Molecular cloning and expression analysis of the MaASR1 gene in banana and functional characterization under salt stress

Background: Abscisicacid(ABA)-,stress-andripening-inducedprotein(ASR)isplant-speci ﬁ chydrophilictranscriptional regulatorsinvolvedinsucrosestressandwoundinginbanana.However,itisnotknownwhetherbanana ASR genesconfer saltstresstolerance.Thecontextsofthestudywastoanalysisthesequencecharacterizationofbanana ASR1 ,andidentify its expression patterns and function under salt stress using quantitative real-time PCR (qPCR) and overexpression in Arabidopsis . The purpose was to evaluate the role of banana ASR1 to salt stress tolerance employed by plants. Results: Afull-lengthcDNAisolatedfrombananafruitwasnamed MaASR1 ,andithada432bpopenreadingframe(ORF) encoding143aminoacids. MaASR1 waspreferentialexpressioninrootsandleavescomparedtolowexpressioninfruits, rhizomes and ﬂ owers. Under salt stress, the expression of MaASR1 quickly increased and highest expression level was detected in roots and leaves at 4 h, and then gradually decreased. These results suggested that MaASR1 expression was induced under salt stress. MaASR1 protein was localized in the nucleus and plasma membrane. MaASR1 was transformed to Arabidopsis and veri ﬁ ed by southern and northern analysis, transgenic lines L14 and L38 integrated one and two copies of MaASR1 , respectively, while overexpression in transgenic lines provided evidence for the role of MaASR1 to salt stress tolerance. Conclusions: This study demonstrated that overexpression of MaASR1 in Arabidopsis confers salt stress tolerance by reducing the expression of ABA/stress-responsive genes, but does not affect the expression of the ABA-independent pathway and biosynthesis pathway genes.

Banana (Musa acuminata L.) plays important roles in tropical and subtropical fruit production and agricultural economy. However, banana plant has shallow roots and a permanent green canopy, and is especially sensitive to unfavorable conditions, such as high salt, drought, and cold [21,22,23,24]. Therefore, understanding the molecular mechanisms of the abiotic stress response is necessary for genetic improvement of stress resistance in banana. Although studies in M. acuminata L. A. Colla and Musa balbisiana L. A. Colla have highlighted mAsr members' role in sucrose stress and wounding [21], the expression patterns and functional characterization of ASR genes in M. acuminata L. AAA group, cv. 'Dwarf Cavendish' (a commercially important Cavendish cultivar) under salt stress remain unknown.
In this study, we obtained a full-length ABA-, stress-, and ripening inducible gene named MaASR1 from banana based on a cDNA fragment that originated from a single clone of a forward suppression subtractive hybridization (SSH) cDNA library of banana fruit [8]. We showed that MaASR1 expression was induced in banana plants under salt stress and overexpression of MaASR1 in A. thaliana improves its tolerance to salt stress. These results suggested that MaASR1 plays an important role in salt stress tolerance.

Plant materials
The ex vitro plants of banana (M. acuminata L. AAA group, cv. 'Dwarf Cavendish') (ITC 0002) ('Dwarf Cavendish' as known as 'Brazilian') were obtained from the banana tissue culture center (Institute of Banana and Plantain, Chinese Academy of Tropical Agricultural Sciences, Danzhou, Hainan, China). Ex vitro banana plants were grown at 28°C with 70% humidity, 200 μmol·m -2 ·s -1 light intensity, and 16 h light/8 h dark cycle. Ex vitro plants with uniform growth at the five-leaf stage were selected and twelve were divided into four groups for salt treatment. Banana grown in soil were irrigated with half-strength Hoag land's solution [25] supplemented with 300 mM NaCl for 0 h, 2 h, 4 h, and 6 h. All samples were separately frozen in liquid N 2 and stored at -70°C for RNA extraction and expression analysis.
The wild-type A. thaliana (Columbia ecotype) seeds were purchased from the Arabidopsis Biological Resource Center (ABRC, Ohio University, Columbus, OH, USA). The DH5α Escherichia coli and the LBA4404 Agrobacterium tumefaciens strains were provided by Professor Jiaming Zhang from the Institute of Tropical Bioscience and Biotechnology, Chinese Academy of Tropical Agricultural Sciences. All Arabidopsis seeds were sown on a 1:1:8 mixture (by weight) of vermiculite, perlite, and peat moss, respectively. Arabidopsis plants were grown at 22°C with 70% humidity and 16 h light/8 h dark cycle (Sylvania GRO LUX fluorescent lamps; Utrecht, The Netherlands).

RNA extraction and cDNA synthesis
Total RNA was extracted from the roots, leaves, rhizomes, flowers, as well as fruits of banana, leaves and roots after NaCl treatments using the RNeasy kit (QIAGEN, Valencia, CA) according to the manufacturer's instructions. First strand cDNA was synthesized from 2 μg total RNA from each sample using M-MLV Reverse Transcriptase (Invitrogen, Carlsbad, CA).

Cloning and sequence analysis of MaASR1
The full-length gene encoding MaASR1 was amplified from a banana fruit 2 d after postharvest with the primers (5′-caagcatcccacactcaatac-3′ and 5′-cacaagcacaagatcgagg-3′) based on the cDNA sequence of MaASR1 isolated from a banana fruit cDNA library [8] with the adapter primers Ptr5′ (ctccgagatctggacgagc) and Ptr3′ (taatacgactcactcactataggg). The MaASR1 cDNA sequences were submitted to GenBank using the web-based submission tool "BankIt" from the NCBI home page (http://www.ncbi.nlm.nih.gov/Banklt/index.html). A comparison of the similarity of the full-length cDNA sequence of the MaASR1 gene was performed in the GenBank database using the BLAST program (http://www.ncbi.nlm.nih.gov/BLAST/). Amino acid sequences were compared using the DNAMAN software package (Version 5.2.2, Lynnon Biosoft, Canada). A homology tree was constructed by the neighborjoining method with a Poisson correction model using MEGA 5.05 software (Arizona State University, Tempe, AZ, USA). The number for each interior branch is the percent bootstrap values calculated from 1000 replicates.  Table 1) were used as a loading control to normalize samples in separate tubes. The qPCR was performed in triplicate for each sample using the primers of MaASR1-F and MaASR1-R ( Table 1). The relative expression level of MaASR1 gene was calculated using the 2 -ΔΔCT method [26].
All data were analyzed using IQ5 software in an iQ5 real-time PCR detection system (Bio-Rad, USA).

Subcellular localization of the MaASR1 protein
The cDNA encoding the ORF of MaASR1 was digested with Nco I and Spe I restriction enzymes and inserted into pCAMBIA1304-GFP expression vector to generate a MaASR1-GFP fusion protein under the control of the cauliflower mosaic virus (CaMV) 35S promoter. The recombinant pCAMBIA1304-MaASR1-GFP plasmid was transferred to the A. tumefaciens strain LBA4404 and introduced into Nicotiana benthamiana leaves as previously described by Goodin et al. [27]. After 48 h incubation on MS at 25°C, fluorescence was examined by fluorescence microscopy (LSM700, Carl Zeiss, Germany).

Plant transformation and generation of transgenic plants
The MaASR1 coding region was inserted into the pBI121 vector by replacing the β-glucuronidase following digestion with BamH I and Sac I. The pBI121-MaASR1 was transferred into A. tumefaciens strain LBA 4404. Transgenic Arabidopsis plants were generated using the floral dip-mediated infiltration method [28]. Seeds from T 0 transgenic plants were plated in kanamycin selection medium (50 mg·L -1 ). Homozygous T 3 lines were used for functional investigation of MaASR1.

Blot analyses
Two kanamycin-resistant transgenic lines from the T 3 generation were used to determine the integration of MaASR1 to A. thaliana genome by Southern blotting analysis. Probes from a partial region (389 bp) of the MaASR1 gene for hybridization were prepared from the PCR product by using the primers (5′-ccgaggagaagcaccaccac-3′ and 5′-gccaccgct gcagcgatctcctc-3′) and used in DIG-dUTP according to the manufacturer's instructions (Roche Applied Science, Mannheim, Germany). Northern blotting was performed according to the manufacturer's instructions (Roche Applied Science, Mannheim, Germany). Probe of northern blotting was labeled using a random primer labeling system (Cat.1093657, Roche Applied Science, Mannheim, Germany). After hybridization, the membrane was washed and exposed to X-ray film (Kodak BioMax MS system) according to the methods of Miao et al. [29].

Salt stress and ABA treatment in wild-type and transgenic plants
For salt stress tolerance analysis, 4 week old plants were irrigated with 300 mM NaCl and survival rates were assessed after 15 d. For expression analysis of ABA/stress-responsive genes in wild-type and transgenic plants, 15 d old seedlings were transferred to 1/2 MS agar plates supplemented with 300 mM NaCl for 12 h or 100 μM ABA for 6 h.
The expression patterns of three ABA/stress-responsive genes (RD29a, RD29b, and RAB18), one ABA-independent pathway gene (DREB2A), one upstream element of ABA signaling pathway (ABI1), and an ABA biosynthesis rate-limiting enzyme gene (AAO3) in the leaves of MaASR1 overexpressing transgenic plants and wild-type A. thaliana after NaCl or ABA treatments were detected by qPCR using corresponding primers (Table 1) and the AtActin as a control. The amplification program consisted of one cycle of 95°C for 1 min, followed by 40 cycles of 95°C for 10 s, 55°C-58°C for 15 s, and 72°C for 30 s. The expression levels of these genes were verified in triplicate and calculated using the 2 -ΔΔCT method [26].

Isolation and sequence analysis of banana MaASR1
A full-length ASR gene was obtained from the banana fruit and designated as MaASR1. The gene was deposited in GenBank under the Accession number AAT35818. Sequence analysis revealed that the full-length MaASR1 cDNA has a 432 bp open reading frame (ORF) (Phred scores N 20) that encodes 143 amino acids. The deduced amino acid sequence of MaASR1 contained the conserved N-terminal DNA binding site and a putative nuclear C-terminal localization signal  ( Fig. 1), and it shared 62 and 66% identities with rice (OsASR1, AAB96681) and maize (ZmASR1, CAA72998), respectively (Fig. 2).

Differential expression of MaASR1 in various banana tissues
There were significant differences in the MaASR1 expression in different banana tissues. MaASR1 expression was detected in roots, leaves, rhizomes, flowers, and fruits. The roots showed the highest gene expression level, together with leaves, fruits, and flowers; the lowest level was found in rhizomes. The MaASR1 expression level in roots was approximately 11-fold higher than that in rhizomes (Fig. 3a).

Phenotype and expression analysis of MaASR1 in banana plants under salt stress
The phenotype and expression of MaASR1 in banana plants at different times under salt treatments were examined to determine the transcriptional response of MaASR1 to salt stress. The results showed that banana roots and leaves exhibited different phenotypes at different times under salt stress. At 6 h under salt stress, the banana roots were black and leaves exhibited obvious brown spots (Fig. 3b). Significant differences in MaASR1 expression were detected in the roots and leaves under salt stress. The expression levels of MaASR1 quickly increased and reached its maximum levels in roots and leaves at 4 h. The expression in roots was approximately 3-fold higher than that in leaves at 4 h, and then gradually decreased over time (Fig. 3c). These results indicated that MaASR1 expression was obviously induced in banana roots and leaves while roots may be more sensitive to salt stress.

MaASR1 localizes to the nucleus and plasma membrane
To determine the subcellular localization of the MaASR1 protein, its ORF was introduced into the pCAMBIA1304-GFP vector upstream of the GFP gene to create a MaASR1-GFP translational fusion construct. The recombinant pCAMBIA1304-MaASR-GFP fusion was infiltrated into the leaves of N. benthamiana. We observed that the green fluorescence MaASR1-GFP was confined to the nucleus and plasma membrane (Fig. 4). These results indicated that MaASR1 is targeted to the nucleus and plasma membrane. transformation of Arabidopsis, two kanamycin-resistant transgenic lines from the T 3 generation were obtained. The copy number of these two transgenic lines was investigated by Southern blotting analysis. These results showed that the L14 line integrated two copies of MaASR1, while the L38 line integrated one copy of MaASR1 (Fig. 5a). Northern analysis confirmed that the MaASR1 transcripts were present in the leaf tissue of two transgenic lines compared to that no expression was detected in wild-type plants (Fig. 5b).

Overexpression of MaASR1 enhances tolerance to salt stress
When mature Arabidopsis plants were subjected to 300 mM NaCl treatment for 15 d, the transgenic plants exhibited better growth and a higher survival rate than that of the wild-type (Fig. 5c and Fig. 5d), where the survival rate of wild-type, L14, and L38 was 10.7, 87.3, and 82.7%, respectively (Fig. 5d). These results showed that overexpressing of MaASR1 in Arabidopsis plants were more tolerant to salt stress than the wild-type.

Overexpression of MaASR1 decreases the expression of ABA/stressresponsive genes by NaCl treatment
To improve our understanding of MaASR1 function during salt stress tolerance, the expression of several ABA/stress-responsive genes was examined in the wild-type plants and the MaASR1 overexpressing transgenic plants (Fig. 6). Without the addition of NaCl, no significant difference was observed in the transcription of tested genes (RD29a, RD29b, RAB18, DREB2A, ABI1, and AAO3) in the MaASR1 overexpressing transgenic plants compared to wild-type plants. Under salt stress, however, transgenic plants exposed to 6 h or 12 h of salt treatment exhibited reduced expression of RD29A, RD29b, and RAB18 compared to wild-type plants that were similarly treated (Fig. 6). The expression of DREB2A, ABI1, and AAO3 revealed similar trends at 6 h between wild-type and transgenic plants under salt stress (Fig. 6). This result indicated that MaASR1 overexpression led to the down-regulation of ABA/stress-responsive genes under salt stress conditions, but didn't affect the expressions of ABA-independent pathway and biosynthetic pathways genes.

Overexpression of MaASR1 enhances the response of plants to ABA by exogenous ABA treatment
Under exogenous ABA treatment, the expressions of several ABA/stress-responsive genes, such as RD29a, RD29b, RAB18, and ABI1, were obviously increased in MaASR1 overexpressing transgenic plants compared to wild-type plants that were similarly treated (Fig. 7). However, the expression of ABA-independent pathway gene DREB2A and ABA biosynthetic pathway gene AAO3 revealed similar trends between wild-type and transgenic plants by ABA treatment (Fig. 7). This result suggested that overexpression of MaASR1 might enhance the response of plants to ABA/stress signal pathway, but was no involved in ABA-independent and ABA biosynthetic pathways.

Discussion
In this study, MaASR1 was identified in banana. MaASR1 contains an ORF encoding 143 amino acids and two highly conserved regions, including Zn 2+ -DNA binding sites at the N-terminus and a nuclear localization signal at the C-terminus (Fig. 1), whose structure was similar to ASR genes from lily [15] and tomato [18], and therefore was characterized as a potential ASR family member. Compared with other banana cultivars, the MaASR1 from the M. acuminata L. AAA group, cv. 'Dwarf Cavendish' (accession no. AAT35818) shared 97 and 86% similarity with Asr amino acid sequences from the M. acuminata L. AAA group cultivars 'Mbwazirume' (accession no. ACZ60129) and 'Williams' (accession no. ACZ50751). Amino acid sequence differences may be because these ASR genes are from different banana cultivars or different ASR family members. Although another mAsr1 (accession no. ACZ60119) was reported in different banana (M. acuminata subsp. burmannicoides) cultivar [21], amino acid sequence differences exist in the N-terminus, which suggests that MaASR1 is different from mAsr1. Compared with other species, MaASR1 shares higher similarities with rice OsASR1 (accession no. AAB96681) and maize ZmASR1 (accession no. CAA72998) (Fig. 2), which indicates that MaASR1 might be a relatively conserved gene.
Distinct ASR family members exhibit variable responses to abiotic stress [21,30]. In tomato, Asr1 and Asr2 are members of the family preferentially induced by desiccation in leaves; Asr2 is the only one activated in the roots from water-deficit-stressed plants [30]. Wheat TaASR1 transcript levels increase after treatments with PEG6000, ABA, and H 2 O 2 [16]. The expression of lily LLA23 is induced following the application of ABA, NaCl, or dehydration [15]. LcAsr was expressed in postharvest uncovered litchi fruit [10]. In banana meristems, mAsr1 and mAsr3 were induced by sucrose stress and wounding, while mAsr3 and mAsr4 were induced by exposure to ABA [21]. In this study, the expression of MaASR1 was induced by salt stress (Fig. 3b and Fig. 3c), consistent with that reported for the tomato Asr1 [18,31], but we found that the expression level of MaASR1 in roots was approximately 3-fold higher than in leaves at 4 h under salt treatment (Fig. 3c). These results indicated that the expression of MaASR1 might be induced in leaves and roots by salt stress but banana roots might be more sensitive to salt stress. Different ASR proteins' subcellular distribution patterns were observed in tomato [5], litchi [10], wheat [16], and lily [15]. The ASR1 from tomato was first reported as a nuclear protein [5]. The result supported the fact that most ASR proteins, such as lily [15], litchi [10], and wheat [16] were found in the nucleus. However, Kalifa et al. [18] reported that tomato ASR1 was localized in both the cytosol and the nucleus. During the early stages of pollen maturation, the ASR protein from lily translocates from the cytosol into the nucleus [14]. In this study, we demonstrated that the MaASR1 protein localized to the nucleus and plasma membrane (Fig. 1c), suggesting that it acts as a part of a transcription-regulating complex. However, further experiments are needed to determine the specific functions and cellular mechanisms of MaASR1 in the nucleus and plasma membrane.
ASRs are involved in abiotic stress tolerance [9,15,16,18]. In our study, to investigate the function of MaASR1, two MaASR1 overexpressing Arabidopsis transgenic lines were generated and confirmed by Southern blot and northern blot ( Fig. 5a and Fig. 5b). Under salt stress, the MaASR1 transgenic plants exhibited increased tolerance to salt stress compared to the wild-types (Fig. 5c). The survival rates of MaASR1 overexpressing transgenic plants were higher than that in the wild-type (Fig. 5d). These results demonstrated that overexpression of MaASR1 genes confers salt stress tolerance to transgenic plants by enhancing the survival rates. The finding will have important theoretical and practical significance for improving the ABA-dependent signal transduction pathways play a crucial role in the adaptation of plants to abiotic stress [32]. The expression of several ABA/stress-responsive marker genes, including RD29a, RD29b, and RAB18D, are induced by abiotic stresses [24,33,34]. DREB2A is a DRE/CRT-binding transcription factor; its expression is not induced by ABA and abiotic stresses [35]. The ABI1 and AAO3 belonged to the upstream element of the ABA signaling pathway and ABA biosynthesis rate-limiting enzyme, respectively [33]. Under water stress, some ABA/stress-responsive genes, such as RD29a, RD29b, and RAB18, were up-regulated in wild-type Arabidopsis plants, indicating that the injury which resulted from water stress induces the expression of ABA/stress-responsive genes [36]. We examined the expression of ABA/stress-responsive genes (RD29a, RD29b, and RAB18), DREB2A, ABI1 and AAO3 in MaASR1 overexpressing transgenic plants and wild-type plants. The expression of DREB2A, ABI1 and AAO3 presented similar trends between wild-type and transgenic lines under salt stress; however, the ABA/stress-responsive genes (RD29a, RD29b, and RAB18) were down-regulated in the transgenic plants subjected to salt treatments in comparison to similarly treated wild-type plants (Fig. 6).
This result suggests that the MaASR1 overexpressing transgenic plants might be involved in enhancing salt stress tolerance through reducing expression of the ABA/stress-responsive genes, but didn't affect the expression of ABA-independent pathway gene, the upstream element of ABA signaling pathway gene, and an ABA biosynthesis pathway gene.
ABA plays important regulatory roles in plant growth, development [37], and fruit ripening [38], particularly in the ability to respond to various unfavorable environmental stresses, including drought, salt [39], and cold [40]. In lily, constitutive expression of LLA23 in transgenic plants significantly reduced ABA sensitivity and enhanced drought and salt resistance [15]. The constitutive overexpression of OsASR1 also involved ABA signaling and increased high salinity stress tolerance in rice [19]. However, it is not known whether overexpression of MaASR1 is involved in ABA signaling and enhances salt stress tolerance. In this study, we compared the expression of RD29a, RD29b, RAB18D, DREB2A, ABI1, and AAO3 in MaASR1 transgenic plants and wild-type by ABA treatment. DREB2A and AAO3 exhibited similar trends between wild-type and transgenic plants under ABA treatment. However, the expression levels of RD29a, RD29b, RAB18D, and ABI1 were higher in MaASR1 overexpressing transgenic plants than that in the wild-type (Fig. 7). These results suggest that MaASR1 overexpression is likely to involve ABA signaling and enhances the salt stress tolerance by altering the expression of the ABA/stress-responsive genes (RD29a, RD29b, and RAB18D) and the upstream element of ABA signaling pathway (ABI1).

Concluding remarks
A full-length cDNA of MaASR1 (accession number: AAT35818) was obtained from banana (M. acuminata L. AAA group, cv. 'Dwarf Cavendish') and it was obviously induced under salt stress. MaASR1 overexpression resulted in enhanced tolerance to salt stress by reducing expression of the ABA/stress-responsive genes, but didn't affect the expression of ABA-independent pathway and biosynthesis pathway genes. Further studies are required to identify the direct target genes of MaASR1 using Chromatin Immunoprecipitation (ChIP). This will enhance our understanding of the molecular interaction mechanisms of MaASR1 in enhancing salt stress tolerance, improving the adaptability of plants to salt stress.

Conflict of interest
All the authors do not have any possible conflicts of interest.