica plants cultivated under normal conditions. However, the root-shoot ratio of plants at St0 and St1 increased, indicating that salt stress significantly reduced the biomass of its aboveground part (Fig. 1b).

Fig. 1.

Phenotypic and physiological responses of E. californica to salt stress. a, General morphology of E. californica plants treated by salt stress (250 mM NaCl solution) at 0 d, 7 d, and 10 d; b-n, Measurements of root-shoot ratio, RWC, chlorophyll content, REC, proline content, soluble protein content, soluble sugar content, superoxide anion (O₂⁻) content, hydrogen peroxide (H₂O₂) content, activities of SOD, POD, and CAT, and MDA content, respectively. The data were expressed as the mean and standard deviation of the three biological replicates. The single asterisk (*) and the double asterisk (**) represented the significant difference (p < 0.05 and p < 0.01, respectively), which were determined by ANOVA

We further analyzed several stress-related physiological parameters. RWC and chlorophyll content under salt stress were significantly lower than those of control plants (CK) (Fig. 1c, d). REC is an index to measure the membrane permeability of plants. After salt treatment, REC increased significantly from 10% to 33% (Fig. 1e), indicating that salt stress increased the ion leakage and the membrane permeability. The proline content increased greatly after salt stress, and the increase rate was slowed down significantly after 7 d, but still maintained at a high level (Fig. 1f). Content of soluble protein decreased after 7 d of salt treatment, and then increased slightly and turned back to similar level to none treated plants at St1 (Fig. 1g). The content of soluble sugar decreased after salt treatment, and increased rapidly after 7 d of salt stress, which was significantly higher than that of the control (Fig. 1h).

We studied degree of oxidative stress under salt stress. SOD can convert O₂⁻ into H₂O₂, while CAT and POD catalyze the decomposition of the latter. Our results showed that the accumulation patterns of O₂⁻ and H₂O₂ were different under salt stress treatment. The O₂⁻ content showed an upward trend (Fig. 1i) at St0 and St0, whereas the H₂O₂ content did not change decidedly at St0, but decreased significantly at St1 (Fig. 1j). which was consistent with the decreased SOD activity (Fig. 1k). The activities of CAT and POD showed a similar increasing trend (Fig. 1l, m) after salt stress. MDA content increased sharply after salt treatment, confirming that the plasma membrane of E. californica seedlings was impaired by the salt stress (Fig. 1n).

Transcriptome sequencing, de Novo assembly, and functional annotation

The leaves under normal conditions and salt stress treatments were sampled for RNA-seq analysis. After data filtering and quality control, a total of 60.35 Gb clean data was obtained from nine samples (Table S2). The percentage of Q20 and Q30 bases in the original data reached 97.37% and 92.38%, respectively. The base distribution after filtration was relatively stable, and the GC content was 41.66%. These results indicated that the sequencing quality was satisfied for subsequent analysis.

The length of the assembled sequence can reflect the quality of the assembly. A total of 136,850 transcript sequences were assembled by using TRINITY, and resulted in 132,018 unigenes after clustering with Corset software. The longest unigene was 16,863 bp, and the average length was 1041 bp. About 86% of the unigenes were less than 2000 bp, and there were 18,773 sequences were longer than 2000 bp, showing high assembly integrity (Fig. S3).

The obtained unigenes were aligned to the seven functional databases of KEGG, NR, Swiss-Prot, Trembl, KOG, GO, and Pfam using DIAMOND BLASTX software. About 53.65% of the unigenes was annotated in at least one database, in which 40,661 (30.80%) unigenes were matched with known proteins in KOG, and the genes number annotated in KEGG, Swiss-Prot, Trembl, GO and Pfam databases was 50,108 (37.96%), 50,008 (37.88%), 69,303 (52.56%), 58,966 (44.67%) and 47,202 (35.75%), respectively (Table S3).

The GO annotation functionally classified genes into three categories of biological process, cellular component and molecular function. Among the 58,966 unigenes annotated in the GO database (Fig. 2a), 48.2% was annotated with GO terms of biological processes, which enriched in cellular processes (GO: 0009987) and metabolic processes (GO: 0008152); 31% of the annotated unigenes were assigned to GO terms of cell component, which enriched in cells (GO: 0005623), cell components (GO: 0044464) and organelles (GO: 0044422). The molecular function contained about 20.6% of the annotated unigenes, with predominant unigenes annotated to terms of binding (GO: 0005488) and catalytic activity (GO: 0003824).

Fig. 2.

Unigene annotations. a GO classification; (b) KOG classification; (c) Statistic of similarities to other plant species based on NR annotation

Approx. 40,661 (30.80%) unigenes annotated in the KOG database can be divided into 25 functional categories, of which general function prediction accounted for the largest proportion (26.08%). Posttranslational modification, protein turnover, and chaperones accounted for about 10.88%, signal transduction mechanisms accounted for about 9.89%, transcription accounted for about 5.94%, and carbohydrate transport and metabolism accounted for about 5.74% of the total KOG annotated unigenes, respectively (Fig. 2b). The homology annotation of E. californica unigenes using the NR database (Fig. 2c) showed that the species with the highest similarity was Macleaya cordata (68.13%), followed by Nelumbo nucifera (5.16%), Aquilegia coerulea (3.13%), Vitis vinifera (2.3%), and Juglans regia (0.7%).

Quantitation of expression levels and analysis of DEGs

The clean reads were mapped to the unigene set to determine gene expression levels. The mapping rate of each sample was greater than 77% (Table S2). PCA (Principal Component Analysis) of all samples indicated that the samples treated by salt stress were clustered distinctly from CK, indicating that the salt stress significantly altered the overall expression levels (Fig. S4). However, the samples of St0 and St1 were partially overlapped, exhibiting their similar genome-wide expression status. Ten unigenes were randomly selected to verify the gene expression quantified by RNA-seq. qRT-PCR analysis showed that these genes exhibited similar expression trend with that of RNA-seq, confirming the accuracy of expression quantification (Fig. S5).

In order to further explore the general expression pattern of DEGs under salt stress, unigenes with fold change (log2) of expression larger than 1 (or less than − 1) and false discovery rate (FDR) < 0.05 were considered as DEGs. Compared to CK, a total of 5728 DEGs at St0 were identified, of which 3451 were up-regulated and 2277 were down-regulated, respectively. There were 9517 DEGs at St1, including 5562 up-regulated and 3955 down-regulated, respectively (Fig. 3a, b). Among them, 4281 DEGs were identified simultaneously at two time points (Fig. 3c). The DEGs number at St1 was almost twice of that at St0. However, there were only 480 DEGs between St0 and St1. These results suggested that the intensity of stress degree of St1 was higher than that of St0, which may just exceed some threshold, triggering a more comprehensive and intense transcriptional reprogramming.

Fig. 3.

Identification of DEGs. a and (b)indicated identification of DEGs between CK and St0, and CK and St1, respectively. Venn diagram (c) showed comparison of the DEGs numbers between two groups

Gene ontology (GO) enrichment of DEG

GO classification of DEGs between CK, St0, and St1 was performed. Between CK and St0, about 60% of DEGs could be annotated to biological process category, in which metabolic process and cellular process accounted for the highest proportion; In the molecular function category, the binding and catalytic activity represented the most abundant terms; Less than 10% of DEGs were classified in the cellular composition category, and cells and cell parts accounted for the highest proportion. The 9517 DEGs between CK and St1 shared similar GO classification pattern as those of CK-St1 group (Table S4, S5).

Further GO enrichment analysis of DEGs was performed. In the biological process category, DEGs in CK/St0 group were enriched in 194 GO terms (P ≤ 0.01), in which photosynthesis related GO terms, protein-chromophore linkage, jasmonic acid metabolic process, jasmonic acid biosynthetic process, drug catabolic process, cellular lipid catabolic process, hormone biosynthetic process, polysaccharide catabolic process, and proline metabolic process were the most significantly enriched Go terms; DEGs in CK/St1 group were enriched in much more GO terms (290), in which half of them were shared with those of CK/St0 (Fig. S6).

In the cell components category, DEGs of CK/St0 group were mainly enriched in 18 GO terms (P ≤ 0.01), in which the GO terms related to photosystem were most significantly enriched; DEGs of CK/St1 group were enriched in 19 GO terms, and 13 of them were also shared by CK/St0 group. In the molecular function category, DEGs of CK/St0 group were enriched in 148 GO terms, in which chlorophyll binding, pigment binding, drug transmembrane transporter activity, organic anion transmembrane transporter activity, amino acid transmembrane transporter activity, and lipase activity were the most enriched. DEGs of CK/St1 group were enriched in more GO terms (189) than those of CK/St0 group, however, the most enriched GO terms in the former were also shared by the latter. These results indicated that both CK/St0 and CK/St1 exhibited some similar responses to salt stress at the transcriptional level, but the latter made more complex changes to cope with longer-term and more serious stress.

KEGG pathway analysis of DEGs

In CK/St0 group, 2125 DEGs were annotated into 136 pathways (Table S6). Biosynthesis of secondary metabolites (ko01110) and Phenylpropanoid biosynthesis (ko00940) were among the most significantly enriched pathways (Corrected p-value = 0), followed by Metabolic pathways, Photosynthesis-antenna proteins, ɑ-Linolenic acid metabolism, Benzoxazinoid biosynthesis, Fatty acid degradation, and Biosynthesis of unsaturated fatty acids (Corrected p-value ≤ 0.01)(Fig. 4a).

Fig. 4.

KEGG enrichment analysis of E. californica under salt stress. a KEGG enrichment scatter diagram after 7 days of salt stress; (b) KEGG enrichment scatter diagram after 10 days of salt stress; (c) Comparison of KEGG enrichment scatter diagram between 7 d and 10 d of salt stress

The 3475 DEGs of CK/St1 group were annotated to 137 KEGG pathways, in which Biosynthesis of secondary metabolites, Phenylpropanoid biosynthesis, and Isoquinoline alkaloid biosynthesis were the most significantly enriched pathways (Corrected_P-value = 0), followed by Metabolic pathways, Photosynthesis - antenna proteins, Photosynthesis, Benzoxazinoid biosynthesis, ɑ-Linolenic acid metabolism, Carbon metabolism, etc. (Table S7, Fig. 4b). Generally, although CK/St1 group has much more DEGs than those of CK/St0 group, the patterns of their KEGG pathway enrichment were similar. With the increase of salt stress treatment time, the biosynthesis of isoquinoline alkaloids (ko00950), plant hormone signal transduction (ko04075), tryptophan metabolism (ko00380), indole alkaloid biosynthesis (ko00901) and MAPK signaling pathway-plant (ko04016) and other metabolic pathways were enriched with more DEGs (Fig. 4c), indicating that these pathways became more active under salt stress in E. californica, and the DEGs involved in these pathways may play essential roles in regulating salt stress responses.

Key genes involved in the response to salt stress

DEGs involved in ion transport

A total of 35 genes encoding cyclic nucleotide-gated ion channel proteins (CNGC2, CNGC4, CNGC15b, CNGC18, CNGC19, CNGC20), glutamate receptor-like proteins (GLR22, 29 and 34) and voltage-gated chloride channel (CLC) were annotated (Fig. S7a). The transcript abundance of most of the non-cation selective channel were down-regulated at both St0 or St1. The KAT3 (AKT4) encoding potassium channel was down-regulated at St0 and St1, and three potassium channel protein transcripts (AKT1 and SKOR) were up-regulated in St1. In addition, aquaporins were also known to play an important role in plant salt stress response. Most of the 19 aquaporins were significantly down-regulated at St1, including tonoplast intrinsic protein (TIP) and plasma membrane intrinsic protein (PIP). The six genes (1 HAK7, 2 HAK11, 1 HAK18, 1 POT10, and 1 POT11) encoding potassium transporters were up-regulated at two time points of salt stress, and the transcript of KT2 (KUP2) was always down-regulated. A total of six cation/H⁺ antiporters (CHXs) and 15 vacuolar H⁺-ATPases were identified in this study. At St0 and St1, a total of eight genes encoding vacuolar H⁺-ATPases were continuously up-regulated, which may confer salt resistance to plants by promoting the Na⁺ transportation (Fig. S7a). A total of 83 DEGs encoding ABC transporters were identified. The majority of these DEGs were significantly up-regulated at St0 and St1 (Fig. S7b).

DEGs involved in osmotic regulation

Osmolytes, such as proline, sucrose, raffinose, mannitol, and trehalose, are important for salinity stress adaptation by maintaining osmotic pressure and stabilizing cellular structure. In this study, 21 DEGs associated with proline synthesis and decomposition pathways were identified, including Δ−1-pyrroline-5-carboxylic acid synthase (P5CS), pyrroline-5-carboxylic acid reductase (P5CR), ornithine aminotransferase (OAT), proline dehydrogenase (ProDH) and P5C dehydrogenase (P5CDH) (Fig. 5a-b). In addition, 12 sucrose synthase (SUS), one sucrose phosphatase (SPP), two sucrose phosphate synthase (SPS) (Fig. 5c-d), eight raffinose synthase (RFS), three mannitol dehydrogenase (MTD) (Fig. 5e-g), five 6-phosphatase trehalose synthase (TPS), and eight 6-phosphatase trehalose phosphatase (TPP) (Fig. 5h-i) were identified at St0 and St1. These transcripts were differentially expressed, indicating significant promotion of osmolytes synthesis of E. californica under salt stress.

Fig. 5.

Expression heat map of osmotic regulation related DEGs under salt stress. Proline (a), sucrose (c), raffinose (e), mannitol (f) and trehalose (h) biosynthetic pathway schematic diagram and proline (b), sucrose (d), raffinose and mannitol (g) and trehalose (i) synthesis pathway related coding gene expression heat map

DEGs involved in oxidative stress and detoxification

Five respiratory burst oxidase homolog (RBOH) genes were significantly up-regulated after salt stress (Fig. S8). In particular, 16 NADPH-related oxidases were up-regulated at St1. A large number of DEGs related to oxidoreductase activity were found, indicating that salt stress induced oxidative stress in E. californica. Under salt stress, genes encoding antioxidant enzymes, such as superoxide dismutase (SOD), catalase (CAT), peroxidase (POD), ascorbate peroxidase (APX), glutathione reductase (GR) and glutathione-S-transferase (GST), were differentially expressed. Among them, 13 genes encoding POD and 13 genes encoding GST were significantly up-regulated, which play important roles in scavenging ROS pathway.

DEGs involved in biosynthesis and signal transduction of plant hormone

KEGG analysis showed significant enrichment of DEG in carotenoid synthesis pathway, α-linolenic acid synthesis pathway, and tryptophan synthesis pathway cysteine, which are closely related to the synthesis of plant hormones abscisic acid (ABA), JA, indole-3-acetic acid (IAA) and ethylene (ET), respectively.

Most DEGs in the ABA synthesis pathways were up-regulated in response to salt stress (Fig. 6a, b) excepting for one phytoene desaturase (PDS), three lycopene ε-cyclase (LYCE), one carotenoid isomerase, one monooxygenases (MO2), one violaxanthin decyclase (VDE), two 9-cis-epoxycarotenoid dioxygenase (NCED4), and one short-chain dehydrogenase reductase 2a (SDR2A). In respect to ABA signal transduction pathway (Fig. 6c), three ABA receptor genes PYR-LIKE (PYL), and four positive regulators SAPK7 of sucrose non-fermenting-1-related protein kinase (SnRK2) family were down-regulated, whereas ten negative regulator protein phosphatase 2 C (PP2C) genes, SAPK2 and SAPK10 of SnRK2 family, and four downstream ABA response element binding factors (AREB/ABF) of bZIP family were up-regulated.

Fig. 6.

DEGs related to hormone biosynthesis and signal transduction pathways responding to salt stress. The schematic diagram of ABA, JA and IAA biosynthesis and signal transduction pathways and the heatmaps of DEGs involved in these pathways were showed

The α-linolenic acid synthesis pathway related to JA synthesis was the most significantly enriched at St0 and St1. Total of 69 unigenes showed differential expression levels at least one of the two stress treatment time points compared with CK. At St0, 53.6% of these unigenes were up-regulated, and 10.1% were down-regulated. Interestingly, 82.6% of these unigenes were up-regulated at St1, including seven acyl-CoA oxidase 1 (ACOX 1/AOX1), six acyl-CoA oxidase 4 (ACOX4/AOX4), seven lipoxygenases (LOX6, LOX21 and LOX31) and three allene oxide synthase 1 (AOS1) (Fig. 6d, e). Several unigenes encoding the major components in the JA signal transduction pathway, i.e. jasmonic acid-amino acid synthase 1 (JAR1) and TIFY transcription factors (Jasmonate ZIM domain-containing protein, JAZ), were up-regulated under salt stress, especially at St1, whereas a MYC2 transcription factor gene was down-regulated (Fig. 6f).

In pathway related to IAA biosynthesis, 16 unigenes were up-regulated at St0 and St1, whereas 24 genes were differentially up-regulated at St1 (Fig. 6g, h). A total of 53 DEGs (Fig. 6i) were detected in the IAA signal transduction pathway. Among the 11 indole-3-acetic acid-amide synthase (GH3) genes, one GH3.6 was down-regulated at St1, and the remaining ten GH genes (six GH3.6, two GH3.5, and two GH3.1) were up-regulated. The expression of unigenes encoding auxin response proteins SAUR21, SAUR50 and SAUR36 was down-regulated, while SAUR32 and SAUR72 were activated by salt stress. In addition, a total of 21 members of the auxin-inducible protein/auxin-responsive protein (AUX/IAA) gene family and eight auxin response factors (ARFs) showed differential expression.

DEGs involved in photosynthetic regulation

The overall DEGs in the photosynthesis-related pathways showed a downward trend. In the photosynthesis (ko00195) pathway, 24 and 53 DEGs were down-regulated at St0 and St1, respectively, including those encoding ATP synthase subunits (ATPA, ATPD, ATPG), chloroplast envelope protein (CEMA), cytochrome b6-f complex, plastocyanin, iron redox protein 1 (Fd1), and iron redox protein-NADPH reductase 1 (FNR1) (Fig. S9a). In addition, the expression levels of 16 DEGs encoding Psa (PsaA, PsaD, PsaE, PsaF, PsaH, PsaL, PsaN, PsaG, PsaO) and 21 DEGs encoding Psb (PsbC, PsbM, PsbO, PsbP, PsbW, PsbR, PsbS, PsbQ) were decreased, and only two iron redox protein 3 (Fd3) genes and one iron redox protein-NADPH reductase 2 (FNR2) gene showed increased expression levels under salt stress.

In the photosynthesis-antenna protein (ko00196) pathway, chlorophyll a/b binding protein (CAB) is also called light-harvesting complex protein (LHC). Salt stress decreased expression of a total of 47 unigenes encoding LHC proteins (Fig. S9b). Among these genes, 74.4% were down-regulated at St0 and 93.6% were down-regulated at St1, respectively. None of unigenes encoding photosynthesis-antenna protein was up-regulated by salt stress (Fig. S9b). These results indicated that salt stress caused significant damage to photosystem I and photosystem II of E. californica leaves, which was consistent with the significantly reduced leaf chlorophyll content under salt stress (Fig. 1d).

DEGs involved in biosynthesis of isoquinoline alkaloids

Genes involved in isoquinoline alkaloid synthesis were significantly up-regulated, including aspartate aminotransferase, tyrosine aminotransferase (TAT2), (S) -desmethylaconitine synthase (NCS1 and NCS2), (RS) -desmethylhengzhouaconitine 6-O-methyltransferase (6OMT), (S) -hengzhouaconitine N-methyltransferase (CNMT), cytochrome P450 (CYP719A2, CYP719A3 and CYP80B2), polyphenol oxidase (PPO), DOPA decarboxylase (DDC1 and DDC2), primary amine oxidase (AOC), (S) -tetrahydroprotoberberberine N-methyltransferase (TNMT), protopine 6-monooxygenase (P6H), and berberine bridge enzyme (BBE8, BBE12, BBE13, BBE18, BBE26 and BBE25), etc. (Fig. 7). These results showed that the biosynthesis of isoquinoline alkaloids was activated by salt stress.

Fig. 7.

Heat map of DEGs related to isoquinoline alkaloid biosynthesis (ko00950) pathway under salt stress

DEGs encoding transcription factors

Transcription factors (TF) play a key role in plant response to stress. Using iTAK software, 3752 transcription factors were predicted from 132,018 unigenes. We screened these DEGs for those having average FPKM value ≥ 1 in at leaset one of the treatments. Three hundred forty-five DEGs in CK/St0 group encoded TFs, in which 215 were up-regulated and 130 were down-regulated (Table S8), and 547 DEGs in CK/St1 group encoded TFs, in which 348 were up-regulated and 199 were down-regulated, respectively (Table S9). A total of 286 DEGs were found simultaneously at both St0 and St1, in which 184 were up-regulated and 102 were down-regulated, respectively. The top 10 TF families in the DEG numbers were NAC (8.7%), WRKY (6.8%), bHLH (5.7%), MYB-related (5.4%), AP2-ERF-ERF (5.4%), AUX/IAA (4.0%), MYB (3.9%), bZIP (3.6%), C2H2 (3.3%), and GARP-G2-like (2.6%) (Fig. 8).

Fig. 8.

Top 10 TF families identified from DEGs under salt stress. Statistics was based on 606 TF coding non redundant DEGs in CK/St01 and CK/St02

Metabolome analysis under salt stress

To analyzing metabolic changes under long term salt stress, quantification of widely targeted metabolites was performed on samples of CK and St1. A total of 1124 primary and secondary metabolites were detected. The types of metabolites from high to low were Alkaloids (19.84%), Flavonoids (19.22%), Phenolic acids (16.01%), Lipids (11.48%), Organic acids (6.94%), Nucleotides and derivatives (5.52%), Amino acids and derivatives (4.98%), Lignans and Coumarins (4%), Tannins (0.36%) and Terpenoids (0.98%) (Fig. S10a). Correlation analysis showed that the biological replicates in each treatment were highly correlated (r > 0.96) (Fig. S10b), which is consistent with the PCA result (Fig. S9c). Additionally, Orthogonal Partial Least Squares Discriminant analysis (OPLS-DA) was performed to analyze the metabolic data. Plotting based on OPLS-DA score separated CK samples from St1 (Fig. S11). All these data showed good repeatability within each treatment group and the salt stress significantly altered metabolites in the E. californica.

Identification of DAMs and KEGG enrichment

VIP (Variable Importance in Projection) and fold change (FC) value were used to discriminate DAMs. When setting VIP ≥ 1, and FC (log2) ≥ 1 or ≤ −1, 381 DAMs were identified, of which 312 DAMs were up-regulated and 69 were down-regulated (Table S10), respectively. The main types of DAM were alkaloids (90, 23.6%), phenolic acids (63, 16.5%), lipids (58, 15.2%), organic acids (31, 8.1%), amino acids and their derivatives (29, 7.6%), flavonoids (25, 6.5%), etc. (Fig. S12a). In addition, the contents of many saccharides, such as L-fucose, gluconic acid, D-mannitol, rhamnose, maltotriose, etc., were differentially accumulated, in which more than 60% were increased.

Alkaloids were apparently the most abundant DAM under salt stress, and most of them (80, 88.8%) showed increased content. The isoquinoline alkaloids (25), indole alkaloids (19), phenolic amines (11), and aporphine alkaloids (7) were the major types, contributing to 27.7%, 21.1%, 12.2%, and 7.7% of the total differential accumulated alkaloids (Fig. S12b), respectively. Interestingly, all of the 25 isoquinoline alkaloids were up-regulated (Table S10).

KEGG annotation showed that 130 DAMs could be classified in 85 pathways. Enrichment analysis indicated that Tryptophan metabolism (ko00380), Lysine degradation (ko00310), Sphingolipid metabolism (ko00600) and Carbon metabolism (ko01200), etc., were the most significantly enriched pathways (P < 0.05) under salt stress (Fig. S12c).

Joint analysis using transcriptome and metabolome data

Several KEGG pathways enriched with both DEG and DAM were identified, such as carbon metabolism (ko01200), biosynthesis of amino acid (ko01230), starch and sucrose metabolism (ko00500), phenylpropanoid biosynthesis (ko00940), cysteine and methionine metabolism (ko00270), glycerolipid metabolism (ko00561), isoquinoline alkaloid biosynthesis (ko00950), and tryptophan metabolism (ko00380), etc. (Fig. S13).

There were 12 metabolites enriched in tryptophan metabolism (ko00380), in which the content of indole alkaloids, such as methoxyindoleacetic acid, serotonin, N-acetyl-5-hydroxytryptamine, Tryptamine, N-hydroxytryptamine, indole, etc., were significantly up-regulated by salt stress. Tryptophan metabolism is essential for biosynthesis of indole-3-acetic acid (IAA), a key growth hormone in most plants. Although the IAA was not detected, a large number of DEG involved in IAA biosynthesis and signal transduction were identified. The α-linoleic acid metabolism (ko00592) is associated with biosynthesis of plant hormone JA. All six DAMs, 13 S-hydroxy-9Z,11E, 15Z-octadecadienoic acid, 9-hydroxy-10,12,15-octadecadienoic acid, α-linolenic acid, 12-oxo-phytodienoic acid, JA, and 2-dodecadienoic acid, in the pathway were up-regulated under salt stress, in which the content of JA was 11.38 times (Log2 fold change) higher that of the CK. Consistently, the pathways of JA biosynthesis and signaling were significantly promoted by salt stress according to DEG analysis. It is interesting that isoquinoline alkaloid biosynthesis pathway was significantly promoted by salt stress at both transcriptome and metabolome levels. All eight DAMs enriched were up-regulated. Among the 90 DEGs enriched in this pathway, 86 were up-regulated.

Correlation analysis between DEGs and DAMs with coefficient > 0.80 or < −0.80 and P < 0.05 in these KEGG were also analyzed. For tryptophan metabolism pathway (ko00380), 250 DEGs were significantly and positively correlated with 12 DAMs, whereas 178 DEGs showed negative correlation, indicating highly dynamic alterations in this pathway under salt stress (Fig. 9a). Most of DEGs in the α-linoleic acid metabolism (ko00592) and isoquinoline alkaloid biosynthesis (ko00950) pathways were positively correlated with DAMs, showing that the two pathways mainly responded positively to salt stress (Fig. 9b, c).

Fig. 9.

Correlation network between DEGs and DAMs of tryptophan metabolism pathway (ko00380) (a), α-linoleic acid metabolism (ko00592) (b), and isoquinoline alkaloid biosynthesis (ko00950) (c). Green nodes indicated DEGs, and red nodes indicated DAMs. Gray line indicated positive correlation and red line indicated negative correlation, respectively

Discussion

About 20% of arable land and 33% of irrigated farmland worldwide are threatened by salinization, leading to reduced crop yields or even complete crop failure [28]. Dissection of plant salt tolerance mechanism will contribute to breeding salt tolerant crops which can stabilize food production in saline areas and alleviate the pressure of land resource shortage [29]. E. californica is extremely tolerant to cold, drought and barren conditions and can grow in coastal, river valley, foothill, and desert areas [27]. However, its responses to salt stress has not been well investigated.

In this study, we firstly analyzed morphological and physiological responses of E. californica to salt treatment. We found that the salt stress treatment with 250 mM NaCl has significantly negative effects on plants growth, reduced leaf RWC, chlorophyll content, and increased root/shoot ratio (Fig. 1). The significantly increased REC and MDA and O₂⁻ content reflected damage on membrane system and oxidative stress, respectively. Antioxidant enzyme system was promoted to scavenge oxidative stress. The activities of POD and CAT were significantly increased, whereas that of SOD was decreased. This result was different to those under drought stress [27]. We also found significantly increased content of proline and soluble sugar, which act as major osmolytes of E. californica under salt stress.

To further investigate the transcriptional regulatory mechanisms in response to salt stress, the comparative transcriptome analysis at different stress time points was conducted. Analysis of DEG and subsequent GO and KEGG annotation and enrichment identified a great number of DEGs involved in pathways related to ion homeostasis, osmotic regulation, ROS synthesis and clearance, hormone biosynthesis and signaling transduction, etc., which constitute the regulatory network for response and tolerance of E. californica to the salt stress.

Maintainers of ion homeostasis under salt stress

Promoting Na⁺ efflux, ensuring K⁺ absorption, and maintaining intracellular Na⁺/K⁺ homeostasis are essential to plant salt tolerance [30]. Under salt stress conditions, high concentrations of Na⁺ enter the cytoplasm through non-selective cation channel (NSCC) [31] and high affinity K⁺ transporter (HKT) proteins on the plasma membrane of plants [32]. With the occurrence of salt stress, plasma membrane potential activates voltage-gated channels and leads to K⁺ efflux, which leads to rapid loss of potassium in the cytoplasm, thereby destroys the homeostasis of cytoplasmic Na/K ratio [33].

GLR and CNGC are large gene families that encode NSCC proteins common in plants [34]. Arabidopsis cngc10 mutant showed stronger salt tolerance than WT, indicating that AtCNGC10 may mediate Na transport and participate in negative regulation of Arabidopsis salt tolerance [35]. Unlike NSCC, potassium channel protein plays positive roles in salt tolerance. Overexpression of PutAKT1 of Puccinellia tenuiflora in Arabidopsis led to stronger salt tolerance [36] Overexpression of OsKAT1 in rice confered salt tolerance by enhancing cell uptake of potassium ions and participating in the regulation of cytoplasmic cation homeostasis [37]. Transgenic Arabidopsis overexpressing a potassium channel gene CmSKOR of melon had better growth and root length development under salt stress [38]. The transcript abundance of most of the NSCC genes were down-regulated at both St0 or St1, but three unigenes (AKT1 and SKOR) encoding potassium channel proteins were up-regulated in St1 (Fig. S7).

Potassium transporter/high affinity potassium transporter/potassium uptake protein (KUP/HAK/KT) family plays a key role in the process of high affinity potassium transport in plants ([39, 40]. We identified five up-regulated DEGs and one down-regulated DEG encoding potassium transporters HAK7, HAK11, HAK18 and POT11, and KT2 (KUP2), respectively.

In higher plants, the discharge of salt from cells and the transfer of salt to vacuoles are completed by anti-transporters, which requires an electrochemical gradient generated by proton pumps, such as vacuolar H⁺-ATPase, to provide a driving force for cation transmembrane transport [41]. Some studies reported that overexpression of vacuolar H⁺-ATPase genes endowed plants with strong salt tolerance [42, 43]. In addition, ATP-binding cassette transporters (ABC transporters) have been shown to act not only as ATP-dependent pumps, but also as ion channels and channel modulators to transport diverse substrates from the cytoplasm across the membrane to the vacuole in a reverse concentration gradient [44, 45]. We identified 15 DEGs annotated as H⁺-ATPases generally showing up-regulated expression in responding to salt stress. A large number of ABC transporter genes showing differential and complex expression patterns were also found, most of them were positively responsive to salt stress. Thus, our data provided a global view of ion channel proteins, transporters, as well as H⁺-ATPases, which may work together to maintain intracellular ion homeostasis under salt stress.

Regulators of osmotic balance

The amino acid and carbohydrate metabolic pathways play important roles in salt stress response, as that the stress-induced accumulation of amino acids or carbohydrates help to maintain cell structure, osmotic balance, and provide energy [46]. Proline, as a well known osmotic regulator, acts on regulating the homeostasis of sodium and potassium, stabilizing cell structure and antioxidant enzyme activity under salt stress [47, 48]. The proline content closely depends on expression patterns of key enzyme genes in the proline biosynthesis pathway, such as Δ−1-pyrroline-5-carboxylic acid synthase (P5CS), pyrroline-5-carboxylic acid reductase (P5CR) [49] and ornithine aminotransferase (OAT) [50], as well as metabolism enzyme genes, like proline dehydrogenase (ProDH/POX) and P5C dehydrogenase (P5CDH) [51]. Overexpressing P5CS of moth bean in carrot [52] and PvP5CS1 and PvP5CS2 of Phaseolus vulgaris in Arabidopsis [53] significantly increased proline levels and salt tolerance. Similarly, protective roles of trehalose [53], sucrose [54], mannitol [55], raffinose [56] and genes related to their biosynthesis were also proved.

Physiological index measurements showed that proline and soluble sugar were predominantly accumulated under salt stress. Our metabolic profiles further revealed the significant salt-induced accumulation of amino acids and carbohydrates, such as proline, L-fucose, D-threonine, gluconic acid, sedoheptulose, D-mannitol, rhamnose, maltotriose, D-ribose, trehalose-6-phosphate, etc. Consistently, the pathways involved in biosynthesis and metabolism of these substances were very active in responding to salt stress, and the identified DEGs involved in these pathways (Fig. 5) formed abundant regulators to maintain osmotic balance and therefore contributed significantly to the salt stress tolerance of E. californica.

Scavenger of oxidative stress

Reactive oxygen species (ROS), mainly composed of free radicals such as hydroxyl (OH⁻), superoxide anion (O₂⁻) and non-radicals such as hydrogen peroxide (H₂O₂), is the inevitable product of normal growth and metabolism of plants, and it is also an important signal for plants to resist environmental stress [57]. But stress-induced excessive accumulation of ROS can lead to oxidative stress [58]. Chloroplasts, mitochondria and peroxisomes are the main subcellular compartments for ROS production during abiotic stress in plants [59–61]. It can also be produced in the extracellular space through NADPH oxidase on the plasma membrane and peroxidases (POXs) on the cell wall [62]. NADPH oxidase, namely Respiratory Burst Oxidase Homologs (RBOHs), is one of the key components of the ROS plasma membrane oxidoreductase system [63]. We detected generally up-regulated expression of DEGs encoding RBOHs and NADPH oxidase, which was consistent with significant ROS (O₂⁻) accumulation under salt stress (Fig. S8).

Antioxidant enzymes, such as SOD, CAT, POD, APX, GR and GST, play important roles in scavenging ROS. Overexpression of their coding genes in sweet potato, peanut, cotton and tobacco plants were proved to enhance the salt tolerance [64–66]. Our physiological investigation indicated that activities of CAT and POD were significantly increased under salt stress, whereas that of SOD was decreased. The increased activities of CAT and POD were in accordance with the generally enhanced expressions of CAT and POD genes, although there were some POD genes showing decreased expression. The contradictions of the SOD activity in responding to salt stress were found among plants. For instance, SOD activity was increased in leaves of Brassica juncea [67], but was diminished in Vigna unguiculata [68], indicating that the the responses of SOD to salt stress may be plant species dependent. This may also be related to the complex expression pattern of the SOD genes. We detected three SOD DEGs, in which those encoding one Copper chaperone for superoxide dismutase (Cluster-4700.44373) and one Cu²⁺/Zn²⁺ superoxide dismutase SOD1 (Cluster-4700.74033) were increased, whereas another unigene (Cluster-4700.78556) encoding Copper chaperone for superoxide dismutase was decreased (Fig. S8). Additionally, multiple GST genes were significantly up-regulated, which together with SOD, CAT, and POD constitutes the main components of the antioxidant system of E. californica under salt stress.

Plant hormone biosynthesis and signaling

ABA is a key phytohormone acting as a central mediator of plant adaptive responses to abiotis stress, including salt stress. Upon salt exposure, ABA levels rapidly increase, triggering signaling pathways that orchestrate crucial physiological and molecular adjustments, including stomatal closure to minimize water loss, induction of protective genes encoding osmoprotectants and antioxidant enzymes, modulation of ion transport (particularly Na⁺/K⁺ homeostasis), and alterations in root architecture, etc [69]. In this study, many DEGs were found to be involved in ABA biosynthesis and signaling. NCED is the key enzyme controlling a rate-limiting step in ABA biosynthesis. Increased NCED transcript levels could promote ABA biosynthesis and increase ABA accumulation in plants [70, 71]. Previous studies proved that NCED3 controls endogenous ABA content under stress conditions [72, 73]. In lettuce, NCED4 was reported to be essential for thermo inhibition of seed germination, but not for stress tolerance, and exhibited opposite responses to stress as those of NCED2 and NCED3 [74]. Similar to these results, we found that two NCED3 genes of E. californica were up-regulated, whereas the two NCED4 were down-regulated (Fig. 6b), suggesting that salt stress activated ABA biosynthesis in E. californica, and consequently the ABA signaling pathway (Fig. 6c). Several key components involved in ABA signaling exhibited differential expression. PYL, PP2C, and ABI showed uniformed down- or up-regulated expression (Fig. 6c). However, SAPKs, which encode SnRK2 family protein kinases, exibited an opposite response pattern to salt stress, in which SAPK2 and SAPK 10 were up-regulated, but the SAPK 7s were down-regulated (Fig. 6c). OsSAPK2, 7, and 10 were reported to play positive roles in regulating tolerance salt stress [75–77]. Thus, the opposite responses of SAPK2, 7, and 10 of E. californica to salt stress suggested that they may function differently from those in rice, and therefore deserve further in-depth research.

Auxin plays crucial roles in regulation of multiple aspects of growth and development [78], and therefore participates in regulating plant developmental alterations in response to salt stress. IAA is the predominant endogenous auxin molecule in plant, with tryptophan as the primary precursor. IPyA pathway is among the most important in IAA biosynthesis [79]. This pathway is regulated by YUC (YUCCA) flavin monooxygenases, catalyzing a rate-limiting step converting IPA to IAA [80, 81]. Salt stress usually reduced IAA levels and down-regulates the expression of YUC [80].

In this study, several DEGs related to IAA biosynthesis were identified and were generally up-regulated (Fig. 6h). Of particular note was that excepting for one YUC5 showing very low expression, five YUCs, annotated as YUCCA10, were up-regulated by salt stress (Fig. 6h). Overexpression of Cucumis sativus CsYUC11 in Arabidopsis increased tolerance to salt stress [76]. Expression of YUCCA6 under stress-inducible [82] and constitutive [83] promoters in poplar and potato resulted in higher auxin levels and tolerance to abiotic stresses. Thus, the up-regulation of YUCCA10s suggested that YUCCA10 may function in IAA homeostasis maintenance under salt stress, and positively contribute to salt tolerance. Besides the IAA biosynthesis, we also identified wide range of DEGs encoding regulators involved in IAA signal transduction, including AUX/IAA, ARF, SAUR, etc., exhibiting complex responses to salt stress (Fig. 6i). These results imply potential critical roles of IAA in salt tolerance of E. californica.

Key salt stress responsive metabolism pathways

Some reports indicated that alkaloids act as ROS scavengers under stress conditions. Many alkaloids interfere with the NADPH oxidase pathway and may produce antioxidant or pro-oxidant effects depending on the concentration and conditions [84]. Recently, the sanguinarine was found to enhance heat tolerance in Arabidopsis [85]. It could regulate the chaperone activities and promote production of HSP17.6CCI protein [86]. These studies suggested the potential pivotal roles of alkaloids in plant resilience mechanisms against abiotic challenges.

E. californica is featured by synthesis of a variety of BIAs [87]. Previous studies on the alkaloids of E. californica mainly focused on their synthesis pathways [88–90]. But their roles in responses to the salt stress have been rarely studied. In this study, metabolome analysis found that isoquinoline alkaloids were among the most abundant class of DAMs and were all up-regulated. Most DEGs enriched in isoquinoline alkaloids biosynthesis pathway were positively correlated with DAMs (Fig. 9c). For example, 31 DEG were positively correlated with dihydrosanguinarine, and only five were negatively correlated with it (Fig. S14a). These results highlighted that salt stress significantly activated metabolic flux for isoquinoline alkaloid biosynthesis in E. californica. Although the roles of BIAs in salt stress tolerance remains unclear, the salt responsive and BIAs-correrlated DEGs identified here may be valuable for revealing their relationship.

The α-linoleic acid metabolism (ko00592) is associated with biosynthesis of plant hormone JA, which has been well known as one of the key mediators of defense to biotic stress [91]. However, its functions in salt tolerance are still debated due to the contrasting results and conclusions from different experimental or plant species. External MeJA treatments reduced salt-induced damage in wheat [92], but enhanced salt-triggered growth inhibition and senescence in Arabidopsis [93]. Endogenous JA accumulation and signaling impaired salt tolerance in rice [94]. However, JA deficiency in tomato aggravates oxidative stress and associated damage [95]. GmLOX6-mediated JA biosynthesis enhances soybean resistance to salinity through modulation of ROS homeostasis and Na + transport [96]. These results suggested important but complex roles of JA in plant salt tolerance. We found significantly up-regualated DAMs in the ɑ-linoleic acid metabolism pathway under salt stress. The DEGs in this pathway were mainly positively correlated with DAM (Fig. 9b). The content of JA was 11.38 times (Log2 fold change) higher than that of the CK. Only two out of 44 DEGs were negatively correlated with JA (Fig. S14b). These results showed that ɑ-linoleic acid metabolism (ko00592) was significantly activated by salt stress at transcriptional and metabolic levels. Thus, the roles of JA in salt stress tolerance of E. californica is of particular interests. However, almost nothing is known about JA’s roles in response to salt stress of Papaveraceae species, and is also very few in other abiotic stresses. Recently, MeJA treatment was found to increase antioxidant capacity, phenolic content, and proline concentration under dry conditions [97], suggesting potential positive roles of JA in abiotic stress tolerance. Therefore, the DEGs significantly correlated with JA could be considered as important research object for revealing the relationship between JA and salt stress tolerance of E. californica in the future.

Interestingly, JA is also the signal inducing synthesis of various secondary metabolites [98]. Regulators involved in JA signaling play important roles in JA mediated regulation of secondary metabolite synthesis, such as nicotine synthesis in tobacco [99], monoterpenoid indole alkaloids biosynthesis in Catharanthus roseus [100], as well as artemisinin synthesis in Artemisia annua [101], etc. Therefore, there may be some association between simultaneous salt-induced promotion of JA biosynthesis and signaling and isoquinoline alkaloid biosynthesis, and their contribution to salt stress of E. californica are fascinating and worthy of particular attention and investigation.

Conclusion

In this study, we conducted comprehensive investigations of E. californica responses to salt stress at morphological, physiological, transcriptomic, and metabolomic levels. RNA-seq analysis of non-stressed, mildly stressed, and severely stressed samples revealed a large number of salt-responsive DEGs. These DEGs included regulators of intracellular ion homeostasis, osmotic regulation, oxidative stress and detoxification, plant hormone biosynthesis and signaling, as well as transcription factors. Several key pathways involved in salt stress response were identified. These results provided the insights into the transcriptional regulatory mechanisms of E. californica in response to salt stress. Additionally, metabolites predominantly accumulating under severe stress were identified. Integrated analysis of transcriptome and metabolome data highlighted JA biosynthesis and signaling, and isoquinoline alkaloid biosynthesis, due to their strong salt stress activation patterns. The DEGs in the two pathways, such as AOX, LOX, and JAZs, as well as NCS1, 6OMT, TNMT, and BBEs, etc., could be employed as breakthrough points to uncover the relationship and functional roles of these pathways in salt stress tolerance of E. californica, and also serve as a valuable resource for future breeding salt-tolerant crops.

Authors’ contributions

Z.H. conceived and administrated the Project and acquired the funding; L.J. performed most of the experiments and data analysis; ZH and LL wrote and revised the manuscript; L.S., X.D., B.Y., Y.F., M.Z., Y.W., and X.T. assisted the experiments, reviewed and approved the final manuscript.

Data availability

The transcriptome data is deposited in the GSA database of China National center (www.cncb.ac.cn) for Bioinformation under accession NO. CRA036883, and is publicly available.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.