Identification of AtSM34, a novel tonoplast intrinsic protein-interacting polypeptide expressed in response to osmotic stress in germinating seedlings

Aquaporins are implicated in a wide variety of plant physiological processes, although the mechanisms involved in their regulation are not fully understood. To gain further insight into the regulatory factors involved in this process, we used a yeast two-hybrid system to screen for potential binding partners to the Arabidopsis tonoplast intrinsic protein (TIP) AtTIP1;1. This was the first protein identified to be associated with high water permeability in vacuolar membranes from Arabidopsis thaliana. Using AtTIP1;1 as bait, a novel binding protein was identified in both yeast and plant cells. This prey protein, named AtSM34, was a 309 aa polypeptide with a predicted molecular mass of 34 kD and contained a single MYB/SANT-like domain. AtSM34 promoter:: GUS histochemical staining analysis detected AtSM34 expression in flowers, stems and leaves, particularly in the vascular tissues, in response to osmotic stress. AtSM34 expression was localized in the endoplasmic reticulum membrane, and sequence deletion analysis revealed that the N-terminal coding region (amino acids 1-83) was critical for this localization. Overexpression of AtSM34 resulted in hypersensitivity to exogenous mannitol, sorbitol and abscisic acid, and caused a significant delay in germination. AtSM34 interacted with AtTIP1;2 and AtTIP2;1, which are essential proteins for modulation of tonoplast permeability and highly expressed in germinating seedlings. These data indicate AtSM34 is a novel TIPs binding protein involved in the osmotic stress response of seedlings at an early stage of development.

Aquaporins are channel proteins involved in the regulation of water permeability of cellular membranes [1] and are actively involved in plant growth and development. Regulation of these proteins and their integrated functions is important for the maintenance of water balance under conditions of stress [2].
Aquaporin activity is regulated by factors that include phosphorylation, heteromerization, pH, Ca 2+ , temperature and solute gradients, which modify their gating behavior [3][4][5][6]. Recent reports have suggested other membrane-associated or cytosolic proteins might be involved in the regulation of aquaporin expression, activity and trafficking or might function as multicomponent protein complexes through protein-protein interactions. There are a number of examples of such interactions in animals and plants. These include the Ca 2+ -dependent binding of two calmodulin molecules with a single aquaporin-0 tetramer, which results in the temporal regulation of channel activity [7], and the interaction of heat shock protein 70 (Hsp70) in the regulation of aquaporin-2-mediated trafficking in rat kidney cells [8]. In plants, the Cucumber mosaic virus (CMV) 1a interacts both with TIP1 and TIP2, which possibly affects CMV replication via interactions in the tonoplasts [9]. Further-more, soybean cytosolic glutamine synthetase interacts with the C-terminal domain of the symbiosome membrane nodulin 26 (NOD26), which is thought to be localized at the cytosolic side of the symbiosome membrane and to promote efficient assimilation of fixed nitrogen and prevent ammonia toxicity [10]. These protein interactions also regulate the amount and localization of aquaporins in response to environmental stress factors. Overexpression of hot pepper Rma1H1, a homolog of E3 ubiquitin ligase, resulted in reduction of the level of AtPIP2;1 and inhibited trafficking of AtPIP2;1 from the endoplasmic reticulum (ER) to the plasma membrane in protoplasts. Downregulation of plasma membrane aquaporin levels conferred tolerance to dehydration in transgenic Arabidopsis plants [11].
In plants, vacuoles are unique and essential organelles with important roles in turgor regulation, osmotic adjustment, cell signaling, storage and digestion. Tonoplast intrinsic proteins (TIPs) were identified as a tonoplast-localized water channel subgroup [12]. Variations in TIP expression levels might influence water movement across the tonoplast and might also affect the responses of plants to abiotic stress [13]. The effects of osmotic stress were investigated in ice plants (Mesembryanthemum crystallinum). Mannitol-induced water imbalance increased expression of McTIP1;2 (McMIPF) proteins and resulted in its redistribution to the endosomal compartment [14], which implicates these changes in response to osmotic stress.
AtTIP1;1 belongs to the Arabidopsis thaliana TIP subfamily and was one of the first proteins identified with high water permeability when expressed in Xenopus oocytes [15]. However, AtTIP1;1 knock-out lines or a double knockout for both TIP1;1 and TIP1;2 in Arabidopsis showed no apparent alterations in macroscopic phenotype. This is likely to be because of the functional redundancy of other TIP homologs [16,17]. These findings indicated undetected functions or regulatory mechanisms are associated with AtTIP1; 1. This study focused on the identification of AtTIP1;1 interacting proteins with the aim of gaining a better understanding of the function of AtTIP1;1. Using AtTIP1;1 as bait in a split ubiquitin yeast screening system, a novel interacting protein was identified and the interaction was confirmed through bimolecular fluorescence complementation (BiFC) studies. This novel protein encoded a putative 309 amino acid (aa) polypeptide with a predicted molecular mass of 34 kD that contained a single MYB/SANT-like domain. The protein was named AtSM34 to reflect this, and it was also inferred as maMYB [18], although it was not among the typical MYB genes that have been identified in A. thaliana [19]. Overexpressing lines were hypersensitive to osmotic stress during germination and early seedling growth. Further study showed expression to be localized in the ER, for which the N-terminal was critical. Subsequent experiments revealed that AtSM34 also interacted with the other TIP isoforms, AtTIP1;2 and AtTIP2;1 which are most abundant in germinating seedlings. These observations indicated interactions between AtSM34 and AtTIPs might affect the natural expression patterns of members of TIP subfamilies during Arabidopsis seed germination. For RNA analysis, 6-d-old seedlings were grown on 1/2 MS medium supplemented with 1% (w/v) sucrose. Seedlings were treated with 1/2 MS liquid medium supplemented to provide the stress conditions investigated.
An A. thaliana NubG-x cDNA library (P02210) was screened to identify protein interaction partners of AtTIP1;1. The yeast strain DSY-1 was used for cDNA library screening, and yeast transformation was performed according to the manufacturer's protocol (DUALsystems Biotech, Schlieren, Switzerland). Transformants were plated onto synthetic dropout (SD) medium lacking leucine (Leu), tryptophan (Trp), and histidine (His) supplemented with 10 mmol/L 3-amino-1,2,4-triazole (3-AT). Positive clones were replated on the SD selection medium and analyzed for β-galactosidase activity. Plasmids of the positive blue colonies were recovered from yeast, transformed into E. coli, and sequenced. To confirm the positive interactions, isolated prey plasmids were transformed into the yeast strain DSY1 with the bait pTMBV4-TIP1;1 or control bait pMBV-Alg5.

Isolation of the AtSM34 promoter region and histochemical GUS assays
To examine the tissue-specific expression of AtSM34, a region containing 1035 base pairs (bp) upstream of the initiation codon (ATG) of the AtSM34 promoter region was amplified by PCR from Arabidopsis genomic DNA using KOD DNA polymerase. The following primers were used: Forward (5′-CGCGGATCCTACTCAAAGCAAACAAAC-GAAG-3′) and Reverse (5′-GGCGAATTCTGTTGCTCTG-TTGCACTGTAGA-3′) (restriction enzyme recognition sites are underlined).
Integrity of the amplified product was confirmed by sequencing, then the fragment was cloned into the BamHI and EcoRI sites of the pCAMBIA1391 vector [22]. The construct was introduced into Agrobacterium tumefaciens strain GV3101 and used to transform Arabidopsis Col-0. GUS staining was performed as described previously [22]. Finally, seedlings were washed with 70% ethanol to remove chlorophyll.

Subcellular localization of the AtSM34-GFP fusion protein
The full-length open reading frame encoding region of AtSM34 was cloned into the XbaI and BamHI sites of the transient expression vector pUC-EGFP [23] downstream from the Cauliflower mosaic virus (CaMV) 35S promoter. The AtSM34∆83aa fragment was excised from pUC-SPYNE-AtSM34∆83aa using the XbaI/BamHI enzymes, and then cloned into pUC-EGFP.
For stable expression, the AtSM34 coding region fragment fused to GFP was cloned into the vector pBI121 [25] and expressed under the control of the 35S promoter. This was introduced into Arabidopsis plants through A. tumefaciens strain GV3101-mediated transformation. The roots of kanamycin-resistant seedlings from transgenic plants containing the 35S::AtSM34-GFP fusion protein were analyzed by direct confocal microscopy. Five independent 35S:: AtSM34-GFP transgenic lines were generated and used to investigate the localization of the fusion protein. The excitation wavelength for GFP detection was 488 nm and emissions were collected between 515 and 530 nm under a Zeiss ×40 oil objective. To visualize the ER, the transformed Arabidopsis mesophyll protoplasts and roots of 35S::AtSM34-GFP transgenic plants were stained with 1 μmol/L redorange-fluorescent BODIPY 558/568-conjugated brefeldin A (B7449, Invitrogen) for 30 min. Fluorescence was detected at an excitation wavelength of 543 nm and emissions collected between 585 and 615 nm.
ACTIN2/8 (At3g18780) mRNA expression was amplified as an internal control. The specific primers used for semiquantitative RT-PCR were: 5′-TCTTCCGCTCTTTCTT-TCCA-3′ and 5′-GAGAGAACAGCTTGGATGGC-3′ [26]. Amplification of ACTIN2/8 and AtSM34 was carried out using 23 and 29 cycles, respectively. The PCR program used was as follows: 94°C for 30 s, 56°C for 30 s and 72°C for 50 s. A final incubation at 72°C for 7 min was performed to complete product synthesis.
The relative expression levels of genes detected were calculated against a standard curve of expression of the ACTIN2/8 calibrator using the 2 ∆∆C T method [28]. Realtime quantitative PCR experiments were independently repeated in triplicate.
Amplified regions were cloned into the XbaI/KpnI sites of the modified pSuper1300 vector [29] and sequenced. To modify the pSuper1300 vector, the FLAG fragment was amplified by PCR and inserted into the pSuper1300 vector using the KpnI and SacI sites. This construct (pSuper1300-AtSM34-FLAG-TNOS) was introduced into A. tumefaciens strain GV3101 and used to transform wild-type Arabidopsis plants using the floral-dip method [30]. Transgenic plants were screened for hygromycin resistance on 25 μg/mL hygromycin plates. Resistant transformants were transferred to soil and grown in a greenhouse. According to the segregation ratios of T 2 and T 3 seeds on hygromycin, T 3 homozygous lines were selected for expression analysis and pheno-type characterization.
Approximately 60 seeds from each of the wild-type and AtSM34 overexpression transgenic lines (OE-6 and OE-10) were sown on 1/2 MS containing 1% (w/v) sucrose media and supplemented with mannitol (400 mmol/L), sorbitol (400 mmol/L) or abscisic acid (ABA) (1 μmol/L). Plants were incubated at 4°C for 3 d before being placed at 22°C under long-day conditions. Germination (emergence of radicles) was scored daily for 6 d. Plant growth was monitored and photographed after 7 d. The experiments were repeated independently three times.

Western blot analysis
Total proteins were extracted by incubating the plant materials in protein extraction buffer. Proteins were fractionated by SDS-PAGE on 12% (w/v) gel using a minigel system (Bio-Rad, Hercules, CA, USA). The separated proteins were transferred from the gel to a polyvinylidene fluoride nitrocellulose membrane (Bio-Rad), and then blotted with an anti-flag (Sigma) primary antibody (diluted 1:10000). The membrane was washed three times with TBS-T buffer followed by incubation for 1 h with the secondary antibody (peroxidase-conjugated affinipure goat anti-rabbit IgG [H+L]). The washing process was repeated before the proteins on the membrane were detected with an enhanced chemiluminescence system (Amersham Biosciences, Uppsala, Sweden).

Yeast split-ubiquitin assay for the interaction between AtSM34 and AtTIP1;1
To search for potential protein interactions, AtTIP1;1 was used as bait to screen the NubG-x cDNA library from A. thaliana 6-d-old seedlings according to the protocol provided for the DUAL membrane yeast split-ubiquitin system (DUALsystems Biotech). Prey plasmids from the positive clones were isolated and sequenced. Further analysis identified a unique clone encoding a 309 aa putative polypeptide (AtSM34, At5g45420) with a predicted molecular mass of 34 kD that contained a MYB/SANT-like domain.

Bimolecular fluorescence complementation assay
The interaction of AtTIP1;1 and AtSM34 in plants was confirmed previously by BiFC assays [20]. TIP1;1 was fused with the C-terminal portion of the yellow fluorescent protein (YFP) and AtSM34 was fused with the N-terminal portion of the same protein. These plasmids were then cotransformed into the Arabidopsis 35S::spRFP-AFVY transgenic lines or wild-type Arabidopsis mesophyll protoplasts. Localized restored YFP fluorescence was detected as punctate structures similar to small vesicles or prevacuolar compartments (Figure 1(c)). Previously, 35S::spRFP-AFVY was reported to localize exclusively in the lumens of the large, central vegetative vacuoles [31]. Vacuolar fusion was dynamic, with tetrapeptide AFVY carrying the RFP proteins that could also label a variety of vesicles, such as proteinstorage vacuoles, dense vesicles or precursor-accumulating vesicles [32] concurrently. The arrows in Figure 1(c) indicate that the YFP fluorescence colocalized with the red separated small vesicles around the central large vacuole (Fig-ure 1(c) upper merge panel). To identify the compartments further, we also stained the wild-type Arabidopsis mesophyll protoplast with FM4-64 dye, which is a useful tool to track endocytosis and vesicle trafficking in living cells [33]. The restored YFP signals colocalized with some vesicles stained with FM 4-64 intracellularly, but were not detected on the central vacuole membrane (Figure 1(c) middle merge panel). The BiFC assays showed that the interaction between AtTIP1;1 and AtSM34 occurred in the small vesicles. These data indicate AtSM34 might be involved in vesicle trafficking.
As a comparison with the above BiFC assays, TIP1;1 was fused with both the C-and N-terminal portions of YFP that were expressed transiently in Arabidopsis mesophyll protoplasts simultaneously stained with FM4-64. Aquaporins are known to form homomers before they can mediate their biological functions. The restored YFP fluorescence was located preferentially in the tonoplasts in both the large central vacuole and an adjacent small peripheral vacuole (Figure 1(c) bottom panel). These findings indicated the interaction between AtTIP1;1 and AtSM34 occurred in the vesicles, but not on the membrane of the large central vacuole.

Expression pattern of AtSM34 in different tissues
Expression patterns of AtSM34 in different organs of Arabidopsis plants were analyzed by RT-PCR. Total RNA was isolated from roots, stems, leaves, flowers, and young siliques. Expression of ACTIN2/8 was used as the loading control. As shown in Figure 2, the transcription levels of AtSM34 detected in the flowers, leaves and stems were higher than the levels detected in roots and young siliques. The expression patterns of AtSM34 were examined in greater detail by generation of a construct encoding the β-glucuronidase (GUS) reporter gene under the control of the AtSM34 promoter (AtSM34promoter::GUS). Following transformation of wild-type Arabidopsis plants, more than three independent lines were used for analyses. Strong GUS signals were observed in cotyledons of the germinating seeds (Figure 2  leaves, particularly the vascular tissues (Figure 2(f)). Strong staining was also observed in mature stems (Figure 2(h)) and in the petals and calyces of flowers (Figure 2(g)). However, GUS activity was virtually absent in the roots and young siliques.

Localization of AtSM34 to the endoplasmic reticulum
The Arabidopsis Information Resource (TAIR) database (http://www.arabidopsis.org) forecast exhibited AtSM34 was localized to endoplasmic reticulum. To examine the subcellular localization of AtSM34, GFP fused either at the C-terminal (AtSM34-GFP) or N-terminal (GFP-AtSM34) to AtSM34 under the control of the CaMV 35S promoter was transiently expressed in Arabidopsis mesophyll protoplasts. Green fluorescent structures similar to the ER were observed by confocal microscopy. Previous investigations showed ER-localized proteins exhibit a network pattern in protoplasts [34]. To confirm the localization of AtSM34 expression, mesophyll protoplasts cells were stained with the ER marker red-orange-fluorescent BODIPY 558/568-conjugated brefeldin A. Overlaid green and red signals confirmed that the GFP signals were colocalized with the ER (Figure 3(a)). Appending GFP to the N-or C-terminus of AtSM34 resulted in an identical pattern of localization (Figure 3(a)).
Further studies of transgenic seedlings expressing AtSM34- GFP were also observed using confocal fluorescence imaging analysis. Young root cells from stably transformed 35S::AtSM34-GFP plants and simultaneously stained with the ER marker showed overlapping green and red fluorescence, predominantly in the ER (Figure 3(a) bottom panel). This was consistent with the localization of maMYB in tobacco epidermal cells [18].

N-terminal amino acids of AtSM34 critical for subcellular localization
Hydrophobicity analysis predicted the presence of two transmembrane domains contained within the 83 N-terminal amino acids of AtSM34 (http://www.ch.embnet.org/software/ TMPRED_form.html). When the GFP was fused to AtSM34∆83aa, the green fluorescent signals were distributed in both the cytoplasm and nucleus (Figure 3(b) middle panel). This observation indicated the N-terminal region might be essential for ER membrane subcellular localization. Ubiquitous expression of the control, 35S::GFP, was observed ( Figure 3

(b) upper panel).
To further elucidate the role of this region in the interaction with AtTIP1;1, Arabidopsis protoplasts were cotransformed with YNE-AtSM34∆83aa and YCE-AtTIP1;1. Restored YFP fluorescence was relocalized to the large central vacuolar membrane (Figure 3(b) bottom panel), which was similar to the tonoplast localizations of some TIP members in protoplasts [12]. This confirmed the N-terminus of AtSM34 as the transmembrane domain was not required for the protein-protein interaction.

Hypersensitivity of AtSM34 overexpression lines to osmotic stress and ABA
To investigate its biological function, AtSM34 was overexpressed under the control of the Super promoter and 20 independent homozygous AtSM34 overexpression lines were prepared. Quantitative real-time PCR analysis indicated the level of AtSM34 transcripts in OE-6 and OE-10 were higher than that found in the wild type (Figure 4(a)). This result was confirmed by Western blot analysis using an anti-flag antibody directed against the flag tag in the transgenic lines. Bands corresponding to the AtSM34-FLAG fusion protein were clearly visible in lines OE-6 and OE-10, whereas no expression was detected in the wild type (Figure 4(b)). Homozygotes from lines OE-6 and OE-10 were selected for use in further experiments.
To further elucidate the possible function of AtSM34, the effects of stressful conditions on seed germination and early seedling growth were investigated. Under normal growth conditions, no significant differences were observed between the transgenic and wild-type plants. However, seed germination and early growth development of transgenic plants (OE-6 and OE-10) were significantly delayed on 1/2 MS medium supplemented with high concentrations of mannitol, sorbitol or ABA ( Figure 5(a)).
After 3 d, more than 70% of the wild-type seeds germinated. In contrast, the germination rate of the transgenic seeds was less than 20% on 1/2 MS medium containing 400 mmol/L mannitol ( Figure 5(b)). Similarly, on 1/2 MS medium containing 400 mmol/L sorbitol, the germination frequencies of the transgenic seeds were only 30%-40%, whereas the germination frequencies of the wild-type seeds were 72% (Figure 5(b)). On 1/2 MS medium containing 1 μmol/L ABA, approximately 40% of the transgenic seeds germinated after 2 d, compared to more than 60% of the wild-type seeds. These results demonstrated that overexpression of AtSM34 conferred hypersensitivity to osmotic stress and ABA, particularly during seed germination and the early stages of plant growth.

Expression profiles of AtSM34 in response to stress
To investigate the impact of osmotic stress on AtSM34 expression, 6-d-old wild-type seedlings were subjected to mannitol treatment for different periods of time and the  Expression profiles of AtSM34 under other stress conditions were also analyzed by real-time PCR. Similar to the effect of osmotic stress, a slight induction in AtSM34 tran-script levels were observed following treatment with ABA, methyl viologen or glucose for 4 or 12 h. These results indicated AtSM34 is induced in response to different types of abiotic stress and therefore might be involved in plant responses to environmental stimuli.
In addition, interactions between AtSM34 and the TIPs were specific. In the yeast split-ubiquitin system assays, AtSM34 interacted with AtTIP1;3, but not with AtTIP2;2 and AtTIP3;2 (Figure 7(b)). Therefore, the possibility of other aquaporins interact with AtSM34 cannot be excluded.

Effect of AtSM34 expression of TIP genes in Arabidopsis under osmotic stress
TIP1;1, TIP1;2 and TIP 2;1 are known to have the highest basal gene expression levels in tonoplast aquaporins groups 1 and 2, respectively [1], and all interacted with AtSM34 in BiFC experiments. Therefore, these isoforms were selected for further investigation of the effect of AtSM34 on aquaporin transcript levels under both normal and stress conditions. Real-time PCR analysis was performed using the TIP gene-specific primers reported in Beebo et al. [17]. As shown in Figure 8, the expression levels of TIP1;1, TIP1;2 and TIP2;1 were modified in the overexpression lines under both normal and osmotic stress conditions. Expression levels of the three TIPs were downregulated in the wild type and OE-6 and OE-10, with rapid downregulation observed in the OE lines under osmotic stress for 12 h. These results demonstrated ectopic expression of AtSM34 disrupts the natural expression patterns of some TIP genes in Arabidopsis plants under osmotic stress.

Discussion
In this study, we identified a novel protein, AtSM34, as a putative interactor with members of the Arabidopsis tonoplast intrinsic protein family by experiments in yeast and plant cells. Although there are a number of reports of aquaporin-interacting proteins in mammalian systems, there are very few reports in plants, with the exception of some aquaporins known to form homomers or heteromers [35]. TIPs are one of the most abundant aquaporin subgroups in plants and, as such, the detection of their interrelated proteins may elucidate the mechanisms of their function and regulation.
It is known that all aquaporin isoforms traffic through the secretory pathway before reaching their final destination compartments during the course of their synthesis and maturation. Reports have indicated expression of TIPs is predominantly located in the tonoplast [12]. However, McTIP1;2 (McMIPF) proteins redistributed to the endosomes in osmotic stress [14]. Recent studies have indicated that aquaporin trafficking is a critical point for regulating aquaporin expression and function [35]. The results of BiFC colocalization suggested that the interactions might occur in small vesicles that are involved in the process of the aquaporin sorting between ER and vacuoles although the mechanism of this process requires further investigation. Changes in the interaction patterns after deletion of the first 83 aa could be due to the deletion of the TM domains of AtSM34, making it cytosolic, thus, able to directly interact with any cytosolic facing domain of TIP1;1. AtSM34∆83 is soluble and thus synthesized by cytosolic ribosomes and does not depend on the secretory pathway anymore. Therefore, their interaction occurred in the vacuole membrane, which was  identical to the observed TIP1;1 localization.
In AtSM34 promoter controlled β-glucuronidase (GUS) transgenic plants, and strong GUS signals were observed in cotyledons of the germinating seeds (Figure 2(a)) and seedlings (Figure 2(b), (c)). This indicated that AtSM34 might be involved in seed germination and early seedling growth.
A role for aquaporins in cell osmoregulation and maturation of the vacuolar apparatus during late seed development and during the early stages of germination has been proposed [15]. This hypothesis is based on the observation that tissue water content alters significantly during seed maturation and germination and knowledge of the expression and regulation properties of tonoplast aquaporins (α-TIP and β-TIP, or TIP3s) that have been detected in seeds [36,37]. Initial expression of aquaporins has been correlated with germination efficiency [38]. During growth and develop-ment, tonoplast aquaporins modulate the tonoplast permeability and mediate transcellular water transport. In this way, aquaporins contribute to water imbibition in seed germination and facilitate water supply to the expanding tissues for cell enlargement [38,39]. The expression profiles of AtTIPs genes obtained by quantitative real-time PCR revealed that the levels of TIP1;1, TIP1;2 and TIP2;1 transcripts are lower in OE lines under conditions of osmotic stress. Therefore, we proposed that the transgenic plants absorb water more slowly than the wide-type, resulting in a delayed rate of seed germination. In addition, at least 10 homologous TIP isoforms have been identified in Arabidopsis. The possibility that other AQPs interact with AtSM34, leading to delayed germination in response to osmotic stress in the early stage of growth cannot be excluded.
To gain further insight into the role of AtSM34 in the response to osmotic stress, the expression of marker genes known to be involved in regulating this response were monitored by real-time PCR analysis. These marker genes included RD22, RD29A, DREB2A, COR47, KIN1, KIN2, ERD10, NCED3, ADH1 and bZIP60. No significant alteration in the expression of these genes was detected, with the exception of a decrease in RD29A in the AtSM34 OE lines under conditions of osmotic stress for 12 h ( Figure S1). RD29A has been reported to have at least two cis-acting elements; one involved in the ABA-associated response to desiccation and the other induced by changes in osmotic potential [40]. These data indicated that AtSM34 is not involved in the usual stress signaling pathway.
In plants, the MYB family has selectively expanded and MYB proteins are key factors in regulatory networks controlling development, metabolism and responses to biotic and abiotic stresses [19,[41][42][43][44]. Novel proteins harboring Myb/SANT-like domains distinct from other described MYB-like transcriptional regulators have been detected in plants gradually. Ectopic expression of the tomato early fruit specific gene, Lefsm1, resulted in severe developmental alterations, retarded growth, and reduced apical dominance during tomato and Arabidopsis seedling development [45]. MYB related genes have been implicated in a wide variety of other plant-specific processes, although their physiological function has not been fully understood, and characterization of a great number of MYB-like proteins remains elusive [46]. So, AtSM34 is likely to be a potential membrane-tethered transcription factor and the possibility remains that AtSM34 interacts with other proteins in addition to aquaporins to delay germination of AtSM34 overexpressing plants under conditions of osmotic stress cannot be excluded.
The precise function of AtSM34 in the response of plants to osmotic stress at the early growth stage remains to be characterized. Further studies of downstream targets of AtSM34 will provide a greater understanding of the molecular function of this protein.