Overexpression of MtCAS31 enhances drought tolerance in transgenic Arabidopsis by reducing stomatal density


Author for correspondence:
Jiangli Dong
Tel: +86 10 62733969
Email: dongjl@cau.edu.cn


  • Dehydrins are a type of late embryogenesis abundant protein. Some dehydrins are involved in the response to various abiotic stresses. Accumulation of dehydrins enhances the drought, cold and salt tolerances of transgenic plants, although the underlying mechanism is unclear. MtCAS31 (Medicago Truncatula cold-acclimation specific protein 31) is a Y2K4-type dehydrin that was isolated from Medicago truncatula.
  • We analyzed the subcellular and histochemical localization of MtCAS31, and the expression patterns of MtCAS31 under different stresses. Transgenic Arabidopsis that overexpressed MtCAS31 was used to determine the function of MtCAS31. A yeast two-hybrid assay was used to screen potential proteins that could interact with MtCAS31. The interaction was confirmed by bimolecular fluorescence complementation (BiFC) assay.
  • After a 3-h drought treatment, the expression of MtCAS31 significantly increased 600-fold. MtCAS31 overexpression dramatically reduced stomatal density and markedly enhanced the drought tolerance of transgenic Arabidopsis. MtCAS31 could interact with AtICE1 (inducer of CBF expression 1) and the AtICE1 homologous protein Mt7g083900.1, which was identified from Medicago truncatula both in vitro and in vivo.
  • Our findings demonstrate that a dehydrin induces decreased stomatal density. Most importantly, the interaction of MtCAS31 with AtICE1 plays a role in stomatal development. We hypothesize that the interaction of MtCAS31 and AtICE1 caused the decrease in stomatal density to enhance the drought resistance of transgenic Arabidopsis.


Abiotic stresses, such as drought, salinity and low temperature, limit the growth of plants. To resist such stresses, plants accumulate a large number of specific proteins. Dehydrins (DHNs) are a class of proteins that respond to a variety of abiotic stresses (Mundy & Chua, 1988; Choi et al., 1999; Nylander et al., 2001; Heyen et al., 2002; Wisniewski et al., 2006; Rurek, 2010). DHNs belong to the late embryogenesis abundant (LEA) proteins and are classified as LEA group 2 or the D-11 family (Hundertmark & Hincha, 2008). Since the isolation of the first DHN gene, RAB21 (Responsive to abscisic acid 21), which is an abscisic acid (ABA)-responsive gene in rice (Oryza sativa L.) (Mundy & Chua, 1988), numerous additional DHNs have been isolated from different plant species. For example, in barley (Hordeum vulgare) 12 DHNs have been identified (Choi et al., 1999; Choi & Close, 2000), and Arabidopsis has at least 10 DHNs (Hundertmark & Hincha, 2008). The structure of DHN is disordered, which means DHN is an intrinsically unstructured protein (Lisse et al., 1996; Kovacs et al., 2008; Mouillon et al., 2008). In some instances, DHNs change their structure when they bind to partner molecules (Koag et al., 2009).

All DHNs harbor a highly conserved Lys-rich K segment (Close, 1997), which can form amphipathic α-helices. This structure confers DHNs with both hydrophilic and hydrophobic characteristics, which might allow DHNs to interact with membranes and partially denatured proteins. The consensus Y-segment is located near the N terminus (Close, 1996; Allagulova et al., 2003). Another consensus sequence in DHNs is the S-segment, which comprises a string of repeated Ser residues that can be phosphorylated (Goday et al., 1994).

Accumulation of DHNs increases abiotic stress tolerance in transgenic plants. For example, wheat (Triticum aestivum L.) WCOR410 (Wheat cold regulated gene 410) proteins improve the freezing tolerance of transgenic strawberry (Fragaia ananassa Duchesne) (Houde et al., 2004), and rice Rab16 (Responsive to abscisic acid 16) proteins enhance the salt and drought tolerance of transgenic tobacco (Nicotiana tabacum L.) (RoyChoudhury et al., 2007). However, the mechanisms that underlie DHN responses to abiotic stress are not well characterized. In previous studies, it was suggested that DHNs might stabilize membranes and macromolecules (Karlson et al., 2003; Puhakainen et al., 2004), act as chaperones (Hara et al., 2001; Kovacs et al., 2008), bind a variety of heavy metal and calcium ions (Heyen et al., 2002; Hara et al., 2005), and aid water retention during dehydration stress (Tompa et al., 2006).

Stomata are present on the leaf surface and allow plants to regulate gas exchange and water lost by evapotranspiration. Stomatal development is regulated both by external environmental cues and by endogenous genes. A number of proteins have been demonstrated to regulate stomatal density. During stomatal differentiation, SPEECHLESS (SPCH) directs the first asymmetric cell division to initiate the stomatal cell lineage, MUTE controls the transition from the meristemoid to the guard mother cell, and FAMA determines the symmetric division to form a pair of guard cells (Bergmann & Sack, 2007; MacAlister et al., 2007; Pillitteri et al., 2007; Serna, 2009). Recent research has demonstrated that ICE1 (INDUCER OF CBF EXPRESSION 1) and SCREAM2 (INDUCER OF CBF EXPRESSION 2) interact with the above three critical transcription factors during Arabidopsis stomatal differentiation (Kanaoka et al., 2008; Serna, 2009).

In our previous study, total RNAs from the roots of Medicago truncatula plants, which had been treated with 180 mM NaCl, were used as samples for microarray experiments. We observed an abundance of up-regulated genes in response to NaCl treatment, of which MtCAS31 (Medicago truncatula cold-acclimation specific protein 31) showed the highest up-regulation of all of the genes (Li et al., 2009). MtCAS31 encodes 312 amino acids and is a Y2K4-type DHN. Only one copy of MtCAS31 is present in the M. truncatula genome. The promoter of MtCAS31 is induced by chilling stress (Pennycooke et al., 2008). In the present study, we analyzed expression of MtCAS31 in response to NaCl, ABA, cold and drought stress. MtCAS31 was highly significantly up-regulated following drought stress. Overexpression of MtCAS31 markedly increased drought tolerance and decreased stomatal density of transgenic Arabidopsis. In addition, we investigated the putative mechanism by which MtCAS31 enhanced drought resistance of transgenic Arabidopsis plants.

Materials and Methods

Plant material and treatments

Seeds of Medicago truncatula Gaertn cv Jemalong A17 were surface-sterilized in concentrated sulfuric acid for 15 min, then rinsed five times with distilled H2O, and chilled at 4°C for 2 d. The seeds were sown on moist filter paper to germinate. After an additional 2 d, the seedlings were transplanted into soil and vermiculite (1 : 1, v/v). All plants were grown in a glasshouse under long-day conditions (16 h : 8 h, light : dark cycle) at 22°C with 50–60% relative humidity.

Four-week-old seedlings were treated with four abiotic stresses. For salt treatment, the plants were treated with 180 mM NaCl, and sampled at 6, 24, and 48 h after treatment. For ABA treatment, seedlings were transferred to filter paper in a Petri dish, sprayed with 200 μM ABA solution containing 0.05% Tween20 (v/v), and sampled at 1, 6, and 24 h after treatment. For cold treatment, the plants were incubated in a growth chamber at 4°C and sampled at 1, 6, and 24 h after treatment. For drought treatment, the seedlings were placed on dry filter paper and sampled after air-drying for 1, 2, and 3 h. Each treatment contained three biological repeats.

RNA extraction and semiquantitative reverse transcription PCR (RT-PCR)

Total RNA of samples were extracted with TRIzol reagent (Invitrogen, USA), and genomic DNA was removed with RNase-free DNaseI (Promega). A sample (2 μg) of the purified total RNA was used for reverse transcription with M-MLV (Moloney Murine Leukemia Virus) Reverse Transcriptase (Promega).

For transgenic Arabidopsis analyses, the cDNAs of transgenic and wild-type Arabidopsis thaliana (L.) Heynh were used for semiquantitative RT-PCR. AtActin2 was used as an internal control and was amplified with the following primers: 5′-TAACAGGGAGAAGATGACTCAGATCA-3′ (forward) and 5′-AAGATCAAGACGAAGGATAGCATGAG-3′ (reverse). The primers for transgenic Arabidopsis were: 5′-GTGATCAAACACGTAGGGTTGATGA-3′ (forward) and 5′-ACACCACCTGTGTGGGTCCC-3′ (reverse).

Quantitative real-time PCR analyses

The cDNAs of M. truncatula plants treated with different abiotic stresses were used for quantitative real-time PCR (qRT-PCR). The qRT-PCR analysis was performed with a CFX-96 Real-Time System (Bio-Rad) using SYBR Premix Ex Taq (TaKaRa). The PCR cycling conditions were as follows: 95°C for 30 s, 40 cycles of 95°C for 10 s, and 51°C for 48 s. Data were collected at 51°C in each cycle by plate reading. The melting curve was prepared from 65 to 95°C to test the amplification specificity. In M. truncatula, Actin was used as an internal control to normalize the amount of cDNA in samples (Li et al., 2011). The gene-specific primers used were as follows: 5′-CCCACTGGATGTCTGTAGGTT-3′ (forward) and 5′-AGAATTAAGTAGCAGCGCAAA-3′ (reverse) for Actin; 5′-AACAAAATACTTATGGGACAGGC-3′ (forward) and 5′-GTTACATACGAACCAACTCACTCA-3′ (reverse) for MtCAS31.

In Arabidopsis, two internal controls, AtActin2 and At1g13320, were used in the qRT-PCR analysis. The primers for qRT-PCR were identical to those used in previous studies (Czechowski et al., 2005; Miura et al., 2007; Miura & Ohta, 2010).

Three independent repetitions were performed under the same conditions per experiment. All data were analyzed with Bio-Rad CFX Manager.

Transformation and regeneration of Arabidopsis

The promoter of MtCAS31 was cloned by PCR from genomic DNA using the PstI and NcoI restriction sites. The amplified fragment was inserted into the pCAMBIA1305.2 vector between the PstI and NcoI sites upstream of the GUS gene to form a chimeric pP31-GUS vector (PMtCAS31:GUS). The open reading frame (ORF) of MtCAS31 was introduced into pCAMBIA1302 at the BglII restriction site to construct the recombinant pG31 vector (PCaMV35S:MtCAS31). pP31-GUS and pG31 were each transformed into Arabidopsis ecotype Columbia using the Agrobacterium tumefaciens-mediated floral dip method as described by Zhang et al. (2006). Seeds of transformed Arabidopsis were selected on Murashige and Skoog (MS) medium containing 80 μm ml−1 hygromycin. Three independent lines of the T3 generation were randomly chosen for further analysis.

Histochemical localization

For histochemical GUS assays, tissues and seedlings of T3PMtCAS31:GUS transgenic Arabidopsis were analyzed following the protocol described by Jefferson et al. (1987), and observed under a microscope.

Drought tolerance analysis of transgenic plants

The T3PCaMV35S:MtCAS31 transgenic and wild-type Arabidopsis seedlings were transplanted into pots containing soil and vermiculite (1 : 1, v/v). The seedlings were grown in a growth chamber for 4–5 wk under short-day conditions (12 h : 12 h light : dark) or long-day conditions (16 h : 8 h, light : dark cycle) at 22°C and 50–60% relative humidity.

For measurement of leaf water loss, the aboveground parts of the plants were excised from the roots and placed on weighing paper. All samples were air-dried slowly on a bench (25°C; 50% relative humidity) and weighed at different time-points, with three replicates per time-point. The percentage water loss at each time-point was calculated as the percentage reduction from the initial weight of the plant tissue.

To test drought tolerance, water was withheld from plants for 15 d. Digital images of the plants and survival were recorded on the 15th day after drought treatment. Three independent experiments were performed.

Stomatal density and index counts

Four-week-old rosette leaves were used for stomatal density and index counts. Stomata are pores in the epidermis that are bordered by a pair of guard cells, which determine the pore size to regulate gas exchange and transpiration. Stomata were counted on the abaxial epidermis at the midpoint of the leaf lamina on the midrib and margin. The stomata were photographed and counted with the aid of a scanning electron microscope (TM-3000; Hitachi, Ibaraki, Japan). The experiment was repeated three times.

To clear the leaves of Arabidopsis, leaves were incubated in ethanol : glacial acetic acid (3 : 1, v/v) for 1 h, then in basic solution containing 7% NaOH (g/v) in 60% ethanol (v/v) for 30 min. An ethanol series of decreasing concentrations was used for rehydration of tissues. The transparent leaves were stained in Lugol’s solution (5% (g/v) iodine and 10% (g/v) potassium iodide) for 5 min, then incubated in water until the leaves were destained. All leaves were observed under an Olympus BX51 microscope. The formula SI = (S/(S)) × 100 was used for calculation of the stomatal index (SI), where S is the number of stomata and E is the number of epidermal cells (Wilson et al., 2009).

Yeast two-hybrid assay

MtCAS31 was fused with the pGBKT7 vector at the NdeI and EcoRI sites as the bait vector, which was used to screen an Arabidopsis cDNA library following the manufacturer’s protocol (Clontech, Mountain View, CA, USA).

AtICE1 and Mt7g083900.1 were cloned into pGADT7 separately as prey vectors. The vectors were transformed into yeast strain AH109 according to the manufacturer’s protocol (Clontech). The transformed yeasts were selected on SD medium containing –Ade/–His/–Leu/–Trp dropout supplement. The colonies that contained two interacting proteins developed after incubation for 3 d at 30°C.

Subcellular localization and bimolecular fluorescence complementation assay

The ORF of MtCAS31 without the termination codon was inserted into the modified pE3025 plasmid (which contained GFP instead of RFP) at the EcoRI and SalI sites to form the pE31-GFP vector. Similarly, the ORF of Mt7g083900.1 was fused with the mGFP of pE3025 to form the pEMt7g083900-GFP vector. The fusion vectors pE31-GFP (35S:MtCAS31-GFP) and pEMt7g083900-GFP (35S:Mt7g083900.1-GFP), and the control vector (35S:GFP) were transformed into Arabidopsis. The full-length MtCAS31 cDNA was inserted into pSY735 to form pSY735-MtCAS31. AtICE1 and Mt7g083900.1 were cloned into pSY736 to form pSY736-AtICE1 and pSY736-Mt7g083900.

Rosette leaves of 3- to 4-wk-old Arabidopsis plants grown under short-day conditions were used for isolation of protoplasts. Transformation was performed as described by Chen et al. (2010). The relevant vectors were used in polyethyleneglycol-mediated transformation of the Arabidopsis protoplasts. From 12 to 18 h after transformation, the protoplasts were assayed for fluorescence. Digital images were captured with a confocal laser scanning system (ECLIPSE TE2000-E; Nikon, Tokyo, Japan).

Simultaneously, pE31-GFP, pEMt7g083900-GFP and the control vector were bombarded into living onion (Allium cepa) epidermal cells with a biolistic device. After culture on MS medium for 16–20 h in the dark, the onion cells were observed with a confocal laser scanning system (ECLIPSE TE2000-E; Nikon).

Identification of ICE1 homologs from M. truncatula

To identify ICE1 homologs from M. truncatula, the protein sequence of AtICE1 was used as the query sequence for blastp analyses (http://blast.jcvi.org/er-blast/index.cgi?project=mtbe). A phylogenetic analysis of ICE1-like proteins from M. truncatula and the ICE family from Arabidopsis was performed. Protein sequences in the ICE1 family were used to generate a midpoint-rooted neighbor-joining tree. The trees were created with mega 4.0 (Tamura et al., 2007) using the default parameters. ClustalX version 2 (Thompson et al., 2002) was used to generate the multiple alignment.


MtCAS31 was induced by NaCl, ABA, cold and drought

Expression of MtCAS31 was induced significantly by NaCl, ABA, cold and drought (Fig. 1). In particular, expression was elevated dramatically in response to drought. After 3 h of drought treatment, MtCAS31 expression was at least 600-fold higher than that of the control. With NaCl treatment, MtCAS31 expression was increased after 6 h but subsequently decreased. Similar results were obtained in response to ABA treatment. Both cold and drought treatment induced an increase in MtCAS31 expression. These results indicated that MtCAS31 is involved in the response to different stresses, particularly drought stress, but the response to these stresses is not notably rapid.

Figure 1.

Expression pattern analyses of MtCAS31 (Medicago Truncatula cold-acclimation specific protein 31) in Medicago truncatula under different abiotic stresses. Quantitative real-time PCR was used to analyze the expression patterns of MtCAS31 in M. truncatula in response to NaCl (180 mM), abscisic acid (ABA; 200 μM), cold (4°C), and drought at selected time-points. The values were normalized against Actin expression. Data represent the mean ± SD of three technical replicates.

Subcellular localization and histochemical localization of MtCAS31

To study the subcellular localization of MtCAS31, MtCAS31 was fused with the N terminus of GFP and transiently expressed in Arabidopsis protoplasts and onion epidermal cells under the control of the CaMV 35S promoter. After 16–20 h, accumulation of the MtCAS31-GFP fusion protein in the cell membrane, cytoplasm and nucleus of Arabidopsis protoplasts was detected (Fig. 2d–f), which was the same as the subcellular localization of the GFP protein (Fig. 2a–c). The same response was observed in onion epidermal cells (Supporting Information Fig. S1). Previous studies found that DHNs can be localized in the nucleus, chloroplasts, vacuole, rough endoplasmic reticulum, mitochondria, cytoplasm and membranes (Heyen et al., 2002; Mueller et al., 2003; Carjuzaa et al., 2008).

Figure 2.

Subcellular localization of MtCAS31 (Medicago Truncatula cold-acclimation specific protein 31). Arabidopsis protoplasts were transfected with a construct carrying GFP or MtCAS31-GFP under the control of the cauliflower mosaic virus (CaMV) 35S promoter. All images were captured with a confocal laser scanning system. (a–c) Expression of GFP protein. (d–f) Expression of MtCAS31-GFP fusion protein.

To investigate the localized expression pattern of MtCAS31, the c. 1.7-kb promoter of MtCAS31 was fused with GUS, and the chimeric vector was transformed into Arabidopsis. The T3 seeds of three independent transgenic lines were germinated on MS medium. The embryo and 4-d-old seedlings were used for analysis. After GUS staining, all three lines gave the same results (Fig. 3). GUS activity was detected in the embryo of the seed, and in the vascular tissue and guard cells of transgenic Arabidopsis plants. Detection of DHNs has been reported in all plant tissues, such as the seed, root, stem, leaf, and flower (Parra et al., 1996; Nylander et al., 2001; Rurek, 2010).

Figure 3.

Histochemical localization of GUS activity in transgenic PMtCAS31:GUS Arabidopsis. (a) A 4-d-old seedling (×1.25); (b) root (×10); (c) embryo of seed (×20); (d) leaf (×20).

MtCAS31 overexpression enhanced the drought tolerance of Arabidopsis

Because MtCAS31 expression was strongly induced by drought in M. truncatula, our study focused on the drought tolerance of PCaMV35S:MtCAS31 transgenic Arabidopsis. The MtCAS31 cDNA was driven by the cauliflower mosaic virus (CaMV) 35S promoter and constitutively expressed in transgenic Arabidopsis. No notable morphological differences were observed between wild-type and transgenic Arabidopsis plants throughout their life cycle. We randomly chose three independent transgenic lines for further analysis. As shown in Fig. 4(a), semiquantitative RT-PCR indicated that transgenic line 2 showed the highest expression of MtCAS31 and line 9 showed the lowest expression.

Figure 4.

Overexpression of MtCAS31 (Medicago Truncatula cold-acclimation specific protein 31) in Arabidopsis increased drought tolerance. (a) Semiquantitative RT-PCR analysis of MtCAS31 expression in transgenic Arabidopsis. (b) Water loss in 5-wk-old detached leaves at different time-points with three replicates. Three independent experiments were performed with similar results. The data represented here were obtained from one experiment. Data represent the mean ± SD. (c) Drought tolerance of 4-wk-old Arabidopsis plants from which water was withheld for 15 d. Survival rate for WT, Line 2, Line 5 and Line 9 was 16.7%, 74.3%, 69.5% and 62.2%, respectively. Survival statistics are presented as (number of plants survived/number of plants tested), respectively. This experiment had three replicates. WT, wild type.

The rate of water loss differed significantly between wild-type and PCaMV35S:MtCAS31 Arabidopsis (Student’s t-test, < 0.05) after 4 h of water stress (Fig. 4b). The fresh weight of PCaMV35S:MtCAS31 plants was reduced to 73.6% for line 2, 68.5% for line 5, 66.9% for line 9, and 59.9% in wild-type plants.

The survival rate of PCaMV35S:MtCAS31 plants deprived of water for 15 d was significantly higher (Student’s t-test, < 0.05) than that of wild-type plants (Fig. 4c). The survival rate was 74.3% for line 2, 69.5% for line 5, 62.2% for line 9, and only 16.7% for wild-type plants.

The results of these experiments confirmed that overexpression of MtCAS31 can greatly improve the drought resistance of transgenic Arabidopsis.

MtCAS31 decreased stomatal density and index

To clarify the physiological mechanism responsible for the increased drought tolerance of MtCAS31-overexpression lines, the stomatal density on the abaxial epidermis of 4-wk-old rosette leaves was studied. The morphology of stomata in wild-type and PCaMV35S:MtCAS31 Arabidopsis was normal, and no heteromorphic stomata were observed (Fig. 5a,b). The stomatal density of wild-type Arabidopsis was 313 mm−2, and that of PCaMV35S:MtCAS31 Arabidopsis was from 180 to 235 mm−2 (Fig. 5c). Compared with wild-type Arabidopsis, PCaMV35S:MtCAS31 Arabidopsis displayed a substantially lower stomatal density. Cleared rosette leaves from 4-wk-old plants were used to calculate the stomatal index. As shown in Fig. 5(d), the stomatal index of PCaMV35S:MtCAS31 Arabidopsis ranged from 22.82 to 24.63%, which was markedly lower than that of the wild type (28.37%). These findings indicated that the lower stomatal density was one of the reasons for the higher drought tolerance of PCaMV35S:MtCAS31 Arabidopsis compared with the wild type.

Figure 5.

Stomatal density and stomatal index in 4-wk-old rosette leaves of wild-type (WT) and transgenic Arabidopsis. (a) Scanning electron micrograph (SEM) of abaxial epidermis of wild-type Arabidopsis. (b) SEM of abaxial epidermis of transgenic line 2. Bars, 200 μm. (c) Stomatal density in wild-type and transgenic Arabidopsis. (d) Stomatal index in wild-type and transgenic Arabidopsis. This experiment was repeated three times. Data represent the mean ± SD (significant difference from the wild type, as indicated by Student’s t-test: *, < 0.05).

MtCAS31 and AtICE1 interact in yeast and in planta

To investigate the molecular mechanism responsible for decreased stomatal density, MtCAS31 was fused with the pGBKT7 vector to form pGBKT7-MtCAS31 as a bait vector to screen an Arabidopsis cDNA library in a yeast two-hybrid system. Twenty putative interaction proteins were obtained from the screening. Because AtICE1 is involved in stomatal development (Kanaoka et al., 2008), AtICE1 was chosen for further analysis. AtICE1 was revealed to interact with MtCAS31 (Fig. 6a,b). Co-transformation of the pGADT7-AtICE1 prey vector and pGBKT7-MtCAS31 bait vector into yeast showed that MtCAS31 and AtICE1 were able to interact in yeast.

Figure 6.

Yeast two-hybrid and bimolecular fluorescence complementation (BiFC) analyses of interaction between MtCAS31 (Medicago Truncatula cold-acclimation specific protein 31) and AtICE1 (Arabidopsis thaliana inducer of CBF expression 1). (a) Interaction of MtCAS31 and AtICE1 in transformed yeast strain AH109 cells grown on SD/–Ade/–His/–Leu/–Trp medium. (b) Analysis of X-α-galactosidase activity of the relative yeast strains. 1, pGBKT7-53/pGADT7-RecT (positive control); 2, pGBKT7-MtCAS31/pGADT7-AtICE1; 3, pGBKT7-MtCAS31/pGADT7; 4, pGBKT7/pGADT7-AtICE1; 5, pGBKT7/pGADT7 (negative control). (c–e) Interaction of MtCAS31 and AtICE1 in the protoplast analysed by BiFC. (c) YFP fluorescence (green); (d) brightfield images; and (e) merged images.

To investigate whether MtCAS31 and AtICE1 interact in planta, a bimolecular fluorescence complementation (BiFC) assay was performed. pSY735 and pSY736, in which the target genes were fused to the N terminus of the split YFP fragments, were used for the BiFC assay (Bracha-Drori et al., 2004). Previously, localization of AtICE1 in the nucleus was confirmed (Chinnusamy et al., 2003), and in the current study GFP-MtCAS31 was detected in the nucleus (Fig. 2). Consequently, we hypothesized that MtCAS31 and AtICE1 are able to interact in the nucleus. YFP fluorescence (green) was detected in the nucleus of Arabidopsis protoplasts co-transformed with the chimeric vectors pSY735-MtCAS31 and pSY736-AtICE1(Fig. 6c–e), whereas no YFP fluorescence was detected in the negative control (data not shown). These results indicated MtCAS31 and AtICE1 were able to interact in the nucleus of Arabidopsis.

AtICE1 is crucial in the cold-response regulatory network of Arabidopsis and directly regulates AtMYB15 (v-myb avian myeloblastosis viral oncogene homolog 15) and AtCBF3 (CBF3: C-repeat binding factor 3) (Chinnusamy et al., 2003, 2007; Zarka et al., 2003). Most importantly, based on the results of the yeast two-hybrid and BiFC assays, MtCAS31 was shown to interact with AtICE1. We detected expression of AtMYB15 and AtCBF3 in transgenic and wild-type Arabidopsis under normal conditions with qRT-PCR. As shown in Fig. 7, in PCaMV35S:MtCAS31 Arabidopsis AtMYB15 expression was promoted, whereas expression of AtCBF3 was depressed. We speculated that the changes in AtMYB15 and AtCBF3 expression were attributable to the interaction of MtCAS31 and AtICE1, and their interaction may impact on downstream genes regulated by AtICE1. In the present study, we demonstrated that MtCAS31 could interact with one important regulator, AtICE1, but the mechanism of their interaction needs further study.

Figure 7.

Expression of AtMYB15 (v-myb avian myeloblastosis viral oncogene homolog 15) and AtCBF3 (C-repeat binding factor 3) in wild-type (WT) and transgenic Arabidopsis under normal growing conditions. Expression of AtMYB15 and AtCBF3 was analyzed by qRT-PCR. AtActin2 and At1g13320 were used as internal controls. Data represent the mean ± SD of three technical replicates.

Identification of Mt7g083900.1 and the interaction of MtCAS31 and Mt7g083900.1 in yeast and in planta

Five AtICE1 protein homolog sequences were obtained from M. truncatula (Mt3g098690.1, Mt3g111290.1, Mt5g030770.1, Mt7g076840.1 and Mt7g083900.1). The results of the phylogenetic analysis showed that Mt7g083900.1 and Mt3g098690.1 were more homologous to AtICE1 (Fig. S3). We compared the amino acid sequences of Mt7g083900.1 and Mt3g098690.1 with AtICE1 proteins. Mt7g083900.1 shared more conserved regions with AtICE1 than did Mt3g098690.1 (Fig. S4). Thus, Mt7g083900.1 was chosen as the AtICE1 homolog in M. truncatula for further analysis.

The pGADT7-Mt7g083900 prey vector and pGBKT7- MtCAS31 bait vector were co-transfected into yeast. MtCAS31 and Mt7g083900 were able to interact in yeast (Fig. 8a,b).

Figure 8.

Yeast two-hybrid and bimolecular fluorescence complementation (BiFC) analyses of the interaction between MtCAS31 (Medicago Truncatula cold-acclimation specific protein 31) and Mt7g083900.1. (a) Interaction of MtCAS31 and Mt7g083900.1 in transformed yeast strain AH109 cells grown on SD/–Ade/–His/–Leu/–Trp medium. (b) Analysis of X-α-galactosidase activity of the relative yeast strains. 1, pGBKT7-53/pGADT7-RecT (positive control); 2, pGBKT7-MtCAS31/pGADT7-Mt7g083900; 3, pGBKT7-MtCAS31/pGADT7; 4, pGBKT7/pGADT7-Mt7g083900; 5, pGBKT7/pGADT7 (negative control). (c–e) Interaction of MtCAS31 and AtICE1 in the protoplast analyzed by BiFC. (c) YFP fluorescence (green); (d) brightfield images; and (e) merged images.

A BiFC assay was performed to investigate whether MtCAS31 and Mt7g083900 interact in planta. As for construction of pSY736-AtICE1, Mt7g083900 was fused to pSY736 to form pSY736-Mt7g083900. In the present study, the subcellular localization of Mt7g083900.1 in the nucleus was confirmed (Fig. S2). YFP fluorescence (green) was detected in the nucleus of transgenic Arabidopsis protoplasts (Fig. 8c–d), but not in the negative control (data not shown). These results indicated MtCAS31 and Mt7g083900.1 were able to interact in the nucleus of Arabidopsis.


In the present study, expression pattern analyses of MtCAS31 demonstrated that MtCAS31 transcripts accumulate to a high level in response to several abiotic stresses, particularly drought stress. This result is similar to the findings of previous studies. OsDhn1 was strongly induced by drought, and transiently induced by cold, salt and ABA (Lee et al., 2005), the expression of PpDHNA increased significantly after treatment with ABA, NaCl or mannitol (Saavedra et al., 2006), and CIDhn was up-regulated by ABA, NaCl and chilling (Pulla et al., 2008). To our knowledge, the up-regulation of MtCAS31 under drought stress shows the most dramatic response to abiotic stress among DHNs currently documented. According to the Medicago truncatula Gene Expression Atlas, MtCAS31 is highly expressed in the seed, and in roots treated with NaCl (Benedito et al., 2008), and MtCAS31 is indicated to be involved in resistance to abiotic stresses in M. truncatula.

Overexpression of MtCAS31 dramatically enhanced the drought resistance of mature transgenic Arabidopsis both above- and below-ground, but seed germination was not significantly enhanced (data not shown). Line 2 showed the highest expression of MtCAS31 and the highest drought resistance. These results demonstrate that overexpression of MtCAS31 can enhance the drought tolerance of transgenic plants, which is consistent with the effect of other DHNs (Brini et al., 2007; RoyChoudhury et al., 2007).

The stomatal aperture was not significantly different between PCaMV35S:MtCAS3-transgenic and wild-type Arabidopsis (Fig. S5). After 10 μM ABA treatment, no differences in stomatal closure were observed between PCaMV35S:MtCAS31 Arabidopsis and the wild type (Fig. S5). These results suggest that the higher drought tolerance of PCaMV35S:MtCAS31 Arabidopsis is unrelated to stomatal aperture. That a lower stomatal density confers higher drought resistance has been demonstrated previously (Ouyang et al., 2010; Yoo et al., 2010; Wang et al., 2011a). Consistent with these studies, higher drought resistance in Arabidopsis was accompanied by fewer leaf stomata in the present investigation. This finding indicates that MtCAS31 overexpression either directly or indirectly reduced the stomatal density and stomatal index to confer increased drought tolerance in PCaMV35S:MtCAS31 Arabidopsis. To our knowledge, this is the first report of a DHN gene affecting stomatal density.

Recently, it was suggested that AtICE1 interacted with AtSPCH, AtMUTE, and AtFAMA to direct the sequential transition of three cell-states during stomatal development (Kanaoka et al., 2008; Serna, 2009). Combination of AtICE1 and AtSPCH is weaker than that of AtICE1 with AtMUTE or AtFAMA (Kanaoka et al., 2008). The interaction of MtCAS31 and AtICE1 is assumed to affect AtICE1 to form heterodimers with AtSPCH rather than interact with AtMUTE or AtFAMA. AtSPCH is essential for initiation of stomatal development in Arabidopsis (MacAlister et al., 2007), which is a likely explanation of why MtCAS31 reduced stomatal density in PCaMV35S:MtCAS31 Arabidopsis without expression of a heteromorphic stomatal phenotype. Just as MtCAS31 interacted with AtICE1 to affect the AtICE1-regulated downstream AtMYB15 and AtCBF3 genes, the interaction between MtCAS31and AtICE1 may alter the formation of heterodimers between AtICE1 and AtSPCH to impact on stomatal development. To better understand how MtCAS31 participated in stomatal development, the interaction of MtCAS31 with AtSPCH, AtMUTE and AtFAMA was investigated. The results suggested that MtCAS31 could not directly interact with AtSPCH, AtMUTE or AtFAMA in yeast (data not shown). According to previous research and the present study, we generated a model for MtCAS31 involvement in the development of stomata (Fig. 9). This model predicts that interaction between MtCAS31 and AtICE1 impacts on formation of heterodimers by AtICE1 and AtSPCH during establishment of the stomatal cell lineage, and causes a reduction in stomatal density to increase the drought resistance of transgenic Arabidopsis. Interaction of the AtICE1 homologous protein Mt7g083900.1 with MtCAS31 indicated that similar regulatory mechanisms are present in M. truncatula and Arabidopsis.

Figure 9.

Model of relationship among AtICE1 (Arabidopsis thaliana inducer of CBF expression 1), MtCAS31 (Medicago Truncatula cold-acclimation specific protein 31) and AtSPCH. In wild-type Arabidopsis, AtICE1 and AtSPCH (Arabidopsis thaliana SPEECHLESS) form heterodimers to direct the initiation of stomatal development. In PCaMV35S:MtCAS31 Arabidopsis, MtCAS31 could interact with AtICE1. This interaction might impair formation of AtICE1 and AtSPCH heterodimers, and thereby impede initiation of stomatal development.

Decreased stomatal density was one factor underlying the enhanced the drought resistance in PCaMV35S:MtCAS31 Arabidopsis. It had previously been found that increased expression of AtMYB15 could enhance the drought tolerance of transgenic Arabidopsis (Ding et al., 2009). Meanwhile, AtCBF3 was involved in response to low temperature and abscisic acid. (Maruyama et al., 2004; Novillo et al., 2007). The expression of AtMYB15 was promoted in PCaMV35S:MtCAS31 Arabidopsis, which may be another possible explanation for the increased drought resistance in PCaMV35S:MtCAS31 Arabidopsis. Moreover, relative expression of some stress response genes was detected. A shown in Fig. S6, expression of AtCOR15A, AtCOR47 and AtRD29A in PCaMV35S:MtCAS31 Arabidopsis was higher than that in wild-type Arabidopsis after treatment with PEG-8000 for 6 h (Fig. S6). This result indicated that the increased expression of these stress-response genes contributed to the increased drought resistance of PCaMV35S:MtCAS31 Arabidopsis. The results of the yeast two-hybrid assay indicated that some other putative proteins could interact with MtCAS31 to increase the drought tolerance of PCaMV35S:MtCAS31 Arabidopsis, such as dehydroascorbate reductase and tonoplast intrinsic protein. Dehydroascorbate reductase might provide antioxidant protection for rice during dehydration (Ji et al., 2011). Tonoplast intrinsic protein has been studied in relation to water stress (Sade et al., 2009; X. Wang et al., 2011). However, other mechanisms by which MtCAS31 enhances the drought tolerance of plants need further study.


This work was supported by the National Natural Science Foundation of China (30671328) and the Hi-Tech Research and Development (863) Program of China (2009AA10Z108 and 2011AA100209). We thank Dr Jean Marie Prosperi and Magalie Delalande (Biological Resource Center for M. truncatula, UMR 1097, INRA, Montpellier, France) for providing seeds of M. truncatula A17. We also sincerely thank Dr Jie Le (Center for Signal Transduction and Metabolomics, IBCAS, China) for his suggestions about stomata.