The Ethylene Response factor SmERF1 Improves Salinity Tolerance and Impacts Seed Size in Tobacco

Background Ethylene response factor (ERF) proteins play vital roles in plant resistance and plant development. However, little is known about the ERF transcription factors of Salvia miltiorrhiza, which is a famous Chinese herb with good resistance to stress. Result Screened from our previous transcriptome data, SmERF1, an ERF transcript factor, was isolated from S. miltiorrhiza. SmERF1 had a single AP domain and was classied in the ERF E3 subfamily. SmERF1-expressing tobacco plants showed slower growth, less biomass and a decrease in chlorophyll only at the seedling stage, and there was no signicant difference in other growth stages. In addition, seeds of tobacco plants with SmERF1 expression were smaller and lighter than those of wild plants, similar to some AP2 TFs. Under NaCl treatment, transgenic tobacco lines showed better tolerance to salinity, and the proline content, SOD and POD activities of transgenic lines were higher than those of wild-type plants, while MDA content was lower than that of wild-type plants. Furthermore, we determined the phytohormones related to plant resistance and showed that transgenic tobacco plants had higher ABA levels and lower GA levels than wild tobacco. The expression of SmERF1 regulated the expression of key enzyme genes related to plant hormone biosynthesis, such as NtSDR, NtGA20ox, NtACO and NtACS. Conclusions The affect plant growth at seedling stage, and increase plant tolerance salt. And the cause tobacco seeds smaller and lighter. Our study suggested that SmERF1 enhanced tobacco tolerance to salt and impacted seed size through an ABA-dependent

Phylogenetic analysis revealed that SmERF1 was classi ed in the B3 subfamily, together with NtERF5 and AaERF1 (Fig. 1B).
ExpressingSmERF1in tobacco resulted in smaller tobacco seed size Young leaves from transgenic plants selected with Basta were used to extract DNA and RNA. After detecting the CaMV35S promoter sequence in transgenic plants, 15 transgenic tobacco lines were con rmed (Fig S1A). Through qPCR veri cation, four lines (OE-3, OE-4, OE-7 and OE-10) with high expression levels of SmERF1 were obtained and used for follow-up experiments (Fig S1B). There were some differences only at the seedling stage, and there was no signi cant difference in other growth stages between transgenic and wild plants (see results in 3.4).
Furthermore, we harvested T2 generation seeds from four tobacco transgenic lines of SmERF1. We found that the seed size of each tobacco transgenic line was smaller than that of the wild-type plants (WT) ( Fig. 2A). According to the statistical analysis, the thousand kernel seed weights of OE-3, OE-4, OE-6, OE-7 and OE-10 were 59%, 63%, 55%, 56% and 58% of the WT, respectively (Fig. 2B).
Expression ofSmERF1in tobacco affected young tobacco seedling growth During the germination stage, white buds appeared on the WT seeds of tobacco at 7 days after sowing, while the buds on tobacco transgenic seeds (T1) appeared later at 9 days. Two weeks after sowing, the growth rate of transgenic tobacco was signi cantly slower than that of wild plants (Fig. 3A). The biomass of transgenic tobacco seedlings was 48% ~ 0.53% (Fig. 3B). In addition, we observed that the leaves of transgenic tobacco seedlings were yellower than those of wild-type seedlings (Fig. 3A). The chlorophyll content of transgenic seedlings was 82%-84% of that of the WT (Fig. 3D); however, there was no signi cant difference in the germination rates between transgenic and WT seeds under normal conditions (Fig. 3C).

ExpressingSmERF1in tobacco enhances salt tolerance
To detect tolerance to salt, two-month-old plantlets of T2 generation transgenic and WT plants were treated with 400 mM NaCl. Leaves of all wild-type plants wilted and rolled up, and some old leaves turned yellow, while leaves of transgenic lines showed better status than that of WT, and only old leaves rolled up (Fig. 4A).
The physiological parameters, including SOD, POD, MDA content, proline content, and transgenic and WT tobacco plants, were compared under both normal and 400 mM NaCl conditions (Fig. 4B). (1) After NaCl treatment, the POD activity of transgenic tobacco plants was signi cantly higher than that of WT plants.
Under normal conditions, the POD activities of transgenic tobacco plants were higher than those of WT plants, but there was no signi cant difference. (2) After NaCl treatment, the transgenic tobacco plants also showed markedly higher SOD activity than that of WT plants. However, under normal conditions, there were no signi cant differences. (3) The MDA content of the transgenic tobacco plants was signi cantly lower than that of WT under both conditions. (4) There was no signi cant difference in proline content between transgenic tobacco plants and WT under normal conditions, while proline content was 1.35-1.48 times higher in transgenic tobacco plants than WT after NaCl treatment.
ExpressingSmERF1in tobacco regulates ABA and GA biosynthesis Abscisic acid (ABA) and gibberellins (GA) are a pair of classic phytohormones that antagonistically mediate several plant developmental processes, including seed maturation, seed dormancy and germination, primary root growth, and owering time control (Shu et al., 2018). We found that the ABA contents of transgenic tobacco lines were 1.53-2.51 times higher than those of WT ( Fig. 4-1c), and the GA contents of transgenic tobacco lines were 0.60-0.74 times lower than those of WT ( Fig. 4-1d). The results from qRT-PCR showed that the expression of NtSDR (short-chain dehydrogenase/reductase, GenBank Accession No. AJ223177.1) was activated in SmERF1-overexpressing tobacco, which encodes a key enzyme of ABA biosynthesis. NtGA20ox (GenBank Accession No. JQ413251), a key enzymeencoding gene of GA biosynthesis, was repressed in SmERF1-overexpressing tobacco. The study showed that the SmERF1-overexpressing seedlings had higher ABA content and lower GA content than WT plants (Fig. 4b), indicating that SmERF1 can regulate the plant hormone response of transgenic tobacco seedlings to drought through the ABA pathway.
Moreover, we also tested the expression of NtACS (GenBank Accession No. NM_001326220.1) and NtACO (GenBank Accession No. NM_001325967), two key enzyme-encoding genes in plant hormone ethylene biosynthesis. Our qRT-PCR analysis showed that the transcription of both NtACS and NtACO increased in transgenic tobacco lines, which suggested acceleration of ET production.

Discussion
Over 10% of traditional Chinese patent medicines and simple preparations contain S. miltiorrhiza. The annual consumption of S. miltiorrhiza has exceeded 16 million kilograms in China [18]. More and more genes involved in S. miltiorrhiza resistance have been studied. For example, genes of three multigene universal stress proteins (SmUSPs) were cloned, and their expression enhanced Escherichia coli tolerance to salt and heat stress [29]. Overexpression of SmSnRK2.6 ( sucrose non-fermenting-1-related protein kinase 2) improved S. miltiorrhiza tolerance to abiotic stresses [30]. Overexpression of SmLEA (late embryogenesis abundant proteins) improves drought and salinity tolerance in S. miltiorrhiza [31]. Ectopic expression of tomato prosystemin (LePS), Arabidopsis DREB1A/CBF3 and AtDREB1A was also used to improve S. miltiorrhiza resistance against biotic and abiotic stress separately [32][33][34].
ERF proteins play vital roles in a variety of stress responses in plants. Some ERF genes, such as NtERF5, TaERF3 [35],GmERF3 [36],GmERF113 [37] and tomato JERF1 [12], were thought to be candidates to improve crop-plant resistance because their overexpression enhances resistance to multiple diseases and improves tolerance to drought, salt, and freezing in transgenic plants [1]. To date, 79 ERFs have been identi ed from the S. miltiorrhiza genome [19] and the functions of four SmERFs in regulation of active ingredient biosynthesis have been investigated [20,21,23]. Few reports have described the resistance functions of S. miltiorrhiza ERFs in plants. In our study, the growth of transgenic tobacco lines with SmERF1 expression was signi cantly better than that of wild-type plants, and the proline content, SOD and POD activities of transgenic lines were higher than those of wild-type plants, while the MDA content was lower than that of wild-type plants. These results implied that SmERF1 expression could enhance plant salt tolerance under salt stress. Some ERFs were shown to regulate hormone levels in plants. For instance, apple MdERF2 can negatively affect ethylene biosynthesis by suppressing MdACS1 transcription [15]. TSRF1 activates the expression of a putative rice abscisic acid (ABA) synthesis gene, SDR, resulting in enhanced ABA levels [12]. Peach PpERF3 regulates ABA biosynthesis by activating PpNCED2/3 transcription [13]. ORA47 of Arabidopsis thaliana can activate ABA biosynthesis genes (NCED3 and NCED9) [14]. Overexpression of AtERF11 resulted in elevated bioactive GA levels by upregulating the expression of GA3ox1 and GA20ox genes [16]. ABA, GA and ET, are well known to play major roles in mediating plant defense responses against biotic and abiotic stresses [17]. Our qPCR results show that the expression levels of NtSDR, a key gene of ABA biosynthesis, was upregulated in SmERF1-expressing tobacco, while the expression of NtGA20ox, a key gene of GA biosynthesis was downregulated. In the SmERF1-expressing tobacco seedlings, the expression of NtACS and NtACO, two key enzyme-encoding genes in the ethylene biosynthesis, were upregulated compared with wild-type plants. It was also implied that the expression of SmERF1 many also cause an increase of ethylene. Our further study also indicated that SmERF1-expressing tobacco seedlings had higher ABA content and lower GA content than wild tobacco. There was antagonism between ABA and GA [38], and ABA could also induce ET biosynthesis [39]. Therefore, we inferred that SmERF1 can regulate the plant hormones of transgenic tobacco seedlings in response to drought through the ABA pathway. ABA is believed to be a general inhibitor of plant growth during the early development of seedlings [40]. It is indicated that the higher ABA content resulted in slower growth and decreased biomass in SmERF1-expressing tobacco seedlings.
Seed size is the most important agronomic traits in crop domestication [41], and it is desirable to increase seed yield because grains represent signi cant sources of food. Many transcription factors have been identi ed as regulators on controlling seed formation. AtMYB56, encoding an R2R3 MYB transcription factor, positively regulates Arabidopsis seed size by coordinately controlling the expansion of endothelium layer and proliferation [42]. ARF2 (AUXIN RESPONSE FACTOR 2) and NGAL2 can negative regulate seed size by suppressing cell proliferation [43,44]. The AP2-type transcription factors Aechmea fasciata AfAP2-2, rice SMALL ORGAN SIZE1 (SMOS1) and Larix kaempferi LkAP2L2 also affect seed size [45][46][47]. Furthermore, SMOS1 may integrate auxin and BR signaling to control rice grain size [46]. In the current research, SmERF1 only contains one AP2 DNA binding domain belonging to the ERF subfamily, differing from members of the AP2 subfamily containing two AP2 DNA binding domains. However, expressing SmERF1 in tobacco also causes seeds to be smaller and lighter than wild-type tobacco, similar to AfAP2-2 [45].
Plant hormones also respond during seed development. ABA play a negative role in seed morphogenesis. ABA suppresses seed development by negatively regulating ABI5(ABA response factor), while ABAde cient mutants (abi5) can produce larger and heavier seeds for the increase of embryo cell number and endosperm proliferation [48]. GAs also play important roles in the developing embryo and endosperm in seeds [49]. Loss-of-function mutation in OsGA20ox2 (a key GA biosynthesis gene) resulted in lower GA levels, and delayed seed morphogenesis and maturation in rice [50]. Overexpression of GASA4 (a member of gibberellic acid-stimulated Arabidopsis family), which is expressed in response to GAs, result in increase of seed size and total seed weight, while the gasa4 mutant has smaller seeds [51]. Our study also indicated that SmERF1-expressing seedlings had higher ABA content and lower GA content than WT plants. We inferred that SmERF1 may regulate the plant hormones of transgenic tobacco seedlings, resulting in smaller seed sizes. Crosstalk exists between ABA and GA, and these two phytohormones antagonistically mediate plant developmental processes [52]. Decreasing the gibberellin/abscisic acid (GA/ABA) ratio can weak seed repression of soybean seed germination [53]. The delayed germination of seeds, slower growth and loss of mass of SmERF1-expressing tobacco seedlings all veri ed the change in GA/ABA ratio. Additionally, ET participates in integration with ABA and Gas [49]. Overexpression of ACC deaminase resulted in reduced levels of ET, GAs and IAAs, and smaller seeds in Brassica napus [54]. In the SmERF1-expressing seedlings, the increase in ET may also contribute to sharp seed shape.

Conclusion
S. miltiorrhiza is a famous medicinal plant in Chian. We isolated one ERF transcription factor, SmERF1, from S. miltiorrhiza, and expressed SmERF1 in tobacco. The SmERF1-expression in tobacco affect plant growth at seedling stage, and increase plant tolerance to salt. And the expression of SmERF1 cause tobacco seeds smaller and lighter. The ABA level were increased and GA levels were decreased in SmERF1-expression tobacco plant. Our study suggested that SmERF1 enhanced tobacco tolerance to salt and impacted seed size through an ABA-dependent pathway.

Clone and analysis of SmERF1
According to the unigene (unigene-25151) sequence from our transcriptome data [28], we isolated the cDNA sequences of SmERF1 from S. miltiorrhiza with accession number: KC405081.1. Multiple sequence alignments were performed using ClustalW2 (https://www.ebi.ac.uk/Tools/msa/clustalw2/), and the results were rendered using the Boxshade Server. The conserved domain of SmERF1 was searched using the Conserved Domain Search Service (https://www.ncbi.nlm.nih.gov/guide/domains-structures/). The subcellular location of SmERF1 protein was predicted with WoLF PSORT (https://wolfpsort.hgc.jp/). The phylogenetic tree was constructed with MEGA X software using a Maximum Likelihood (ML) method based on an JTT model, and 500 bootstrap test replicates were used during the construction with other parameters as default [55].

Plant Materials
Mature seeds of S. miltiorrhiza Bunge were surface sterilized as described previously (Yan and Wang, 2007) and cultured on MS basal media (Murashige and Skoog, 1962) for germination. One-month-old seedlings were used for RNA isolation.

Extraction Of Rna, Cdna Synthesis
Genomic DNA was isolated from young leaves of 2-month-old S. miltiorrhiza seedlings or tobacco seedlings with the Biospin Plant Genomic DNA Extraction Kit (BioFlux, China). Total RNA was extracted from transgenic tobacco and wild type plants, using BIOZOL reagent (BIOER, Hangzhou) according to the manufacturer's instructions. The quality and concentrations of genomic DNA and RNAs were determined by 1.0% agarose gel electrophoresis and analysis on a spectrophotometer (SHIMADZU UV-2450, Japan).
cDNA was synthesized with Revert Aid First Strand cDNA Synthesis Kit (Takara, Dalian).

Plant Transformation And Transgenic Plant Veri cation
We cloned the SmERF1 sequence with the attB recombination sites and inserted into pDONOR221 via the Gateway BP clonase (Invitrogen Corporation) reaction to create entry clones. Then, the sequence was moved into the destination vector pEarleyGate100 using Gateway LR clonase (Invitrogen Corporation) to create the expression vector (pO-ERF). The new constructs of the pO-ERF vector were transformed into Agrobacterium tumefaciens strain GV3101 via a freeze-thaw method [56]. Then, the transient transformation of tobacco leaves was accomplished according to Sparkes et al. [57]. The expression vector pO-ERF contained a Basta resistance selection marker; therefore, transformants were selected on 1/2 strength MS medium supplemented with IBA 0.5 mg/L and Basta 10 mg·mL − 1 .
To determine whether the recombinant vector was inserted into the plant, we detected the CaMV35S promoter sequence in all transgenic tobacco lines (T0, T1 generation) by qPCR, primers were shown in Table 1.  Transgenic and wild-type tobacco lines with good growth for approximately 6 weeks in tissue culture were transplanted to the arti cial culture room after seedling re ning. The culture temperature was 24 °C, the humidity was 60%, and the light/darkness was 14/10 h.

Expression analyses of SmERF1 in transgenic tobacco
After selective culture, the transgenic plants were transferred to MS culture. Approximately 6 weeks after subculture, young seedlings (T0 generation) were transplanted into humus soil. Seeds (T1 and T2 generation) were collected when they were mature. The sizes of T2 seeds were observed with a stereomicroscope. The thousand seed weights of tobacco transgenic lines were also evaluated.
Detection of ABA and GA levels in transgenic plants of tobacco (T2) Whole plant samples were collected from each transgenic plant. The contents of ABA and GA were detected according to the manuals of ELISA kits from ShaanXi Maiyuan Biotechnology Co., Ltd.

Detection of chlorophyll content of transgenic plants of tobacco (T2)
Two-week seedlings after germination of T2 generation seeds were used to calculate biomass and detecte the chlorophyll content. Chlorophyll (Chl) content was measured according to the method described by Frank et al. with slight modi cation [58]. Brie y, 200 mg of fresh seedlings was collected and ground with sterile mortar and pestle. The samples were transferred into 10 mL tubes, and then 10.0 mL of 80% acetone was added. After mixing well, the tubes were strati ed at 4 °C for 24 h. The samples were centrifuged at 12000 g at 4 °C for 3 min, and then the OD value of the supernatant was assayed with a spectrophotometer (

Resistance Analyses Of Transgenic Tobacco
The seeds of the T1 generation of transgenic and wild-type tobacco were germinated on lter paper. When the seedlings grew to approximately 1 cm in size, they were transplanted into the sowing and seedling substrate, one in each pot, cultivated in the arti cial culture room, and watered once every 3 days. Two months after transplantation into the soil, the tobacco seedlings were watered with 400 mmol/L NaCl solution every other day. After 7 days of treatment, the growth states of the tobacco lines were observed, and the aerial parts of these seedlings were quickly frozen with liquid nitrogen for the detection of physiological indexes as follows.
Quantitative Real-time Pcr (qrt-pcr) The quantitative reactions were performed on an IQ5 real-time PCR detection system (Bio-Rad), using SYBR® Premix Ex Taq™ (Perfect Real Time) (TaKaRa). PCR ampli cations included the following conditions: 50 °C for 2 min and 95 °C for 30 s, followed by 40 cycles of 95 °C for 10 s and 60 °C for 30 s. Gene-speci c primers for qPCR are listed in Table 1. Nicotiana tabacum actin (Accession No. U60495.1) was used as an internal control. Their relative expression was calculated via the 2 −ΔΔCt method [62]. Each sample was collected from ve independent plants. Three independent biological replicates were used for each plant sample. Each data point represents the average of three experiments.

Data analysis
Data were analyzed using one-way analysis of variance (ANOVA) followed by the Turkey-Kramer test (P < 0.05) using SPSS software (version 19.0). Levels of statistical signi cance were marked with *P < 0.05, **P < 0.01 and ***P < 0.001.

Consent for publication
Not applicable.

Availability of data and material
All data generated or analyzed during this study are included in this published article and its supplementary information les. data.
All authors read and approved the nal manuscript.