Introduction

During growth and development, plants are inevitably subjected to low temperatures, a main abiotic stress factor decreasing product yields. Many species are able to survive chilling or freezing environments through a series of response and recovering mechanisms in plant itself such as synthesis of cold-responsive (COR) and antifreeze proteins (AFPs) [1, 2], changes of membrane composition and structure [3], phospholipid desaturation [4], increase of small moleculars and osmotic pressure [5], as well as the participation of various secondary metabolites [6]. All these above mechanisms rely in part on gene regulation. At present, the C-repeat-binding factors (CBF) pathway is deemed to one main genetic way, which induces about 12–20 % relative genes in cold response reactions in Arabidopsis thaliana [7].

Since reported in A. thaliana, the CBF pathway has been generally understood [8, 9]. In spite of work specialization between CBF1-3, the transduction of cold signal in plants follows calcium channel [10], phosphorylation and sumoylation of ICE (inducer of CBF expression) [11], activation of CBFs and regulation of COR genes [12]. To date, ICE1 is the most upstream transcription factor in the CBF pathway, however, being constitutively expressed in Arabidopsis [12]. The transcripts of CBFs are induced a great degree by cold stress. Then through binding the C-repeat/dehydration-responsive (CRT/DRE) elements of promoters, CBFs activate many downstream COR genes such as COR15a, LTI78, COR47, BIN2, ERD10 and so on [13]. Thus, via CBFs, the cold response signal is much amplified. As much as 79 % metabolite changes elicited during cold hardening were AtCBF3 dependent [9]. Thus, CBFs may be major components and signal intersections of cold hardening in plants.

Jatropha curcas L., is a multipurpose, perennial shrub tree belonging to the family Euphorbiaceae, with the greatest reputation mainly from its biofuel potential [14, 15]. In China and India, the cultivation and exploration of Jatropha is a national development strategy. As a tropical and subtropical plant distributing wildly in the middle regions of Asia and Africa, J. curcas is sensitive to cold stress. In regions of high altitude (>1,700 m) or latitude (>28.5°), chilling temperatures becomes the bottleneck for cultivation and commercialization of J. curcas. For instance, in Xinyang city of Henan province (latitude about 29.0°N) of China, Jatropha seedlings introduced from the south were eventually frozen to death by the frost climate in winter. So the chilling tolerance improvement becomes the urgent issue for Jatropha.

We have previously studied network reactions of this plant to cold stress. Among its molecular web of cold response, CBFs transcription levels rose most dramatically, up to hundred folds above the control. And JcCBF2 might play indispensable roles for chilling tolerance in J. curcas [16]. However, whether and how the JcCBF2 gene functions remain unverified. So in the present report, JcCBF2 is further studied through conserved element analysis, ectopic expression and the regulation activity to confirm its real roles.

Materials and methods

Materials and treatments

The J. curcas seeds harvested from Liangshan of Sichuan province were germinated in greenhouse conditions (24000 Lx, 60 % humidity, 25 °C, 16 h; dark, 80 % humidity, 18 °C, 8 h) as described by Gao et al. [6]. Two months old seedlings were cultivated in 4 °C chilling stress and their second leaves were selected to extract nucleic acids along with stress duration of 0–48 h.

Wild Col-0 and transgenic A. thaliana seeds of T2 generation were germinated in moist filter papers and cultured in peat blended with vermiculite (4:1). The culture condition was as above with exception that it’s 22 °C during the day. The 20 day old seedlings were freezed at −2 °C (24000 Lx, 60 % humidity). After 0, 1, 3, 6, 12, 24 and 48 h durations of treatment, the A. thaliana plants were recovered 3 days in normal temperature. Then the entire seedlings were collected for physiological measurements and RNA purification.

Isolation of JcCBF2 and sequence analysis

The genomic DNA and total RNA were isolated using the DNA secure Plant Kit and RNAprep pure Plant Kit (TIANGEN, Beijing), respectively. The first-strand cDNA was synthesized using the Quant cDNA First-Chain Kit (TIANGEN, Beijing). Then primers Jc-p-S and Jc-p-A (Table 1) were designed according to sequence regions of JcCBF2 (No: Jcr4S02476.10) from Jatropha Genome Database (http://www.kazusa.or.jp/jatropha/) and used to get the JcCBF2 open reading frame (ORF) fragment with J. curcas DNA and cDNA as templates. Both purified products were subcloned into pMD-19T vectors (TakaRa, Dalian, China) and sequenced (Invitrogen, Shanghai, China).

Table 1 Primers used in the study
Full size table

The sequenced nucleic acids from DNA and cDNA of J. curcas were checked in the genome database on line (http://www.kazusa.or.jp/jatropha/) and compared with each other to analyze intron and exon structure of JcCBF2. Afterwards the homologous similarity of JcCBF2 was analyzed in NCBI by BlastP on line (http://blast.ncbi.nlm.nih.gov/). Typical conserved and core functional motifs for CBFs were predicted on line (BCM, http://www.hgsc.bcm.tmc.edu/; BioNavigator, http://www.bionavigator.com/) and in line with literatures [17, 18]. Based on similarity assignment by Clustal W and MEGA 4.0 soft packages, the phylogenetic analysis was performed in the way of neighbor-joining (NJ) method.

Vector construction and plant transformation

For over-expression assays, the digestion product of JcCBF2 ORF from PMD-19T-JcCBF2 was ligated into XbaI-SmaI sites of pBI121 expression vector with the 35S promoter (Novagen). The 35S::JcCBF2 was then transformed into cells of Agrobacterium tumefaciens GV3101 by freeze–thaw treatment. The transgenosis steps was performed according to the floral dipping method with little adjustment [19].

To locate the JcCBF2 protein in A. thaliana, primer Jc-K-S and Jc-K-A was designed to amplify the JcCBF2 with endonuclease sites XhoI and PstI (Table 1). Then the PCR product and transient expression plasmid KNTG were used to construct the report vector 35S::JcCBF2-GFP with a green fluorescent protein (GFP). And the recombinant plasmid was transformated into wild A. thaliana.

After antibiotic screening by kanamycin in 1/2 MS solid medium, the T2 generation transgenic A. thaliana seedlings were transferred into the soil medium. PCR and sequencing analysis was further carried out to confirm the integration and expression of the transgenes.

Protein location in cellular level

Shortly after seed germination in dark, fresh white hypocotyls and radicles (<1 cm) were taken from 35S::JcCBF2-GFP transgenic A. thaliana (T2 generation) sprouts and used as samples. On the glass slides, the materials were soaked in sterile water and observed under OLYMPUS 1X71 fluorescence microscope with green fluorescence light.

Transcriptional activation assay

For transcriptional activation assay, the full open reading frame (ORF) and a series of deleted JcCBF2 sequences were generated by PCR and fused in frame to the GAL4 DNA-binding domain in pGBKT7 by recombination reactions (Invitrogen). The empty pGBKT7 vector was used as a negative control. Then these constructs were transformed into yeast strain AH109. The transformants were streaked on the SD/Trp− and SD/Trp−/His − medium. After incubation at 28 °C for 3 days, the growth status of the transformants was evaluated. The colony-lift filter assay (β-galactosidase assay) was performed as described by the manufacturer (Invitrogen).

QRT-PCR analysis of gene transcriptional level

To study transcription profiles of JcCBF2, primers Q-S and Q-A were designed (Table 1). Then the leaves, stems and roots of chilled Jatropha seedlings were all tested by quantity RT-PCR (QRT-PCR) method. The progress was conducted on a Bio-Rad iCycler MyiQ Real-Time PCR System in a 20 μl volume containing appropriate cDNA from each sample, 10 μl of SsoFastTM EvaGreen Superemix (Bio-Rad, USA), 1.0 μl of each primer (Table 1) and 7 μl doubly distilled H2O. The PCR profile included one cycle of 95 °C for 30 s, 40 cycles (95 °C for 5 s, 60 °C for 20 s) and a final melt curve profile (65–95 °C, 0.5 °C/S). Quantification was determined with ΔΔCT method. Each data represents the average of three repeats and QRT-PCR was done in triplicate.

In transgenic A. thaliana, to exploit responses of JcCBF2 and other related genes, long amplification of semi-QRT-PCR was carried out after freezing process. And the primers for JcCBF2, RD29A, COR15A and COR6.6 were designed, the β-actin gene being the control reference (Table 1). Following the optimization of PCR programs, products were collected after 27, 25, 29, 28 and 26 reaction cycles respectively. And 5 μl mixture was used for product detection by 1.5 % agarose gel and stained with ethidium bromide for UV visualization. Samples represented three plants and PCR progress was repeated five times.

Determination of physiological indexes

Arabidopsis thaliana seedlings were tested at 10 am after freezing stress and growth recovering. Leaf conductivity, malondialdehyde (MDA) content and superoxide dismutase (SOD) activity of entire plants were measured as described [20].

Results

Isolation of JcCBF2 and its expression

Both cDNA and DNA of J. curcas exposed in chilling stress were used, respectively, to amplify sequences of JcCBF2 gene, with the same primer pair Jc-p-S/Jc-p-A. After assignment between the two PCR product, no difference of nucleotide type and length was found, indicating that the JcCBF2 gene had no intron just like other CBFs [17, 21]. The ORF of JcCBF2 contains 693 bp nucleotides, encoding a 230 amino acid (AA) chain with a calculated molecular mass of 26.28 kDa and theoretical isoelectric point of 5.54.

Using primers Q-F and Q-R, QRT-PCR tests displayed that JcCBF2 was expressed mainly in leaves, hardly any in stems and roots (Fig. 1a). After 12 h of 4 °C chilling treatment, the transcriptional level of JcCBF2 raised maximum 118 folds than the control, suggesting its drastic response to chilling stress in J. curcas (Fig. 1b). Obviously, the JcCBF2 gene is expressed tissue-specifically and inducible by low temperature in J. curcas.

Fig. 1
figure 1

Transcriptional profiles of JcCBF2 in different tissues (a) and leaves (b) of chilled J. curcas The second leaf in 4 °C chilling stress was tested by QRT-PCR. Values represented the mean ± SE of three replicates, not significantly different (p < 0.05)

Full size image

Sequence and structure analysis of JcCBF2

Screened in NCBI on line through association of BlastP and Specialized Blast bl2seq, JcCBF2 was found to share 100.0 % similarity with Ricinus communis DREP1 (XP_002509703.1), Malus domestica CBF/DREB2 (AGL07698.1) and Prunus persica DREB (ABR19831.1), more than 98 % similarity with DREB68 from Populus hopeiensis (AGH33879.1), CBF1 from Morus alba var. multicaulis (AFQ59977.1) and CBF from Ageratina adenophora (ABN58746.1). Moderate similarity was found between JcCBF2 and Betula pendula CBF2 (ABP98988.1), Vitis riparia CBF (AAW58104.1) and Eucalyptus gunnii CBF (ABB51638).

Then based on Clustal W and MEGA 4.0 software package, the phylogenetic tree was performed using the NJ method (Fig. 2). The dendrograms showed that JcCBF2 was clustered closely with RcDREB1 (XP_002509702.1) from R. communis and CBFs (ABP64695.1, CBJ94893.1) from Populus. In spite of high similarity with JcCBF2, CBFs of A. thaliana clustered in another separate group, with large genetic distance. The connotation of graphical dendrograms was similar with consequences resulted from other Jatropha genes [22, 23].

Fig. 2
figure 2

Phylogenetic analysis of JcCBF2 MEGA4.0 soft package was used with neighbor-joining (NJ) method. Species and their accession number in GenBank were following: AtCBF1, A. thaliana CBF1 (AAC49662); AtCBF2, A. thaliana CBF2 (AAD15976); AtCBF3, A. thaliana CBF2 (ABV27138.1); AtCBF4, A. thaliana CBF4 (ABV2755.1); AaCBF, Ageratina adenophora CBF (ABN58746.1); NtDREB4, Nicotiana tabacum DREB4 (ACE73696.1); LhVBF1, Lycopersicon hirsutum LhCBF1 (BAE17131.1); StVBF5, Solanum tuberosum CBF5 (ACB45083.1); VvCBF, Vitis vitifera CBF (AAG59618); BpCBF2, Betula pendula CBF2 (ABP98988.1); MaCBF1, Morus alba var. multicaulis CBF1 (AFQ59977.1); EguCBF, Eucalyptus gunnii (ABB51638); PaDREB1, Prunus avium DREP1 (BAC20184.1); PdCBF2, Prunus dulcis CBF2 (AFL48191.1); MdCBF2, Malus domestica CBF/DREB 2 (AGL07698.1); PmCBF, Prunus mume CBF (AFX97738.1); RcDREB1, Ricinus communis DREB1 (XP_002509702.1); PhDREB68, Populus hopeiensis DREB68 (AGH33879.1); PtCBF4, Populus trichocarpa CBF4 (ABP64695.1); CtCBF1, Citrus trifoliata CBF1 (ABH08746.1); RcDREP1, Ricinus communis DREP1 (XP_002509703.1); PbCBF2, Populus balsamifera CBF2 (CBJ94884.1); PtCBF2, Populus trichocarpa CBF2 (CBJ94893.1)

Full size image

Depended on literatures and online prediction, conservative motifs of JcCBF2 were diagrammatized (Fig. 3). Results revealed this short chain consisted of the core AP2 element (consensus structure of AP2 transcription factor super-family), N-terminal signal peptide and C-end characteristic components. In AP2 sequence, the fourteen (Valine, Val) and nineteen (Glutamic acid, Glu) are relatively conservative in the CBF family [17, 21]. What’s more, JcCBF2 had typical amino acid motifs, KKPAGRKKFRETRHP (CLUSTER 1), DSAWR (CLUSTER 2) and other three conservative ones (CLUSTER 3–5), which were characteristic of CBF transcription factors. Among them, the element CLUSTER 1 and 2 are unique to CBF transcription factor, the former locating in the N side of their protein chains, functioning as signals for nuclear localization [24]. Amino acid A is often candidate of X1 and X2 in CLUSTER 3 [21, 25], so being it in JcCBF2 (Fig. 3).

Fig. 3
figure 3

Multiple sequence alignment and conserved elements of JcCBF2 Clustal W software was used. Shaded zones were amino acids with not less than 50 % similarity. Inside the black box and asterisk marked were conserved motifs or sequences for CBFs. Underlined sequences were the AP2 domain (ethylene response element binding factor/APETELA2 DNA binding motif). Abbreviations for plant species and their sequence accession numbers in GenBank were as Fig. 2

Full size image

Summarized from similarity of nucleotides and amino acids, as well as the deduced functional elements, JcCBF2 is most likely to be a CBF family factor protecting J. curcas leaves from stress injury in cold environment.

Transcriptional activation analysis of JcCBF2

To investigate the transcriptional activation ability of JcCBF2, we constructed a series of deletion sequences of JcCBF2 fused to the DNA -binding domain of pGBKT7 plasmid and transformed them into yeast strain AH109 (Fig. 4a). The deletion in the C-terminal region abolished the transactivation activity (Fig. 4b). These results indicated that JcCBF2 had the transcription activated ability.

Fig. 4
figure 4

The transcriptional activation analysis on JcCBF2. a Schematic construction of serial deletion of JcCBF2 fused to the GAL4 DNA-binding domain in the yeast vector pGBKT7 for transactivation assay. b Transactivation analysis of the full, N-terminal or C-terminal of JcCBF2 in yeast AH109. The transformants were streaked on the SD/Trp− and SD/Trp−/His− medium. The plates were incubated and subjected to a β-gal assay

Full size image

Transcription assay of JcCBF2 in frozen A. thaliana

To confirm the roles predicted by sequence and element analysis, the 35S::JcCBF2 was transformed into wild-type Col-0 A. thaliana. While the wild and transgenic seedlings of T2 generation were stressed in −2 °C condition, transcripts of JcCBF2 were tested by semi-QRT-PCR. The results displayed that after 1 h freezing treatment, JcCBF2 transcripts in transgenic A. thaliana increased rapidly, and maintained the high level until 6 h. Hereafter its expression declined gradually (Fig. 5a). The data indicated that JcCBF2 was indeed existed and might play considerable roles in early response to frozen stress in transgenic A. thaliana.

Fig. 5
figure 5

Damage and responses of A. thaliana (T2 generation) to freezing stress (a) was semi-QRT-PCR results (the PCR cycles were within parentheses), 0–48 represented stress time in −2 °C freezing condition. β-ACTIN (GenBank no GR991088) was the reference gene. Primers of each gene were described in Table 1. b was the recovered seedlings after 48 h freeze stress; 35S::JcCBF2, Wild and pBI121 represented respectively transgenic A. thaliana with JcCBF2, wild and transgenic A. thaliana with pBI121. c was the MDA content, leaf conductivity and SOD activity of whole seedlings of freezed A. thaliana

Full size image

After continuous 48 h freezing and 3 days recovery, the leaves of wild-type and control (with pBI121 plasmid transgenic) plants partly shriveled, faded green to gray and white (Fig. 5b). Leaf physiology study displayed that the MDA content, conductivity and SOD activities of these seedlings changed 1–3 folds during the freezing treatment (Fig. 5c). Compared with the wild-type and control seedlings, the 35S::JcCBF2 transgenic plants grew as well as the original state (Fig. 5b). In addition, the JcCBF2 transgenic lines conducted less MDA (50 %), lower leaf conductivity (50 %) and higher SOD activities (175 %), indicating that they were in a better growth state with existence of JcCBF2 (Fig. 5c). Results suggested that JcCBF2 could evidently increase the freezing tolerance of A. thaliana.

Functional orientation of JcCBF2

To study where JcCBF2 plays his role, a 35S::JcCBF2-GFP report protein was expressed ectopically in A. thaliana bud. In white hypocotyls and radicles under green fluorescence, the JcCBF2-GFP protein was observed mainly in nucleus (Fig. 6). Results concluded that JcCBF2 might function as an expression-regulator for cold response genes, which had been described in other plants [1, 12].

Fig. 6
figure 6

Cellular localization of JcCBF2-GFP recombinant protein (a) was structure diagram of recombinant 35S::JcCBF2-GFP sequence. (b) was fluorescence microscopic photography of GFP protein. The samples were radicles and hypocotyls of transgenic A. thaliana. The above were pictures in white light, followed by green fluorescence below

Full size image

To avoid the infection of chloroplast fluorescence, the orientation observation was done using hypocotyls and radicles. Interestingly, the ectopic expression of JcCBF2-GFP was observed clearly in heterologous plant system.

Transcripts of freezing-related genes in transgenic A. thaliana

How CBF factors participate in the freezing tolerance of plants. In Arabidopsis, CBFs have been demonstrated to regulate the transcription of many downstream genes involved in cold stress, such as RD29A, COR15A and COR6.6 and so on [26]. In our study, these three genes were not detected by semi-QPCR assay in normally growing wild-type Arabidopsis (Fig. 7). However, in 35S::JcCBF2 transgenic Arabidopsis plants they were transcribed constitutively. During freezing treatment, their transcripts were all induced in wild and transgenic seedlings, as other papers reported [27]. In wild plants, RD29A and COR6.6 were activated after 6 h of freezing treatment. But in transgenic plants their transcripts rise was detected rapidly from 1 h, lasting until 3–6 h. Then after 6 h, the transcriptional activities of COR genes seemed to reduce a little.

Fig. 7
figure 7

Transcriptional responses of three COR genes to freezing stress in A. thaliana Freeze stress condition was 24000Lx, 60 % humidity and −2 °C. 0–48 h exposed seedlings were recovered in normal environment for 3 days. The cDNA from total RNA of whole seedlings was used by semi-QRT-PCR (the optimal cycle numbers within the parentheses). Transgenic was transgenic A. thaliana with JcCBF2 gene, while Wild being wild Col-0. Genes’ accession numbers in GenBank were: RD29A, NM124610; COR105A, AY057640; COR6.6, X55053

Full size image

It was noticed that in wild-type Arabidopsis, COR15A transcripts were induced also within 1 h of freezing exposure, probably by CBFs in Arabidopsis itself (Fig. 7). However, in JcCBF2 transgenic Arabidopsis, COR15A transcripts rise faster and higher, obviously due to induction of JcCBF2. It is thus clear that, during responding to freezing environment, JcCBF2 is able to rapidly increase reactions of multiple cold regulator genes.

Discussion

Conserved motifs of JcCBF2 specific to cold responses

In the AP2/ERF super-family, CBF factors were reported firstly in A. thaliana. These family genes contain pieces of conserved sequences which could recognize and bind the CRT/DRE of COR genes. Except AP2 DNA-binding domain, several motifs have been summarized and studied as showed in Fig. 2. These domains are vital components for function switch in CBFs. It is noticed that two AP2/ERF factors, JcERF and JcDREB, were cloned from J. curcas previously and proved to participate in cold-response in Arabidopsis [28, 29]. However, their amino acid chains have little sequence similarity with JcCBF2, except for an AP2 element (Supplemental Fig. S1). In Arabidopsis, the AP2/ERF family consists of more than 100 members. They respond to different environment signals such as salty, drought, ABA and low temperature [30, 31]. In addition, one AP2/ERF factor could respond to a few external signals. In Jatropha, JcERF was able to response to salt and cold. And JcDREB could be induced by low temperature, salt and drought. So, JcERF and JcDREB play more colorful roles than JcCBF2 in J. curcas. Given the chilling response network, they might well just be passive participants rather than the dominant animateur for cold tolerance. On the contrary, with various motifs binding to promoters of COR genes, JcCBF2 is more like a master and positive actor to transmit the cold signals from environments.

In the early stage of freezing stress JcCBF2 might perform its roles

In this present study, in transgenic seedlings the period when the transcription level of three COR genes were acutely induced was 1–6 h during freezing stress. After the early phase, transcripts of COR genes in transgenic samples were not more than that in wild seedlings (Fig. 7). These data matched results from expression profiles of JcCBF2 itself in frozen Arabidopsis seedlings (Fig. 5a). Furthermore, no DRE/CRT cis-acting element, G/ACCGAC, was screened in the promoter of JcCBF2, indicating that it could not auto-regulate itself [32, 33]. So it is concluded that during early stages of freezing responses, JcCBF2 is induced rapidly. As a member of the reaction chain, whereafter, it quickly increases the reactions of multiple COR genes. After activation of downstream reactions, the functions of JcCBF2 were weakened, probably because of the repression of other factors [34].

This regulating style resembles other CBFs. In Arabidopsis, the transcriptional levels of CBF1-3 genes increased within 15 min of low temperature stress, followed by accumulation of COR gene transcripts at about 2 h [13]. In Lolium perenne, its LpCBF3 was responsive to cold stress mainly from 15 min to 2 h, and was reduced transcripts from 4 h of cold stress [33]. It is easy to understand the rapid efficiency of CBFs. As a small molecular of 26 kDa with just 230 AAs, JcCBF2 has convenience to synthesis, transport, binding to target gene and even degradation. Besides, the action earlylization of CBFs is accord with their positions in a wide response network to environment stresses.

More than cold responses, plus functions might JcCBF2 play

Through a GFP-marked plasmid, JcCBF2-GFP, the JcCBF2 protein was observed in hypocotyls and radicles of tender Arabidopsis seedlings (Fig. 6). It suggests that this novel transcription factor may be existed in the universal tissues of heterologous system. The results are supplementary to data from JcCBF2’s expression profiles in senior Jatropha plants (Fig. 1a). CBFs often express their proteins in plant leaves to contend with stress factors, which is always there in harsh environment.

Interestingly, in our late study, JcCBF2 is revealed to react to drought and high temperature stress in a transcription-positive way (Data not shown). The probable negative regulation of CBFs was also reported in Arabidopsis [35]. CBF is a multi-gene family in A. thaliana, which contains at last 4 members [25]. Exposed in the stress-comprehensive nature, CBFs are both relatively isolated and cooperative with each other. For instance, although AtCBF2 being the major responder to low temperatures, AtCBF1, AtCBF3 and AtCBF4 are also sensitive to cold. As the independent, AtCBF4 is particularly dedicated to the ABA signal reaction pathway rather than other members. In our study, the transcriptional activities of JcCBF2 would be down-regulated when Jatropha seedlings are stressed by salt or high temperature. So it is speculated that JcCBF2 plays a bit part in the network of CBF family while responding to drought/high temperatures. So the deeper function insights might partly depend on equivalent researches of other CBFs in J. curcas L.