Functional characterization of two almond C-repeat-binding factors involved in cold response PDF
Preview Functional characterization of two almond C-repeat-binding factors involved in cold response
Tree Physiology 32, 1113–1128 doi:10.1093/treephys/tps067 Research paper Functional characterization of two almond C-repeat-binding factors involved in cold response D o w n lo Pedro M. Barros1,2, Nuno Gonçalves1,2, Nelson J. M. Saibo1,2 and M. Margarida Oliveira1,2,3 ad e d fro m 1Genomics of Plant Stress Laboratory (GPlantS), Instituto de Tecnologia Química e Biológica, Universidade Nova de Lisboa, Av. da República, 2780-157 Oeiras, Portugal; 2IBET, h Apartado 12, 2781-901 Oeiras, Portugal; 3Corresponding author ([email protected]) ttp s Received December 27, 2011; accepted June 8, 2012; published online July 25, 2012; handling Editor Chunyang Li ://a c a d e m Low temperature plays a crucial role in seasonal development of woody plants and may directly impact crop production, more ic particularly in temperate fruit trees. Given its high genetic variability and adaptability to different climatic conditions, almond .ou p (Prunus dulcis Mill.) is an interesting model to understand the mechanisms regulating low temperature sensing in fruit trees. In .c o m this paper, we report the cloning and characterization of two genes (PdCBF1 and PdCBF2) belonging to the C-repeat-binding /tre factor (CBF) family of transcription factors. Southern blotting analysis showed that this family is composed of at least five mem- e p h bers. In almond shoots propagated in vitro, transcription of these genes was rapidly induced by low temperature, suggesting an y s involvement in cold acclimation. Transactivation assays showed that PdCBF1 and PdCBF2 could bind to dehydration responsive /a rtic element/C-repeat containing sequences, as activators of gene expression. In addition, induction of both PdCBFs by cold was le -a higher towards the end of the day, which agreed with the expression pattern of PdDehydrin1, a predicted CBF target gene. b s Furthermore, PdCBF1 and PdCBF2 were also transiently induced by abscisic acid and drought treatments. Considering the bin tra c mapping analysis that correlated PdCBFs and PdDHN1 (respectively in linkage groups 5 and 7) with two different quantitative t/3 2 trait locicontrolling blooming time, it is relevant to perform further association studies that may validate their effect on this trait. /9 /1 1 1 3 Keywords: bin mapping, CBF transcription factors, Dehydrin1, gene expression, Prunus dulcis, transactivation. /1 6 5 2 1 6 3 b y g u Introduction es specificgenesandaccumulationofcryoprotectiveproteinsand t o n Cold stress is a major environmental stress that limits crop pro- other metabolites involved in cell homeostasis (Thomashow 1 0 ductivity in cultivated areas. Plant species may have different 1999, Nakashima and Yamaguchi-Shinozaki 2006). The A p levels of susceptibility either to chilling (0–15 °C) or freezing C-repeat-binding factor (CBF)/dehydration-responsive ele- ril 2 0 (<0 °C). Nevertheless, they can increase their freezing toler- ment-binding factor 1 (DREB1) family of transcription factors 1 9 ance upon exposure to low non-freezing temperatures, in a pro- (TFs) mediates one of the best-studied mechanisms of low- cess known as cold acclimation or cold hardening (Levitt 1980). temperature signalling in plants (Yamaguchi-Shinozaki and Perennial trees from temperate or boreal environments are fre- Shinozaki 2006, Medina et al. 2011).TheseTFs,firstidentified quently exposed to chilling and freezing conditions, particularly in Arabidopsis (Stockinger et al. 1997, Liu et al. 1998), contain during fall and winter dormancy. In fact, maximal cold hardiness a highly conserved AP2/ERF (APETALA 2/ethylene responsive levels are achieved during winter dormancy by a sequential pro- factor) domain, which acts as a DNA-binding motif, and an cess induced by photoperiod and/or temperature conditions acidic C-terminal sequence conferring transacting activity (Weiser 1970, Welling and Palva 2006, Tanino et al. 2010). (Stockinger et al. 1997, Medina et al. 2011). Transcriptional Cold acclimation involves extensive physiological and induction of CBFs occurs rapidly after temperature decrease, molecularmodifications,suchastranscriptionalmodulationof and this is closely followed by the increased expression of their © The Author 2012. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected] 1114 Barros et al. target genes (the CBF regulon) which play a fundamental role In addition to the role in seasonal cold acclimation, the pro- in cold acclimation. CBF expression was also shown to be longed exposure to cold temperatures during winter also deter- modulated by light quality and the circadian clock, during cold mines the timing of dormancy break (Metzger 1996). In response or, to a lower extent, in control conditions (Fowler Rosaceae fruit trees this timing affects the extent of the growing et al. 2005, Franklin and Whitelam 2007, Bieniawska et al. season and also crop production. Thus, understanding the 2008). Several reports have also shown transient induction of mechanisms underlying cold signalling and response in these CBFs in response to abscisic acid (ABA), drought or mechani- plants is crucial. Almond (Prunus dulcis Mill.) is the earliest fruit cal stress (Gilmour et al. 1998, Knight et al. 2004, Yamaguchi- tree to bloom in winter/spring, having low chilling requirements Shinozaki and Shinozaki 2006). for dormancy break (Arús et al. 2009). However, this species The CBF-mediated cold signalling pathway seems to be a has a wide adaptability to different climatic conditions (Alonso conserved mechanism among distantly related species. et al. 2005) associated with its high genetic variability, thus D C-repeat-binding factor TFs have been targeted in a wide vari- being an interesting model to study low-temperature response in ow n ety of herbaceous and woody plants (Welling and Palva 2006, Prunoideae. Considering that the CBF-mediated signalling path- lo a Medina et al. 2011), deepening the knowledge related to varia- way is still poorly understood in this species, and in general de d tions in freezing tolerance occurring within different plant spe- within the Prunusgenus,inthispaperwedescribetheidentifica- fro m cies throughout evolution. Factors like the expression level of tion and functional characterization of two almond CBF genes. h different CBF members and the size and diversity of the CBF ttp s regulon have been suggested as crucial in this natural variabil- Materials and methods ://a c ity (Zhang et al. 2004, Hannah et al. 2006, Champ et al. 2007, a d Badawi et al. 2007, Carvallo et al. 2011). Concerning woody Plant material em ic species, the involvement of CBF genes in cold acclimation has Almond shoots were obtained from in vitro axillary budding of .o u p been studied in sweet cherry (Prunus avium L.) (Kitashiba et al. adult cv. Verdeal. Shoots were subcultured every 3 weeks on .c o 2004), birch (Betula pendula Roth) (Welling and Palva 2008), micropropagation medium as described by Miguel et al. (1996) m poplar (Populus trichocarpa Torr. & Gray) (Benedict et al. and maintained under 16/8 h photoperiod at 22/20 °C. Genomic /tre e p 2006), eucalyptus (Eucalyptus gunnii Hook.f. and Eucalyptus DNA was extracted as described by Martins et al. (2003). RNA h y s globulus Labill.) (El Kayal et al. 2006, Gamboa et al. 2007, was extracted using the RNeasy® Plant Mini Kit (Qiagen, Valencia, /a Navarro et al. 2009), grape (Vitis spp.) (Xiao et al. 2006, CA, USA), treated with RNAse-free DNase I (Qiagen) and quanti- rtic le 2008), citrus (Poncirus trifoliata (L.) Raf. and Citrus paradisi fiedusingNanoDrop(ThermoScientific,Wilmington,DE,USA). -a b s Macf.) (Champ et al. 2007), highbush blueberry (Vaccinium The Arabidopsis cell suspension line used for protoplast isolation tra c corymbosum L.) (Polashock et al. 2010), dwarf apple (Malus was maintained as described by Mathur and Koncz (1998). t/3 2 baccata (L.) Borkh.) (Yang et al. 2011) and apple (Malus x Protoplast isolation was performed according to Anthony et al. /9 /1 d omestica Borkh.) (Wisniewski et al. 2011). (2004). Genomic DNA from Arabidopsis thaliana (L.) Heynh., 1 1 3 Dehydrins (DHNs) are a sub-group of the late embryogenesis ecotype Columbia, was obtained from fully expanded leaves with /1 6 abundant proteins, widely studied in woody plants due to their DNeasy® Plant Mini Kit (Qiagen). 5 2 1 predicted role in protection against cellular dehydration and cold 6 3 acclimation (Welling and Palva 2006). The presence of C-repeat/ Identification of PdCBF-like sequences by g dehydration responsive element (CRT/DRE) motifs in the pro- TocloneCBF-likesequencesfromalmond,specificprimers u e s moter region of several cold-regulated DHNs suggests that CBF-F and CBF-R (Table 1) were designed based on the nucle- t o n these are predicted players in the CBF-mediated signalling path- otide sequence of sweet cherry PaCBF1 (accession no. 1 0 way (Puhakainen et al. 2004, Wisniewski et al. 2006, Bassett AB121674). Two-week-old propagated shoots were transfered A p et al. 2009). The precise function of these proteins remains to to a 5 ºC chamber and collected and frozen in liquid nitrogen ril 2 0 be determined, although several studies elucidated their cryo- at selected timepoints (30 min, 1 h, 2 h, 4 h, 8 h and 12 h). 1 9 protective and antifreeze properties (Wisniewski et al. 1999, Four micrograms of total RNA pool obtained from the cold Rorat 2006). Seasonal patterns of DHN expression appear to be treatment was used for cDNA synthesis with SuperScript® a common feature in woody plants (Welling and Palva 2008), First-Strand Synthesis System (Invitrogen, Carlsbad, CA, USA) namely in fruit species (Yamane et al. 2008, Bassett et al. 2009, following the manufacturer’s instructions. Polymerase chain Garcia-Bañuelos et al. 2009). In peach, three DHN genes were reaction (PCR) mixture was prepared in 15 µl volume using previouslyidentifiedandtwowereshowntobecold-responsive 1 µl cDNA as template and 1 U GoTaq® DNA Polymerase (Wisniewski et al. 2006, Bassett et al. 2009). More particularly, (Promega Corp., Madison, WI, USA) according to the manufac- PpDNH1 showed high levels of expression during fall and winter turer’s instructions. Samples were analysed by electrophoresis in bark tissues, in close correlation to the development of cold on 1% agarose gel. The predicted amplicons were excised hardiness (Artlip et al. 1997). fromgelandpurifiedusingtheWizard® SV Gel Clean-Up Tree Physiology Volume 32, 2012 CBF transcription factors in almond cold response 1115 Table 1. Sequences of the primers used for molecular cloning of PdCBF1, PdCBF2 and PdDHN1, RACE-PCR and inverse PCR. Genes Experiments Primer Primer sequence PaCBF Gene cloning CBF-F 5′-GCCCCAGTCGAGTTTGTTGTC-3′ CBF-R 5′-AGCATTGCGATGGAGAAAGAAG-3′ PdCBF1 3′-RACE CBF-3F 5′-CAAGTGGGTGTGTGAAATGAG-3′ 5′-RACE CBF1-5R1 5′-AATGAAAATCAAATTACAACGGTC-3′ CBF1-5R2 5′-TGGAGAAACTCCACAATTTGACATC-3′ Inverse PCR iPCR1-F 5′-GCTTGTCTGCCTCAACTTCC-3′ iPCR1-5R1 5′-GGACCAAGAAGCCCTTAGAGTC-3′ iPCR1-5R2 5′-CGGACAACGAACTTGACTCG-3′ PdCBF2 5′-RACE CBF2-5R1 5′-ACCCATGTCATCCAAATTCATGTC-3′ CBF2-5R2 5′-CAAACAATAATCGCTCGCACA-3′ D o Inverse PCR CBF2-5F 5′-ACTTTGCTGACTCCGCGTTG-3′ w n iPCR2-5R1 5′-AAGAAGGTCCCTGAGTCGTG-3′ loa d iPCR2-5R2 5′-GGGTCGGAAAGTTGAGACAAG-3′ e d – RACE Adp-R 5′-CCACGCGTCGACTAGTAC-3′ fro AdpPoliT-3R 5′-CCACGCGTCGACTAGTACT VN-3′ m (18) h AdpGGI-5F 5′-GCCACGCGTCGACTAGTACCGGGGIIGGII GGGIIG-3′ ttp s ://a c a d e System (Promega), cloned into pCR2.1 (Invitrogen) following regions (Supplementary Figure S1 available as Supplementary m ic the manufacturer’s instructions and sequenced (STABVida, Data at Tree Physiology Online). In silico restriction analysis was .o u p Caparica, Portugal). performed to search for enzymes that could cut at least once, .c o downstream the primer located towards the 3′ end. Two micro- m Rapid amplification of cDNA ends grams of almond genomic DNA were digested with 10 U NcoI /tre e p Full-lengthcodingsequenceswereobtainedbyrapidamplifi- and NotI for 6 h in 60 µl total volume at 37 °C, followed by h y s cation of cDNA ends (RACE-PCR) according to Frohman et al. enzyme inactivation by 10 min incubation at 65 °C. DNA circu- /a (1988),withminormodifications.MessengerRNAwaspurified larization was carried out for 16 h at 14 °C using 15 µl of the rtic le using the PolyATract® mRNA Isolation System III (Promega) digestion reaction in 100 µl ligation reaction with 4 U T4 DNA -a b s from the cold-induced total RNA pool. Sixty nanograms of this ligase (Invitrogen). Eight microlitres of each ligation reaction tra c mRNA were then used for cDNA synthesis using the wasusedinafirstPCRroundin100µl volume with either t/3 2 SuperScript® First-Strand Synthesis System (Invitrogen). The iPCR1-F/iPCR1-5R1 or PdCBF2-5F/iPCR2-5R1 primer combina- /9 /1 3′-RACE was performed according to Borson et al. (1992), tions (Table 1). Polymerase chain reaction was performed as 1 1 3 usingaspecificforwardprimer(CBF-3F)forbothgenes.For follows: 5 min incubation at 95 °C, followed by 30 cycles of /1 6 5′-RACE, cDNA was synthesized using CBF1-5R1/CBF2-5R1 90sat95°C,60sat56°Cand3minat72°C,andafinal 52 1 gene-specificprimersaccordingtoBekman et al. (2000). For extension step of 5 min at 72 °C. After this, a nested PCR was 6 3 theamplificationof5′ cDNA ends, PCR was performed using prepared with 4 µl of the previous reaction mixture using the by g gene-specificreverseprimersCBF1-5R2/CBF2-5R2andfor- same forward primers and the nested reverse primers iPCR1- u e s wardprimerspecificfora5′ anchor region previously synthe- 5R2 or iPCR2-5R2 (Table 1).Amplificationwasperformedas t o n sized.Allamplificationreactionswereperformedusing0.4U above, decreasing the cycling extension time to 150 s. 1 0 Phusion® High-Fidelity DNA Polymerase (Finnzymes Oy, Amplificationproductswereanalysed,clonedandsequenced A p Espoo, Finland) according to the manufacturers’ instructions. as previously mentioned. The 5′ genomic end was located ril 2 0 Amplificationproductswereseparatedbyelectrophoresison immediately downstream of the enzyme recognition site. 1 9 1%agarosegelsandfragmentsofinterestwerepurified, Transactivation analysis cloned and sequenced as mentioned above. All the primers used for 3′- and 5′-RACE are listed on Table 1. For transient expression and transactivation analysis two types of vectors were generated. Effector vectors were obtained by Promoter cloning cloning PdCBF1 or PdCBF2 coding sequences into pDONR221 ThepromoterregionsoftheidentifiedPdCBFs were targeted (Gateway®, Invitrogen) and recombined with the pK7WG2 des- using inverse PCR (iPCR), as described by Ochman et al. tination vector (Karimi et al. 2002),downstreamthecauliflower (1988). Specific primers were designed for each coding mosaic virus (CaMV) 35S promoter. Reporter vectors were sequence 5′ and 3′ ends, facing opposite orientation in a way obtained by cloning the promoter regions of almond Dehydrin1 that primer extension could occur outwards the known genomic (PdDHN1, −770 to −126 bp from ATG, accession no. JQ317156) Tree Physiology Online at http://www.treephys.oxfordjournals.org 1116 Barros et al. and Arabidopsis RD29A (−943 to −23 bp, At5g52310), through analysed using databases of cis-regulatory elements available restriction-based cloning into pLUCm35GUS (Figueiredo et al. online (PLACE (www.dna.affrc.go.jp/PLACE/), Higo et al. 1999; 2012). This vector contained the β-glucuronidase (GUS) PlantPAN (plantpan.mbc.nctu.edu.tw), Chang et al. 2008). reporter gene driven by the minimal 35S (m35S), in addition to Southern blotting the luciferase (LUC) gene under the control of CaMV35S. promPdDHN1 and promAtRD29A fragments were PCR ampli- Ten micrograms of almond genomic DNA were digested with fiedfromalmondandArabidopsis genomic DNA, respectively. EcoRI (Fermentas Inc., Ontario, Canada), EcoRV and SstI (Jena PstI and SalI recognition sites, present right upstream of the Bioscience, Jena, Germany) and separated on a 0.8% agarose m35S::GUS cassette in pLUCm35GUS, were used to clone the gel at 1.4 V/cm during 12–16 h. Digested DNA was blotted onto corresponding promoter fragments. Hybond N+ membrane (GE Healthcare, Uppsala, Sweden) by Polyethylene glycol (PEG)-mediated transfection was per- alkaline transfer (Sambrook et al. 1989). Probes (Supplementary D formed according to Anthony et al. (2004) using 3 : 1 (6 µg : Figure S1 available as Supplementary Data at Tree Physiology ow n 2 µg) equimolar ratios of effector : reporter vectors. Equal pro- Online) were labelled with [α-32P] dATP by random priming lo a toplast volumes (50 µl) were incubated with the corresponding using Klenow Fragment (New England Biolabs, Ipswich, MA, de d vector combinations, prior to the addition of 150 µl PEG solu- USA) according to the manufacturer’s instructions. Membranes fro m tion (25% [w/v] PEG 6000; 0.45 M mannitol; 0.1 M Ca[NO ] ). were pre-hybridized for 2 h in Church buffer (Church and 32 h Reactions were incubated for 10 min in the dark and PEG was Gilbert 1984) at 62 or 66 °C (for medium or high stringency) ttp s removed by centrifugation after serial dilutions with 0.275 M and hybridized overnight with Church buffer supplemented with ://a c Ca(NO3)2 solutions. Protoplasts were resuspended in the cul- ~2 × 107 cpm ml−1 of denatured probe. After hybridization, the ad e ture medium and incubated in the dark for 12 h. Three replicate membranes were washed twice in 2× SSC for 5 min, and once m ic transfections were prepared for each vector combination and in 2× SSC +0.1%SDSfor10min,at62/66°C.Afinalwash .o u p controls. Protoplasts were lysed using CCLR buffer (100 mM was performed in 1× SSC + 0.1% SDS for 10 min, at room tem- .c o K PO , 1 mM EDTA, 7 mM 2-mercaptoethanol, 1% [v/v] Triton, perature.MembraneexposuretoAmershamHyperfilm™MP m 102% 4glycerol) and cell debris was removed by centrifugation. (GE Healthcare) was carried out at −80 °C for 12 and 48 h. /tre e p β-Glucuronidase activity was assessed by monitoring cleavage Membranes were used for multiple hybridizations by stripping h y s of the GUS substrate 4-methylu mbelliferyl β-d-glucuronide the probe in 0.1% SDS at 68 °C for 2 h. /a (MUG) (Jefferson et al. 1987). Protein extracts (20 µl) were rtic le mixed with 0.5 µl MUG 50 mM and incubated for 1 h at 37 °C. Mapping analysis -a b s The reaction was stopped with 180 µl Na2CO3 200 mM. Mapping analysis was performed using the bin mapping strat- tra c Fluorescence was measured on a FLUO-star plate reader (BMG egy described by Howad et al. (2005) on the Prunus reference t/3 2 Labtechnologies Inc., Durham, NC, USA) at 455 nm when map, constructed from an almond (cv. Texas) × peach (cv. /9 /1 excited at 365 nm. As internal transfection control, the same Earlygold) F2 population. Cleaved amplified polymorphic 1 1 3 volume of each protein extract was assayed for LUC activity sequence markers were generated for each gene in order to /1 6 using a Modulus™ Microplate Luminometer (Turner Biosystems) discriminate parental cultivars and determine the genotype of 5 2 1 with dual injectors. Extracts were mixed with 150 µl LUC assay six F2 individuals previously selected for bin mapping. Genomic 6 3 reagent (10 mM Tricine pH7.8, 5 mM MgCl2, 0.1 mM EDTA, DNA from the parental cultivars, the F1 hybrid and the F2 indi- by g 3.3 mM DTT and 2 mM ATP) and 0.5 µM Luciferin (Biotium vidualswasusedasatemplateforPCRusinggene-specific u e s Inc., Hayward, CA, USA) and luminescence was measured with primers (CBF-3F and CBF1-5R3 for PdCBF1; CBF2-5F and t o n 10 s integration time. β-Glucuronidase and LUC values were CBF2-5R2 for PdCBF2; PromDHN1-F and SalDHN1-R for 1 0 calculated as the average of three independent readings. PdDHN1; see Table 1) and the corresponding amplicons were A p β-Glucuronidase/LUC ratios were calculated and results are further digested with NotI (PdCBF1), TaqI (PdCBF2) or XbaI ril 2 0 shown as the average for each transfection replicate. (PdDHN1). The genotypic pattern obtained for the F2 plants 1 9 was then compared with those previously established by In silico sequence analysis Howad et al. (2005),whichdefineaspecificregion(bin) All sequences were edited using EditSeq and in silico restric- among the eight linkage groups (LGs) from TxE reference map. tion analysis was performed using MapDraw, both from Stress treatments Lasergene software package (DNASTAR Inc., Madison, WI, USA). Pairwise and multiple sequence alignments were made Expression analysis of both PdCBFs during cold exposure was using ClustalW (www.ebi.ac.uk/Tools/msa/clustalw2) (Larkin conducted using 2-week-old in vitro almond shoots, previously et al. 2007) and phylogenetic analysis was performed by the adapted to a 12/12 h photoperiod (100 µmol m−2 s−1) at neighbour-joining method using MEGA4 (www.megasoftware. 22/20 °C for 5 days. In two independent experiments, cold was net) software (Tamura et al. 2007). PdCBF promoters were applied after dawn or after dusk by decreasing the temperature Tree Physiology Volume 32, 2012 CBF transcription factors in almond cold response 1117 to 5 °C and light intensity to 60–70 µmol m−2 s−1. After 0, 1, 2, 8, during 3′-RACE, an alternative 3′ extension was obtained and, 12, 16 and 24 h of treatment, four plantlets were pooled and after 5′-RACE,theidentificationofanadditionalmemberofthe frozen in liquid nitrogen. Cold acclimation experiments were con- almond CBF gene family (PdCBF2)wasconfirmed.PdCBF1 ducted with plants adapted to 12/12 h (SD) at 20 °C for 5 days contained an open reading frame (ORF) of 729 nucleotides, (D1 − 5). Plants were then transferred to 12 °C and after 5 more encoding a 242 amino acid protein of 27.45 kDa with a pI 7.69. days (D6–10) temperature was again reduced to 4 °C (D11). At The deduced PdCBF2 ORF was slightly smaller than PdCBF1, D5, D6, D10 and D11, samples were collected and frozen in liq- with 717 nucleotides, encoding a 238 amino acid protein with uid nitrogen at 2, 8, 16 and 24 h (after dawn). Abscisic acid and 26.59 kDa, pI 5.29. PdCBF1 and PdCBF2 shared 82% amino dehydration treatments were performed by transferring plants to acid identity. the culture medium containing 100 µM ABA and medium-free Multiple sequence alignment of PdCBF1 and PdCBF2 pre- cultureflasks,respectively.Mockcontrolswereobtainedby dicted proteins with other CBFs from different species con- D transferringtheplantstonewflaskscontainingfreshmicropropa- firmedtheconservationoftheAP2DNA-bindingdomainaswell ow n gation medium. After 10 min, 30 min, 1 h, 2 h, 8 h and 12 h four asconservedmotifsCMIII-1to4aspreviouslydefinedby lo a plantlets were pooled and frozen in liquid nitrogen. Nakano et al. (2006) (Figure 1a). These conserved motifs de d included the putative nuclear localization signal and the DSAWR fro Semi-quantitative reverse transcriptase-PCR m signature(CMIII-3)flankingtheAP2/ERFregion.Inphyloge- h Two micrograms of total RNA were used for cDNA synthesis netic analysis, the sequences belonging to the Prunoideae sub- ttp s with SuperScript® First-Strand Synthesis System (Invitrogen) family clustered together, demonstrating their high similarity ://a c following the manufacturer’s instructions. cDNA samples were (Figure 1b). Within this cluster peach CBFs grouped separately a d e diluted 1 : 3 in Milli-Q H O (Merck Millipore, Billerica, MA, from PdCBF1, PdCBF2 and PaCBF1. The CBF/DREB1 peptide m 2 ic USA)andusedastemplateforPCRwithgene-specificprimers sequences from dicot species clustered in three major groups .o u p (Table 2), in 20 µl total volume. PdCBFswereamplifiedusing (I, II and III; see Figure 1b). Interestingly, CBF-like proteins from .c o 2 µl cDNA, while 1 µl was used for the remaining genes. the same species (B. pendula, P. trichocarpa) or genus (M. x m PdActin (AM491134) was used as internal control. Total reac- domestica and M. baccata) showed up in two different groups /tre e p tion volumes were analysed by electrophoresis on 1.2% aga- (I and II) suggesting a different evolutionary origin. The Prunus h y s rose gel stained with ethidium bromide. Images were captured CBFs isolated so far clustered in group II. /a using the ChemiDoc XRS System (Bio-Rad, Hercules, CA, The almond Dehydrin1 (PdDHN1) promoter and partial cod- rtic le USA). At least two technical replicates were performed for ingsequencewasalsoisolatedbyPCRusingspecificprimers -a b s each gene. targeting the reported sequence from peach, PpDHN1 (Artlip tra c et al. 1997) (data not shown). The PdDHN1 sequenced region t/3 2 was highly similar to the PpDHN1, sharing 97% nucleotide /9 Results /1 identity up to the initiation codon. The presence of two CRT/ 1 1 3 DREelementswasconfirmedat−238 and −191 bp from the /1 Isolation of two almond CBF-like genes deduced TATA-box, similar to the PpDHN1 promoter 65 2 1 An almond CBF-like sequence (PdCBF1) was isolated from (Wisniewski et al. 2006). 6 3 cDNA obtained from cold-treated in vitro almond plantlets, by using a PCR-based approach with primers designed for the PdCBF1 and PdCBF2 promoter analysis gu e s sweet cherry PaCBF (Kitashiba et al. 2004). Homology to To gain an insight into the signalling mechanisms involved in t o n CBF/DREB1familyofTFswasconfirmedusingBLASTXagainst PdCBF1 and PdCBF2 transcription regulation we isolated their 1 0 the GenBank database. Among the multiple clones sequenced corresponding promoter regions by iPCR. This approach A p ril 2 0 Table2.Sequencesandcorrespondingamplificationconditionsandampliconsizeoftheprimercombinationsusedforsemi-quantitativeRT-PCR. 19 Gene Primer sequence Anneal. (°C) No. of PCR cycles Ampl. (bp) PdCBF1 F 5′-CGCTAATGAACAGGTTCTTCTCTCA-3′ 58 32 550 R 5′-TTCACACTATCCTTCTTCTTCTTCTTC-3′ PdCBF2 F 5′-CTCTAATGGACTTGTCTCAACTTTC-3′ 58 32 540 R 5′-CCAAGTTCACACTACCCTTCTTG-3′ PdDHN1 F 5′-CTTATGGTGGGGCTGGGTA-3′ 62 23 453 R 5′-TCGGTGGTCACAGACCTACA-3′ PdActin F 5′-AGCAAGGTCCAGACGAAGAA-3′ 58 24 385 R 5′-TGTAGGTGATGAAGCCCAATC-3′ F/R, forward/reverse primers; Anneal., annealing temperatures; Ampl., amplicon size. Tree Physiology Online at http://www.treephys.oxfordjournals.org 1118 Barros et al. D o w n lo a d e d fro m h ttp s ://a c a d e m ic Figure 1. (a) Multiple alignment of CBF amino acid sequences from Prunus spp. with other representative CBF sequences from Arabidopsis .o u (AtCBF1–4; AtDDF1), poplar (PtCBF3 − 4), dwarf apple (MbDREB1) and citrus (PtrCBF), performed with ClustalW; (b) Phylogenetic relationships p between CBF/DREB1 and DREB2 amino acid sequences from herbaceous and woody plant species. A consensus neighbour-joining tree was gen- .co m e(AraYt6e6d7, 2b4a7s)e, dA otCn BmF2u lt(ipAlYe6 a6l7ig2n4m7e),n tA, twCBithF 3b o(oAtYs6tr6a7p2 4a7n)a,l yAsitsC BwFit4h (1N0M0_01 2re4p5l7ic8a)t,i oAnstD. DThFe2 s(eNqMue_1n0ce1s1 3u1s)e, dA tfDorR EthBe2sAe (aNnMaly_s0e0s1 0w3e6re7:6 A0t)C BaFn1d /tre e AtDREB2B (NM_111939) from A. thaliana; BpCBF1 (EF530204), BpCBF2 (EF530205) and BpCBF3 (EF530206) from B. pendula; EguCBF1A p h (DQ241820), EguCBF1B (DQ241821), EguCBF1C (EU794855) and EguCBF1D (EU794856) from E. gunnii; GhDREB1A (AY321150) from ys Gfroomss yPp. iturimch hoicrasurptuam; P; tLrCeCBBF F(1D Q(A7Y9003848497)3 f)ro fmro mP. tLryifcoolipaetar;s iVcvoCn BeFs1c u(lAeYn3tu9m0 3M7i2ll.);, PVtvCCBBFF12 ((EAFY135910435766),) PatnCdB VFv2C (BEFF31 5(A1Y435970) 3a7n5d) PfrtoCmB FV3it i(sE vFi1ni5fe1r4a5 8L.); /article OsCBF3 (AF300970) from Oryza sativa L.; MdCBF1 (HM992942) from M. × domestica; MbDREB1 (EF582842) from M. baccata; PaDREB1 -a b (AB121674) from P. avium; PpCBF1 (HM992943) and PpDREB1 (EF635424) from Prunus persica (L.) Batsch; and PdCBF1 (JQ317157) and s PdCBF2 (JQ317158) from Prunus dulcis. CBF/DREB1 sequences from dicotyledon species clustered in three groups, I, II and III. Bootstrap values tra c are indicated on the branches. The bottom scale represents 0.1 amino acid substitutions per site. t/3 2 /9 /1 allowedtheidentificationofsequenceswith1401and1417bp, Genomic analysis of PdCBFs and PdDHN1 1 1 3 upstream of the deduced transcription start sites (TSSs) of /1 6 PdCBF1 and PdCBF2, respectively. The putative TATA-box To investigate if additional genes similar to PdCBF1 and 5 2 1 motifs were located ~30 bp upstream of the TSS (Figure 2a PdCBF2 could be found in the almond genome, we performed 6 3 and b), close to two putative CAAT-box regions (not shown). In a Southern blotting hybridization using the complete PdCBF1 by g silico analyses of the −1000 bp upstream TATA-box revealed coding sequence as probe. Genomic DNA was digested with u e s the presence of several core sequence motifs related to envi- EcoRI, EcoRV and SstI, which did not cut within the probe, and t o n ronmental and endogenous signalling (Figure 2a and b). In par- hybridization was performed under medium stringency condi- 1 0 ticular, PdCBF1 and PdCBF2 promoters contained 6–7 MYC tions(62°C).Thisapproachallowedtheidentificationofat A p motifs, some of which also contained G-box core motifs related least eight to nine bands in the EcoRI lane, seven in EcoRV ril 2 0 to light/circadian clock-regulated transcription (Martínez- and four in SstI (Figure 3). In the last, the occurrence of four 1 9 García et al. 2000). The PdCBF1 promoter contained a con- bandswasconfirmedinareplicateanalysis(datanotshown). served motif of ICEr2 close to the TATA-box, which in To determine PdCBF1 and PdCBF2 copy number in the almond Arabidopsis was found to be a determinant for cold induction of genome,DNAblotswerehybridizedtotwoprobesspecificfor AtCBF2 (Zarka et al. 2003). In the PdCBF2 promoter, an ICEr2- the promoter and 5′-UTR regions (Supplementary Figure S1a like motif was also found in a similar position. The PdCBF1 available as Supplementary Data at Tree Physiology Online) of promoter contained at least six ABA-responsive elements PdCBF1 and PdCBF2, respectively, also with no recognition (ABREs), while in PdCBF2 only three were found. Several MYB- sites for the selected enzymes. In both cases, and under high- related consensus motifs were also found in both promoters, stringency conditions (66 °C), the hybridization pattern of each as well as light-responsive elements such as GT1 and SORLIP, PdCBF revealed the presence of at least two bands which were more abundant in the PdCBF2 promoter. (Supplementary Figure S2 available as Supplementary Data at Tree Physiology Volume 32, 2012 CBF transcription factors in almond cold response 1119 Arabidopsis protoplasts (Figure 5). Reporter vectors containing the AtRD29A or PdDHN1 promoter regions (promAtRD29A or promPdDHN1) cloned upstream of the m35S::GUS cassette were used as baits (Figure 5a). As a control to normalize trans- fectionefficiency,reportervectorsalsocontainedthe35S::LUC cassette. Basal GUS expression determined in protoplasts transfected with promAtRD29A-m35S::GUS or promPdDHN1- m35S::GUS, without effectors, was slightly higher than that of the empty reporter vector (m35S::GUS, Figure 5b). However, in the presence of PdCBF1 or PdCBF2, GUS expression con- trolled by either promAtRD29A or promPdDHN1 was upregu- D lated >2-fold as compared with no effector experiments ow n (Figure 5b). In addition, GUS/LUC ratios also increased in the lo a protoplasts cotransfected with m35S::GUS and either one of de d the effectors. A detailed DNA sequence analysis of the original fro m vector backbone revealed the presence of a CRT/DRE-like h motif, −180 bp upstream the m35S (Figure 5a, grey arrow- ttp s head), which may act as a binding site for CBF TFs. Overall, ://a c these results suggest that PdCBF1 and PdCBF2 are functional a d e transactivator proteins. m ic .o u Expression of PdCBF1 and PdCBF2 is induced by low p .c temperature om To determine the expression pattern of PdCBF1, PdCBF2 and /tre e p their predicted target, PdDHN1, under low-temperature treat- h y s ment, we performed a gene expression analysis on in vitro /a almond shoots submitted to cold. Plants were cultured for 5 rtic le Figure 1. (Cont.). days in a 12/12 h photoperiod and temperature decrease -a b s was imposed immediately after dawn. By semi-quantitative tra c Tree Physiology Online). Considering that almond is highly het- RT-PCR, we observed that PdCBF1 and PdCBF2 expression t/3 2 erozygous (Arús et al. 2009), this pattern may correspond to was induced at 2 and 1 h, respectively, after temperature /9 /1 two putative allelic forms. Thus, the occurrence of additional reduction (Figure 6a). PdCBF1 transcript accumulation 1 1 3 bands in the medium stringency blot (Figure 3) thus suggests showed an increase with time, reaching the highest level after /1 6 that the almond CBF family may include two to three additional 16 h treatment (4 h after dusk). For PdCBF2, the results 5 2 1 CBF-like members. suggest that transcription induction may have started earlier 6 3 The genetic mapping of PdCBF1, PdCBF2 and PdDHN1 was than PdCBF1, but showed a similar expression pattern during by g performed by selective (bin) mapping analysis on the the 24 h. u e s almond × peach Prunus reference map. Using this approach, we Since the increase in transcript accumulation of both PdCBF t o n observed that both PdCBFs shared the same genotypic pattern genes was observed towards the night period, we investigated 1 0 (Supplementary Figure S3 available as Supplementary Data at if transcript induction could be positively regulated during the A p Tree Physiology Online), which matched to that of bin 5 : 21, night period, by imposing cold stress immediately after the light ril 2 0 located within LG5. In turn, PdDHN1 associated with bin 7 : 41 period. In these conditions we observed that both PdCBF1 and 1 9 in LG7. Figure 4 shows the schematic representation of LGs 5 PdCBF2 expression reached a peak at 8 h after stress, declin- and 7 with the location of bins 5 : 21 and 7 : 41. Although with ing to a steady state thereafter (Figure 6b). During light-to-dark low precision, both bins overlapped with the relative positions cold treatment, PdDHN1 transcripts exhibited an accumulation of two quantitative trait loci (QTLs) related to blooming time and after 8 h and increased during the night period (Figure 6a). chilling requirements. However, in dark-to-light cold treatment, PdDHN1 transcript lev- els were already increased at 0 h (Figure 6b), faintly decreasing Transient expression of PdCBF1 and PdCBF2 after 2 h of dark and increasing again at 8 h and until the end of in Arabidopsis protoplasts the treatment. Analysis of PdDHN1 expression at room temper- To test whether PdCBF1 and PdCBF2 are functional TFs, their ature (RT) during the same period showed that this gene is activity was evaluated through transient expression in actively transcribed at the end of the day period, but decreasing Tree Physiology Online at http://www.treephys.oxfordjournals.org 1120 Barros et al. D o w n lo a d e d fro m h ttp s ://a c a d e m ic .o u p .c o m /tre e p h y s /a rtic le -a b s tra c t/3 2 /9 /1 1 1 3 Figure 2. Analyses of the nucleotide sequence of the promoter regions of PdCBF1 (a) and PdCBF2 (b) including the position of some predicted /1 cis-elements. The sequences homologous to MYC and GT1 recognition sites are highlighted in grey and open boxes, respectively. The putative 65 2 TATA-boxes are indicated within circles. The arrows show the putative transcription initiation point predicted by 5′−RACE. 1 6 3 b y g during the night (Figure 6c). To clarify the role of light/circadian When temperature decreased to 12 °C, expression of PdCBF1 u e s clock in the induction of PdDHN1 under RT conditions, further and PdCBF2 was upregulated (D6, Figure 7b), with transcripts t o n expression studies were conducted in plants grown under con- being detected at least 2 h after temperature decline. In con- 1 0 trol light/dark conditions followed by 24 h under continuous trast to RT, PdDHN1 expression did not decrease 24 h after A p dark (Supplementary Figure S4 available as Supplementary exposure to 12 °C (D6, Figure 7b), suggesting that the repres- ril 2 0 Data at Tree Physiology Online). Consistently, PdDHN1 tran- sion of transcription occurring at RT before dawn is prevented 1 9 scripts accumulated towards the end of the day and decreased by cold. After 4 days at 12 °C, both PdCBFs were still expressed during the night. However, during the subjective day under con- but at a lower level as compared with D6 while PdDHN1 tinuous dark, PdDHN1 transcript accumulation remained at low showed a constitutive expression pattern (D10, Figure 7b). basal levels. These results suggest that under RT conditions When temperature decreased to 4 °C (D11, Figure 7b) PdCBF1 PdDHN1 transcription is also controlled by light. and PdCBF2 transcript level was again induced while no differ- To investigate whether both PdCBFs could be differentially ences were observed in PdDHN1 transcription. Collectively, regulated during a two-stage cold treatment, we acclimated these results highlight the involvement of PdCBF1 and PdCBF2 almond in-vitro-propagated shoots for 5 days at 12 °C, fol- in temperature perception, being induced when environmental lowed by 24 h at 4 °C (Figure 7a). PdCBF1 and - 2 exhibited temperatures decline to 12 °C, but keeping the responsiveness low levels of transcript accumulation at RT (D5, Figure 7b). to lower temperatures. Tree Physiology Volume 32, 2012 CBF transcription factors in almond cold response 1121 D o w n lo a d e d fro m h ttp s ://a c a d e m ic .o u p .c o m /tre e p h y s /a rtic le -a b s tra c t/3 2 /9 Figure 2. (Cont.) /1 1 1 3 /1 6 5 2 1 PdCBF1 and PdCBF2 are transiently induced by Discussion 63 b dehydration and ABA y g In perennial fruit trees, temperature signals can modulate u e s The effect of drought and ABA on PdCBF1 and PdCBF2 dormancy state and cold acclimation, which are determinant t o n expression was further investigated by transferring in vitro factors for winter survival. In Prunoideae, two CBF genes 1 0 plantlets to dehydrating conditions (medium-free culture involved in cold acclimation were previously reported in sweet A p flasks)ortoculturemediumsupplementedwith100µM cherry (Kitashiba et al. 2004) and peach (Wisniewski et al. ril 2 0 ABA. In both conditions, PdCBF1 and PdCBF2 were slightly 2011), and were mainly studied through heterologous expres- 1 9 andtransientlyinducedduringthefirsthourafterexposure, sion. In the present work, we focused on almond, which is not but transcript levels declined thereafter (Figure 8). The mock only the earliest fruit tree to break dormancy but also shows control showed a faint increase in the transcript level, indicat- the widest range of blooming dates among all the fruit and ing that these genes could also be responding to mechanical nut species (Socias i Company and Felipe 1992). PdCBF1 stress, as also observed in other species (Gilmour et al. and PdCBF2geneswereidentified,sharingahighhomology 1998). PdDHN1 transcript levels showed an increase through- to PaCBF1. Phylogenetic analysis (Figure 1b) showed that out the 12 h treatment in the three conditions (Figure 8). Prunus CBF sequences grouped within the same cluster, yet Nonetheless, during dehydration and ABA, transcript levels the two peach CBF sequences (PpCBF1 and PpDREB1) were reached much higher levels, suggesting the involvement of the most dissimilar. The phylogenetic tree obtained sug- this gene in the plant response to osmotic stress. gested that CBF/DREB1 proteins could segregate due to Tree Physiology Online at http://www.treephys.oxfordjournals.org 1122 Barros et al. additional identity factors other than the speciation effect alone. In fact, Prunus CBFs grouped with poplar (PtCBF3 and PtCBF4), dwarf apple (MbDREB1), birch (BpCBF3) and Arabidopsis (AtDDF1) CBF members in cluster II, but other members from the same genus/species were also found in cluster I. Although further studies need to be performed in this regard, the phylogenetic distribution obtained may relate the genes within each cluster to a common ancestor and/or functional aspect. Mapping analysis in the almond × peach Prunus reference map revealed that PdCBF1 and PdCBF2 were co-localizing D within a 5.8 cM region in LG 5 (Figure 4). By Southern blot- ow n ting analysis, we demonstrated that the almond CBF family lo a mightbecomposedofatleastfourtofivemembers(Figure de d 3). Considering that the three enzymes used did not generate fro m the same number of bands, it is possible that the putative h genes could be located within the same region, not being fully ttp s resolved by restriction. In fact, a tandem organization of CBF ://a c genes has been previously observed in other species, such as a d e Figure 3. Southern blot analysis of almond genomic DNA. The PdCBF1 Arabidopsis (Shinwari et al. 1998) and tomato (Zhang et al. m coding sequence was used as probe, in medium stringency condi- ic 2004). With the advent of peach genome sequencing, avail- .o tions. The consensus pattern obtained from replicate analyses is rep- u p resented on the right. The numbers on the right indicate the size of the able since April 2010 (International Peach Genome Initiative, .c o fragments from ladder DNA (Fermentas). http://www.rosaceae.org),geneidentificationinPrunus spe- m cies has been facilitated. As determined for the almond /tre e p genome by Southern blotting, the peach genome revealed the h y s presenceoffiverelatedsequences,locatedwithinthesame /a genomic region (Supplementary Table S1 and Supplementary rtic le Figure S5 available as Supplementary Data at Tree Physiology -a b s Online). tra c CBF proteins from several plant species bind to CRT/DRE t/3 2 cis-elements to regulate the expression of their target genes /9 /1 (Jaglo-Ottosen et al. 1998, Dubouzet et al. 2003, Qin et al. 1 1 3 2004, Xiao et al. 2006, Champ et al. 2007). The functional /1 6 activity of PdCBF1 and PdCBF2 was tested through transient 5 2 1 expression in Arabidopsis protoplasts, together with reporter 6 3 vectors carrying the m35S::GUS cassette under the control by g of the CRT/DRE-containing sequences from promPdDHN1 u e s and promAtRD29A. AtRD29A is a widely known stress t o n responsive gene in Arabidopsis (Yamaguchi-Shinozaki and 1 0 Shinozaki 2006), while PdDHN1 is homologous to PpDHN1, A p which was shown to be cold responsive (Wisniewski et al. ril 2 0 2006). When protoplasts were transfected with reporter 1 9 vectors carrying promAtRD29A or promPdDHN1, an increase in GUS expression was observed, as compared with the basal expression determined for the m35S::GUS cassette (Figure 5). These results suggested that the stress condi- tions (e.g. osmostic stress) induced in protoplasts during Figure 4. Predicted position of bins 5 : 21 and 7 : 41 (dark grey rect- PEG-mediated transformation might themselves promote angles)andflankingmolecularmarkerswithinlinkagegroups(LG)5 and 7, as described by Howad et al. 2005. The approximate locations GUS expression in the presence of promAtRD29A or promP- of QTLs related to blooming time are represented on the left side of dDHN1. DREB2 proteins, which also recognize the CRT/DRE LG5 and LG7 as open vertical rectangles (Fan et al. 2010) and a thin present in both of the promoters, are among the endogenous vertical line (the thick horizontal line represents the site of the highest LOD) (Dirlewanger et al. 1999). cM, Centimorgan. regulators that may be induced under those conditions Tree Physiology Volume 32, 2012
Description:The list of books you might like
