Regulation of miR319 during cold stress in sugarcane

Authors

  • FLÁVIA THIEBAUT,

    1. Laboratorio de Biologia Molecular de Plantas, Instituto de Bioquímica Médica, Bl.B-33A
    2. Pós-Graduação em Biotecnologia Vegetal, Universidade Federal do Rio de Janeiro, Rua Rodolpho Paulo Rocco s/n, CCS, Bl.K-32, Cidade Universitária, Rio de Janeiro 21941-590, Brazil
    Search for more papers by this author
  • CRISTIAN A. ROJAS,

    1. Laboratorio de Biologia Molecular de Plantas, Instituto de Bioquímica Médica, Bl.B-33A
    2. Universidade Federal da Integração Latino-Americana, Av. Tancredo Neves 6731, Bl.4, Foz do Iguaçu, Paraná 85867-970, Brazil
    Search for more papers by this author
  • KARLA L. ALMEIDA,

    1. Laboratorio de Biologia Molecular de Plantas, Instituto de Bioquímica Médica, Bl.B-33A
    Search for more papers by this author
  • CLÍCIA GRATIVOL,

    1. Laboratorio de Biologia Molecular de Plantas, Instituto de Bioquímica Médica, Bl.B-33A
    Search for more papers by this author
  • GISELLI C. DOMICIANO,

    1. Laboratorio de Biologia Molecular de Plantas, Instituto de Bioquímica Médica, Bl.B-33A
    Search for more papers by this author
  • CAREN REGINA C. LAMB,

    1. Fundação Estadual de Pesquisa Agropecuária – Centro de Pesquisa em Agroindústria, Fazenda Souza, Caxias do Sul, Rio Grande do Sul 95125-000, Brazil
    Search for more papers by this author
  • JANICE DE ALMEIDA ENGLER,

    1. INRA, UMR 1301 and CNRS, UMR 6243, Interactions Biotiques et Santé Végétale, Université de Nice Sophia-Antipolis, Sophia Antipolis 06903, France
    Search for more papers by this author
  • ADRIANA S. HEMERLY,

    1. Laboratorio de Biologia Molecular de Plantas, Instituto de Bioquímica Médica, Bl.B-33A
    2. Pós-Graduação em Biotecnologia Vegetal, Universidade Federal do Rio de Janeiro, Rua Rodolpho Paulo Rocco s/n, CCS, Bl.K-32, Cidade Universitária, Rio de Janeiro 21941-590, Brazil
    Search for more papers by this author
  • PAULO C. G. FERREIRA

    Corresponding author
    1. Laboratorio de Biologia Molecular de Plantas, Instituto de Bioquímica Médica, Bl.B-33A
    2. Pós-Graduação em Biotecnologia Vegetal, Universidade Federal do Rio de Janeiro, Rua Rodolpho Paulo Rocco s/n, CCS, Bl.K-32, Cidade Universitária, Rio de Janeiro 21941-590, Brazil
    Search for more papers by this author

P. Ferreira. Fax: +55 21 3204 2085; e-mail: paulof@bioqmed.ufrj.br

ABSTRACT

MicroRNAs (miRNAs) are part of a novel mechanism of gene regulation that is active in plants under abiotic stress conditions. In the present study, 12 miRNAs were analysed to identify miRNAs differentially expressed in sugarcane subjected to cold stress (4 °C). The expression of miRNAs assayed by stem–loop RT-PCR showed that miR319 is up-regulated in sugarcane plantlets exposed to 4 °C for 24 h. The induction of miR319 expression during cold stress was observed in both roots and shoots. Sugarcane miR319 was also regulated by treatment with abscisic acid. Putative targets of this miRNA were identified and their expression levels were decreased in sugarcane plantlets exposed to cold. The cleavage sites of two targets were mapped using a 5′ RACE PCR assay confirming the regulation of these genes by miR319. When sugarcane cultivars contrasting in cold tolerance were subjected to 4 °C, we observed up-regulation of miR319 and down-regulation of the targets in both varieties; however, the changes in expression were delayed in the cold-tolerant cultivar. These results suggest that differences in timing and levels of the expression of miR319 and its targets could be tested as markers for selection of cold-tolerant sugarcane cultivars.

INTRODUCTION

A novel mechanism of gene regulation has emerged recently, the mechanism of RNA interference (Fire et al. 1998; Tijsterman, Ketting & Plasterk 2002). A central component of this regulatory pathway is a molecule of single-stranded RNA, namely microRNA (miRNA). MiRNAs are small endogenous RNAs 20–25 nucleotides in length (Zhang et al. 2006) and they play a role in regulating the expression of messenger RNA (mRNA) targets by cleavage or repression of translation. MiRNA-guided gene regulation is essential for normal growth and development of the plant (Jover-Gil, Candela & Ponce 2005; Garcia 2008; Rubio-Somoza et al. 2009), as well as for adaptation to stress conditions (Lu, Sun & Chiang 2008; Zhou et al. 2008; Ding et al. 2009a). In plants, the main mechanism of action of miRNAs is the regulation by direct cleavage of target mRNA (Kidner & Martienssen 2005). In this process, miRNA coupled to the ribonucleoprotein complex called RNA induced silencing complex (RISC) recognizes the target mRNA by perfect or near-perfect Watson & Crick pairing (Lu & Huang 2008). The RISC complex contains many proteins, but the best characterized is AGO1, which is required for mRNA target cleavage (Baumberger & Baulcombe 2005).

Several studies have reported the involvement of miRNAs in plants in response to stress conditions, and numerous miRNA targets are genes related to stress response (Jones-Rhoades & Bartel 2004; Phillips, Dalmay & Bartel 2007). Under stress conditions, the modulation (up- or down-regulation) of miRNAs modifies the abundance of their target genes (Jagadeeswaran, Saini & Sunkar 2009; Lv et al. 2010). Moreover, the repression of miRNA can lead to accumulation of its mRNA target, which could be a positive regulator of stress tolerance (Sunkar & Zhu 2004).

Under natural conditions, plants are often exposed to stresses caused by the environment. Low temperature is one of the most common factors limiting the geographical distribution and growing season of many plant species (Guy 1990). Recent studies have shown that cold acclimation can be regulated at post-transcriptional level (Chinnusamy, Zhu & Zhu 2007). Important crops such as rice, corn, cotton and tomato are sensitive to low temperatures and unable to tolerate ice formation in their tissues (Thomashow 1999). Cold-induced genes can also be induced by drought stress and, in many cases, these genes are also regulated by the phytohormone abscisic acid (ABA) (Nogueira et al. 2003). The application of exogenous ABA in plants mimics a stress condition, suggesting that the signalling pathways triggered by these stresses and ABA share common elements (Leung & Giraudat 1998). However, there are also reports that describe genes that are induced by drought and cold, but which are not sensitive to ABA treatment. This indicates the existence of both dependent and independent ABA regulatory pathways with respect to the transcriptional control of the response to drought and cold (Shinozaki & Yamaguchi-Shinozaki 1996).

Sugarcane is generally considered a plant sensitive to low temperatures, but field observations have shown that this sensitivity is variable among the different varieties of sugarcane (Nogueira et al. 2003). Sugarcane has one of the most complex plant genomes and it has a variable ploidy number (Ingelbrecht, Irvine & Mirkov 1999). The molecular analysis of sugarcane subjected to cold stress may help to better understand the molecular mechanisms of response of this crop and related species, and contribute to breeding programs directed towards cold and drought tolerance. Our goal was to study the involvement of miRNAs and cognate targets in the tolerance to low temperatures in sugarcane. The results reported here indicated the regulation of miR319 and its targets in sugarcane subjected to cold stress. Furthermore, these results suggest differences in the timing and intensity of this regulation between sugarcane cultivars that are tolerant and sensitive for cold stress.

MATERIALS AND METHODS

Plant material and RNA isolation

Sugarcane plantlets (Saccharum sp. Cv SP70-1143) free of micro-organisms were obtained by sterile meristem culture and micropropagated according to the method of Hendre et al. (1983). In vitro-grown SP70-1143 rooted sugarcane plantlets, free of micro-organisms, were subjected to the following treatments: (1) Cold stress: plants were exposed to 4 °C and plantlets were harvested at 0, 3, 6, 12, 24 and 48 h after treatment. A set of five plants was collected for each point of the experiment and the pooled material was analysed. (2) In a second experiment of cold stress, a set of sugarcane plants was exposed to 4 °C for 0, 12, 24 and 48 h, and roots and shoots were collected separately. In both experiments, pools of five plants per time point were grown in vitro at 28 °C as controls. (3) In the experiment with the hormone ABA, plants were treated 100 µm for a period of 0, 3, 6, 12, 24 and 48 h. A set of three plants was collected for each point of the experiment and it has been done in duplicate. Plants grown in medium, without the addition of ABA, were used as controls for each time point.

To compare the expression of miRNAs and their targets in sugarcane cultivars with different cold sensitivities, we used two different varieties: TAMBO FEPAGRO and RB931011. TAMBO FEPAGRO is an indigenous cultivar from the far south of Brazil (29°04′34″S; 56°18′38″W and 123 m), grown locally by small farmers. This cultivar was identified by breeders as the only one, among several others grown in the same region, capable of resisting temperatures as low as 3 °C, or even frost. RB931011 is a recently released cultivar, selected for its high productivity in the northeast of Brazil and grown in regions with an average temperature of 25 °C. Stalks were germinated and grown in a greenhouse in a pot containing sand and vermiculite (2:1). After 6 weeks, the plants were exposed to cold stress (4 °C). Treated and control shoots were collected after 0, 24 and 48 h of treatment. Plants were immediately frozen in liquid nitrogen and ground into a fine powder. Total RNA from each sample was extracted with Trizol reagent (Invitrogen®, Carlsbad, CA, USA) as described by the manufacturer.

Validation of experimental stresses

The cold-responsive protein (CoReP) and SsNAC23 genes (Nogueira et al. 2003, 2005) have been reported to be induced by low temperature and were used here to verify the efficiency of cold stress treatment in all experiments. Treatment with ABA was also verified with gene SsNAC23. Specific primers for genes SsNAC23 and CoReP were (reverse: GCCCTCCCTTCTTGTTGTAG; forward: CGAGAAGACCAACTGGATCA) and (reverse: AGCAGCGAATCTGGTAGGAT; forward: CGGGAGATGCCATTACTACA), respectively. First strand cDNA was synthesized using the kit Taqman First Strand cDNA Synthesis according to the manufacturer (Applied Biosystems, Foster City, CA, USA) in reactions with final volume of 25 µL. For quantitative real-time PCR (qRT-PCR) reaction were used 5 µL of the first strand cDNA; 5 µL of the primer pair to be analysed, at a concentration of 3.6 mm each; and 10 µL SYBR Green PCR Kit Master Mix (Applied Biosystems) in a final volume of 20 µL in each well. The reactions were carried out in a 96-well plate (MicroAmp Optical 96-Well Reaction Plate, Applied Biosystems). The program used was 94 °C for 4 min, 40 cycles of 94 °C for 1 min, 55 °C for 1 min and 30 s and 72 °C for 1 min and 30 s. The result was analysed using the program Sequence Detection version 1.3 from Applied Biosystems, and the relative expression was calculated using the control for the same time point according to Livak & Schmittgen (2001). As housekeeping gene, glyceraldehyde-phosphate dehydrogenase (GAPDH) was used to normalize the concentration of first strand cDNA used in all qRT-PCR reactions.

Analysis of the expression of mature miRNAs by pulsed RT-PCR

The expression profiles of 12 sugarcane mature miRNAs (miR156, miR159, miR160, miR166, miR167, miR169, miR172, miR319, miR393, miR394, miR408, miR528) were assayed by stem–loop reverse transcription-PCR. The stem–loop reverse transcription primers were designed following the method described by Chen et al. (2005) and Varkonyi-Gasic et al. (2007). Synthesis of the first strand cDNA was done with 2 µg RNA, 1 µL of oligodT (500 ng µL−1), 1 µL of 10 mm dNTP and 1 µL each RT primer (1 µm). The RT primers were specific primers for each of the miRNAs to be analysed, and this specificity was provided by a sequence of six nucleotides found in the region of the 3′ end of the primer. In the same reaction, RT primers specific for all miRNAs analysed were used (Supporting Information Table S1). To analyse the expression profile of mature miRNAs, qRT-PCR was used with SYBR Green PCR Master Mix (Applied Biosystems). To each well were added 2 µL of first strand cDNA, 3 µL of water, 10 µL of SYBR Green solution, 1 µL of the forward primer (10 µm) – specific miRNA that was being analysed (Supporting Information Table S2) – and 1 µL of reverse primer (10 µm). The reverse primer (5′ GTGCAGGGTCCGAGGT3′) was the same for all the reactions.

Northern blot analysis

Total RNA of root (plantlets of second experiment) was fractionated on a denaturing 15% polyacrylamide gel, transferred to HybondTM-N+ membrane (Amersham Biosciences, Piscataway, NJ, USA) by capillary method and fixed by ultraviolet cross-linking. Blots were hybridized using the miRNA StarFire® Oligonucleotide Labeling kit (Integrated DNA Techonologies, Inc., San Diego, CA, USA). Radioactive (α-32P-dATP) probes were designed against miR319 and miR159.

Analysis of miR319 targets

The identification of putative miR319 target genes was carried out using the software Web MicroRNA Designer (http://wmd.weigelworld.org) and miru: Plant microRNA Potential Target Finder (http://bioinfo3.noble.org/miRNA/miRU.htm), using the sugarcane sequences dataset deposited in these banks. A similar search was also made for miR319 targets in the database of rice. Specific primers for three putative target genes were selected (Supporting Information Table S3). Expression profile of these putative targets was determined by qRT-PCR.

5′ RACE assay

The 5′ RACE was performed with the GeneRacer kit (full-length, RNA ligase-mediated rapid amplification of 5′ and 3′ cDNA ends, RLM-RACE, Invitrogen®) that allows mapping the 5′ end of the cleaved targets. Briefly, RNA (5 µg) from sugarcane plants treated for 48 h at 4 °C was ligated to a 5′ RACE adaptor. Random hexamer primers were then used for cDNA synthesis. PCR amplification of a cDNA fragment containing the cleavage site of the targets was carried out by Nested PCR. The first PCR was carried out using the 5′ RACE outer primer and a gene-specific outer primer (Supporting Information Table S4). The second PCR was carried out using the initial PCR reaction as a template, the 5′ RACE inner primer and gene-specific inner primer (Supporting Information Table S3). RACE fragments were cloned into a pGEM T-easy vector (Invitrogen®) and sequenced.

Whole-mount in situ hybridization (WISH)

In situ hybridization was performed essentially as previously described by de Almeida Engler, Van Montagu & Engler (1994). Sugarcane SP70-1143 plants submitted to 24 h of treatment at 4 °C and of control plants kept at 28 °C were used in these analyses. Roots were hybridized with digoxigenin-labelled GAMyb gene specific antisense, and sense RNA as control probe. After hybridization, slides were washed as described in the same protocol. Slides were dipped in photographic emulsion containing 4.5 µL of nitro blue tetrazolium (NBT) [75 mg mL−1 NBT in 70% v/v (water/dimethylformamide)], 3.5 µL of BCIP (5-bromo-4-chloro-3-indolyphosphate; 50 mg mL−1 in 100% dimethylformamide) and stopped when hybridization signal was detected.

RESULTS

Expression profile of miRNAs in cold-stressed sugarcane plants

In order to study the expression of miRNAs in sugarcane under cold stress, we subjected in vitro-grown plants to 4 °C for 0, 3, 6, 12, 24 and 48 h. To validate the low-temperature treatment, we first investigated by qRT-PCR the expression of two marker genes –SsNAC23 and CoReP– in control and treated plants. The results showed that SsNAC23 was induced approximately 3-fold, 9.5-fold and 16.14-fold at 12, 24 and 48 h, respectively. CoReP was induced approximately 2.5-fold, 5.6-fold and 4.5-fold at 3, 24 and 48 h, respectively (Supporting Information Fig. S1), indicating that cold stress treatment was effective. Next, the expression profile of 12 miRNAs was carried out by the stem–loop RT-PCR method (Varkonyi-Gasic et al. 2007). These miRNAs – miR156, miR159, miR160, miR166, miR167, miR169, miR172, miR319, miR393, miR394, miR408, miR528 – have been implicated in stress caused by cold, salt and drought (Liu et al. 2008; Zhou et al. 2008; Ding et al. 2009a). Analysis of the relative expression of all these miRNAs (Fig. 1) showed that only miRNA319 was differentially expressed, with an approximately threefold increase in expression after 24 h of cold stress, returning to the baseline at the point of 48 h of treatment. Other miRNAs showed small differences in expression compared with the control points at 24 and 48 h, and therefore their regulation was not assessed in this work.

Figure 1.

Analysis of sugarcane miRNA expression in response to cold. Entire plants of sugarcane SP70-1143 grown in vitro were subjected to 4 °C. A set of five plants was collected after 0, 3, 6, 12, 24 and 48 h of treatment. Every point of the experiment had its own control. Analysis of the expression of 12 mature miRNAs was performed by pulsed real-time PCR (RT-PCR). The level of transcript is represented as a ratio (relative expression) of the absolute value of expression of each miRNA divided by the absolute value of expression in its respective control. The housekeeping gene used as control was sugarcane GAPDH. Error bars show range of duplicate analyses of sample in qRT-PCR.

To examine whether miR319 induction by cold stress was differentially regulated in the plant organs, we analysed the miRNA expression profile in roots and shoots by stem–loop RT-PCR method. The experiment was first validated by verifying the expression profile of the cold stress marker genes (Supporting Information Fig. S2). The results of qRT-PCR showed that there was an induction of miR319 in the set of plants treated for 12 and 24 h in both tissues (Fig. 2a). This induction was higher in the root (2.5-fold, approximately). In addition, after 48 h at 4 °C the expression of miR319 was down-regulated in roots and remained at baseline in the shoot. Then, we used the northern blot analysis to confirm these results in the root (Fig. 2b). The result of miR319 expression was similar to that verified by stem–loop RT-PCR, and we noted an unambiguous differential expression of miR319 in the roots. In contrast, northern blots also showed the constitutive expression of miR159 (in roots of sugarcane subjected to cold stress).

Figure 2.

MiRNA expression analysed separately in root and shoot of sugarcane SP70-1143 in response to cold. Plants of sugarcane grown in vitro were subjected to 4 °C. A set of five plants was collected after 0, 12, 24 and 48 h of treatment. (a) Analysis of miR319 expression was performed by pulsed real-time PCR (RT-PCR). Expression in the plantlets of 0 h and the control plantlets at each time was set equal to 1. The level of transcript is represented as a ratio (relative expression) of the absolute value of expression of miR319 by the absolute value of expression respective control. The housekeeping gene was GAPDH. Error bars show range of duplicate analyses of sample in qRT-PCR. (b) Northern blot of miR319 and miR159 in root samples.

MiR319 regulation by ABA

Cold stress is regulated by both ABA-dependent and ABA-independent signalling pathways (Nogueira et al. 2003). To verify whether miR319 was regulated by ABA, we performed the analysis of expression in two separate experiments in that plants were treated with this hormone for 0, 3, 6, 12, 24 and 48 h. Firstly, we analysed the gene expression profile SsNAC23. NAC family genes are induced by ABA (Fujita et al. 2004; Tran et al. 2004). The rice ortholog of SsNAC23, OsNAC6, is regulated in response to treatment with exogenous ABA. Increased levels of relative expression of SsNAC23 were observed in plants treated with ABA (Supporting Information Fig. S3). Curiously, miRNA319 induction by ABA showed a two-step response (Fig. 3). In sugarcane (SP70-1143) treated with ABA, miR319 expression increased approximately twofold after 3 h; decreasing to half the initial levels at 6 h. At 12 h, miRNA319 levels started to increase again, with a maximum of 2.5-fold induction at 24 h. A second independent experiment also showed a two-step response (Supporting Information Fig. S4). However the maximum expression of miR319 was observed in 48 h.

Figure 3.

MiR319 expression profile in response to phytohormone abscisic acid (ABA) treatment. Sugarcane plantlets SP70-1143 grown in vitro were treated with ABA. A set of three plants at each time point was collected after 0, 3, 6, 12, 24 and 48 h of treatment. The control plants were maintained in medium containing Murashige and Skoog (MS) salts and rooting hormones. Analysis of the expression of miR319 was performed by pulsed real-time PCR (RT-PCR), as described in Methods. The level of transcript is represented as a ratio (relative expression) of the absolute value of expression of miR319 divided by the absolute value of expression in the respective control. The housekeeping gene was GAPDH. Error bars show range of duplicate analyses of sample in qRT-PCR.

Identification of miR319 target genes in sugarcane

The high degree of homology between the mature sequence of the miRNAs and the cleavage site in their targets allows the identification of putative miRNA targets in the databases. In addition, the identification of miRNA targets provides important clues to the biological processes involved. To identify the putative miR319 targets in sugarcane, we used two different bioinformatic tools. The miRU: Plant microRNA Potential Target Finder did not identify potential miR319 targets in the sugarcane genome. However, using the Web MicroRNA Designer, six potential targets were identified (Table 1). Because the rice cDNA database is more complete, we also performed a search for possible targets of miR319 in rice. Four potential targets were identified, and a comparison with the sugarcane database identified two of the potential targets, one Myb transcription factor (GAMyb, TC94752) and one TCP transcription factor (PCF6, TC111376). In rice, another TCP transcription factor (PCF5, TC95766) was identified as a putative target. However, as the sugarcane homologue expressed sequence tag (EST) is not full length, the putative miR319 cleavage site could not be identified.

Table 1.  Putative miR319 targets genes found in sugarcane and rice
Gene familyAccess codes for riceAccess codes for sugarcane
  1. Gene families of miR319 targets found on the Web MicroRNA Designer and annotation of the expressed sequence tag (EST) deposited in the database. In this table are reported the targets in the rice genome that also appear in the database of sugarcane using the tool Blast Gene Index and all possible access codes for the sequence of sugarcane.

TCP-PCF6Os07g05720.1; Os03g57190.1TC111376
TCP-PCF5Os01g11550.1TC95766
GAMybOs01g59660.3TC94752
Protein kinaseTC67079
Unknown proteinCA281259
Fasciclin-like proteinCA210555

As a first step to understand the regulation of the putative miRNA319 targets, the expression profiles of three possible targets (PCF5, PCF6 and GAMyb) were verified by qRT-PCR using the same cDNA samples used for the analysis of expression of mature miRNAs. The results showed a negative correlation among the expression patterns of the three possible targets and miR319 (Fig. 4). Plants that received cold treatment for 48 h showed repression of putative target gene expression; the relative expression of PCF6 was about 50% of the control. A 20% reduction in expression of the GAMyb gene in plants was exposed to cold for 24 h (20%), and a more pronounced reduction (approximately 45% repression) was observed after 48 h of cold. PCF5 showed the lowest decrease in gene expression, with a repression of about 25% compared with controls. These results indicate a possible regulation by miR319.

Figure 4.

Expression profiling of three putative miR319 targets. Sugarcane SP70-1143 grown in vitro was subjected to 4 °C. A set of five plants at each time point was collected after 0, 3, 6, 12, 24 and 48 h of treatment. Expression in the plantlets at 0 h and in control plantlets was set equal to 1. Controls were plants of the same age that received treatment and were maintained at 28 °C. The analysis of real-time PCR was performed as described in Materials and Methods. The level of transcript is represented as a ratio (relative expression) of the absolute value of expression of potential target genes divided by the absolute value of expression in the respective control. The housekeeping gene was GAPDH. Error bars show range of duplicate analyses of sample in real-time PCR.

In order to corroborate whether these genes are direct targets of repression by miRNA319-directed cleavage of the mRNA, we used the modified 5′ RACE technique for mapping the cleavage sites. We used RNA from plants incubated for 48 h at 4 °C, because at this point the largest repression of targets was observed. Cleavage of the miR319 targets PCF6 and GAMyb was confirmed by 5′ RACE assay. PCR RACE products were cloned and sequenced. Seven of ten PCF6 gene clones sequenced showed the cleavage site of the target after the tenth nucleotide from the 5′ end of the miR319 (Fig. 5a). Out of the ten GAMyb gene clones sequenced all showed the same site of cleavage following the ninth nucleotide from the 5′ end of the miRNA (Fig. 5b). These results show that both genes were targeted for cleavage by miRNA319.

Figure 5.

Experimental validation of predicted miRNA targets. Bottom strands depict target mRNA sequences from PCF6 (a) and GAMyB (b); top strands depict the miR319. Matches (Watson–Crick pairing) are indicated with vertical dashes, mismatches are unmarked. Arrows indicate cleavage sites verified by 5′ RACE assay, with the number of cloned products shown above.

In order to verify whether the regulation of miRNA319 was spatially regulated in plants subjected to cold stress, the expression profile of these two target genes were analysed in roots and shoots of sugarcane. The results showed different patterns in the relative expression of GAMyb and PCF6 (Fig. 6). PCF6 was induced in both roots and shoots incubated at 4 °C for 24 h, followed by a repression at 48 h. In contrast, GAMyb was down-regulated in both organs after 24 h until 48 h of cold stress.

Figure 6.

Expression profiling of miR319 targets (PCF6 and GAMyB). Sugarcane SP70-1143 grown in vitro was subjected to 4 °C. A set of five plants was collected at each time after 0, 12, 24 and 48 h of treatment. Expression in the plantlets at 0 h and in control plantlets was set equal to 1. Controls were plants of the same age that received treatment and were maintained at 28 °C. The analysis of real-time PCR was performed as described in Materials and Methods, and expression of the targets of miR319 was analysed separately for root and shoot. The level of transcript is represented as a ratio (relative expression) of the absolute value of expression of potential target genes divided by the absolute value of expression in the respective control. The housekeeping gene was GAPDH. Error bars show range of duplicate analyses of sample in real-time PCR.

MiR319 target genes during ABA stress in sugarcane

In order to verify whether the targets of miR319 were also part of the ABA-dependent regulatory pathway, we analysed the expression profile of PCF6 and GAMyb in the samples treated with ABA. The results showed that PCF6 had a 20% reduction in relative expression in plants treated with ABA at 6 and 24 h (Fig. 7). At 48 h PCF6 mRNA levels were similar to control plants. The expression of GAMyb was quickly reduced after 3 h of treatment, rising again to the control level after 24 h. There was a repression of gene expression in plants that received treatment for 3, 6 and 12 h (about 30, 20 and 40%, respectively). Similar results were observed in the second experiment, but the GAMyb was weakly induced in plantlets treated for ABA during 24 h (Supporting Information Fig. S5). These results may mean that, despite the accumulation of miR319, these genes had their expression enhanced by ABA.

Figure 7.

Expression profiling of two miR319 targets (PCF6 and GAMyB) treated with phytohormone abscisic acid. A set of three SP70-1143 plants was collected at each time after 0, 3, 6, 12, 24 and 48 h of treatment in vitro. Expression in the plantlets at 0 h and in control plantlets was set equal to 1. Controls plants were maintained in medium containing Murashige and Skoog (MS) salts and rooting hormones and were the same age as those that received treatment. Analysis of real-time PCR was performed as described in Materials and Methods. The level of transcript is represented as a ratio (relative expression) of the absolute value of expression of potential target genes divided by the absolute value of expression in the respective control. The housekeeping gene was GAPDH. Error bars show range of duplicate analyses of sample in real-time PCR.

Regulation of miR319 in sugarcane varieties submitted to cold stress

To explore the possibility that miR319 and its targets could be used as molecular markers for selection of cold-tolerant sugarcane varieties, we analysed the regulation of miR319 in two varieties of sugarcane subjected to cold stress. The two varieties are considered to have contrasting tolerance to cold stress; RB931011 is a recently released new cultivar, developed in the northeast of Brazil, where the average temperature is approximately 25 °C and very rarely drops to below 20 °C. In contrast, TAMBO FEPAGRO is a native cultivar that grows in the mountain regions in the south of Brazil, where the winter temperature will occasionally reach 0 °C. In long-term field trials, TAMBO FEPAGRO is capable of resisting severe cold stress. In particular, the damage to lateral buds at freezing temperature was less severe than in other varieties (data not shown).

Plants germinated from the stalk were grown in sand and vermiculite (2:1) for approximately 6 weeks at 28 °C. Whole plants in pots were then incubated at 4 °C for 24 and 48 h and RNA was extracted from shoots. In both varieties, miR319 was up-regulated in plants that received the treatment of cold stress for 48 h. In RB931011 there was an early increase in mature miRNA levels after 24 h (1.67-fold), peaking at 48 h (4.2-fold) (Fig. 8a). In contrast, miRNA319 levels in TAMBO PEPAGRO at 24 h were similar to control plants, rising only after 48 h (Fig. 8b). Next, we tested whether the change in miRNA319 expression was followed by inverse regulation of its targets. The results showed that while GAMyb was down-regulated, PCF6 was up-regulated in both sugarcane varieties subjected to cold stress for 24 and 48 h (Fig. 8b).

Figure 8.

Regulation of miR319 in genotypes of sugarcane of contrasting cold tolerance. TAMBO FEPAGRO – tolerant; RB931011 – sensitive. Plants of sugarcane were germinated from the stalk and grown in sand and vermiculite (2:1) for about 6 weeks and then received cold treatment for 0, 24 and 48 h. Expression in the plantlets at 0 h and in control plantlets was set equal to 1. (a) miR319 expression was performed by the method of pulsed real-time PCR (RT-PCR). (b) Expression profiling on two miR319 targets (PCF6 and GAMYB). Analysis of qRT-PCR was performed as described in Materials and Methods. The level of transcript is represented as a ratio (relative expression) of the absolute value of expression of the potential target gene divided by the absolute value of expression in the respective control. The housekeeping gene was GAPDH. Error bars show range of duplicate analyses of sample in qRT-PCR.

In situ localization of GAMyb transcript

We observed that the GAMyb mRNA is regulated by miR319 in different varieties of sugarcane exposed to the cold stress. To examine the spatial distribution of GAMyb expression, in situ hybridization with roots of sugarcane cultivar SP70-1143 submitted to 24 h of treatment at 4 °C and roots of plants of control plants kept at 28 °C was carried out. The location of the GAMyb expression in root sugarcane tissue is shown in Fig. 9. In situ hybridization with sugarcane roots showed that GAMyb transcripts are accumulated in the root tips and the signals are strongly reduced by cold stress. No signals were detected by hybridization with the sense probe (data not shown).

Figure 9.

In situ localization of GAMyb in sugarcane root. The signal is seen as a purple stain under bright-field optics. (a) is a root of sugarcane which was submitted to cold stress, and (b) is a root of sugarcane which has not received the cold stress treatment (it kept at 28 °C). In both cases, the roots were hybridized with dioxigenin-labelled GAMyb gene specific antisense. Bars = 100 µm.

DISCUSSION

Genes that are regulated by low-temperature stress in sugarcane have been identified; however, little is known about the epigenetic regulation of cold stress (Nogueira et al. 2003, 2005). Recent reports have shown the involvement of miRNAs in response to abiotic stress in plants such as Arabidopsis, rice and maize (Zhao et al. 2007; Liu et al. 2008; Zhou et al. 2008; Ding et al. 2009a; Covarrubias & Reyes 2010). Furthermore, it has been shown that miRNAs induced by abiotic stress are repressors of their target genes, which are possible determinants of the stress response (Phillips et al. 2007). In this study, we observed that miR319 was differentially expressed during cold treatment for different periods of time (Fig. 1). The data also show that miR319 was up-regulated by cold stress in both roots and shoots (Fig. 2a,b). However, this induction was stronger in roots because in vitro liquid culture was used which might induce a greater response in root system.

Our result corroborates studies in Arabidopsis in which miR319 was up-regulated in response to cold stress (Sunkar & Zhu 2004; Liu et al. 2008). MiR319, previously known as ‘miR-JAW’, was first described in Arabidopsis because of its involvement in the control of leaf morphogenesis (Palatnik et al. 2003). In addition, miR319 is implicated in numerous different biological processes, ranging from GA signalling, transition to flowering, control of gamete formation, regulation of the biosynthesis of the hormone ethylene, ABA signalling and morphogenesis of the leaf response to cold stress and salinity (Rhoades et al. 2002; Jones-Rhoades & Bartel 2004; Sunkar & Zhu 2004; Schwab et al. 2005; Reyes & Chua 2007; Liu et al. 2008).

In this study, we observed that miR319 was up-regulated by both cold and treatment with ABA (Fig. 3). The expression of genes regulated by abiotic stresses can also be regulated by ABA, and molecular analysis has revealed that ABA has an important role in plant adaptation to abiotic stresses (Shinozaki & Yamaguchi-Shinozaki 1996; Seki et al. 2007). Moreover, Nogueira et al. (2003) showed that in sugarcane there are pathways for cold response that are dependent on and independent of ABA.

In Arabidopsis, miR319 targets genes from two classes of transcription factors, TCP-like and MYB-like, described as plant growth regulators (Palatnik et al. 2003, 2007; Woodger et al. 2003; Navaud et al. 2007; Schommer et al. 2008; Zhang et al. 2008). Furthermore, Genevestigator analyses of microarray experiments from Arabidopsis suggest that many of the miRNAs target genes involved in regulating plant growth and development can be cold-responsive genes (Zimmermann et al. 2004). Likewise, in sugarcane we identified two genes regulated by miR319: TCP6 and GAMyb– that belong to these two families of transcription factors. However, phylogenetic analyses of described TCP- and MYB-like genes show that the sugarcane genes group in a separate clade (Supporting Information Fig. S6), peculiar to the Poaceae family, suggesting that they may regulate signalling pathways unique to this family. The cleavage sites were mapped in the miR319 target genes PCF6 and GAMYB (Fig. 5a,b, respectively) confirming mi319-directed cleavage. Palatnik et al. (2007) already demonstrated statistically and experimentally that miR319 is efficient in cleavage of both TCP and MYB genes in Arabidopsis. Moreover, as described previously for Arabidopsis, the cleavage of genes of the TCP family by miR319 occurs after the tenth nucleotide; and in the genes of the MYB family after the ninth nucleotide from the 5′ end of the miRNA paired with the target (Palatnik et al. 2003).

Decrease in steady-state levels of mRNAs from both targets occurred in plants grown in vitro and exposed to cold for 48 h. Furthermore, regulation of the cold response mediated by miR319 appears to be functioning in the whole plant, because induction of miR319 and repression of PCF6 and GAMyb occur in both roots and shoots. However, in ABA-treated plantlets the miR319 target genes were differentially expressed, but a direct regulation by the miRNA was not observed in all points of the experiment. While a small decrease in mRNA levels occurred after 12 h, especially for GAMyb, from 24 to 48 h there was an increase in mRNA levels of both targets (Fig. 7). It is possible that, while both messengers are still cleaved by miR319, the genes are induced by ABA and that this induction in expression balances the destruction of mRNAs. Indeed, up-regulation by ABA has been described for members of the family of MYB transcription (Ding et al. 2009b). Alternatively, under stress conditions, Jeong et al. (2009) have shown that target mRNA levels are not at all times reduced by the increase of miRNA expression even when cleavage of the matching target mRNA by miRNA occurs. Thus, they concluded that post-transcriptional regulation of target by miRNA-directed cleavage is not always limiting for the accumulation of mRNA under the stress conditions.

Based on this study, a hypothetical model of miR319 regulation in sugarcane exposed to 4 °C was proposed (Fig. 10). Exposure of sugarcane to cold induces increased miR319 expression, which regulates the expression of the target genes PCF6 and GAMyb by cleavage. Based on studies of the TCP gene family in Arabidopsis, PCF6 gene repression may result in abnormal development of leaves (Koyama et al. 2007). In barley, the GAMyb gene is responsible for anther development, stem elongation, early flowering and seed development (Woodger et al. 2003). Thus, a decrease in GAMyb mRNA levels in sugarcane plants exposed to cold may result in inhibition of growth and plant development. In addition, ABA also seems to regulate miRNA319 signalling responses in sugarcane. A two-step regulation of miRNA319 is observed, with an early increase after 3 h, and a second peak later. However, we observed that treatment with ABA does not decrease expression of genes targeted by miR319 at all points of the experiment. Thus, this result suggests that there is a balance between these two regulatory pathways, and as a result the expression profile is different from that observed for plants receiving cold treatment.

Figure 10.

Hypothetical model of the sugarcane response to cold and phytohormone abscisic acid (ABA) treatment. In cold stress the miR319 is up-regulated, leading to repression of its targets, which in turn inhibits the growth and development of the plant. In plants treated with ABA, miR319 (arrow in grey – solid line) is up-regulated. This miRNA directs the cleavage of its targets; however, it is possible that ABA treatment induces the expression of genes PCF6 and GAMyB (arrows in grey – dashed line). The balance of this gene expression may be a result of the balance between induction by treatment with ABA and cleavage by miR319.

MiR319 regulation was also observed in plants grown in a greenhouse. For the analysis of miR319 regulation, only leaves were used, because previous results, by stem–loop RT-PCR with plants grown in vitro, showed that similar changes in miRNA319 regulation occurred in both roots and shoots. MiR319 was up-regulated in both varieties of sugarcane exposed to cold stress. However, there was a temporal difference in expression between varieties, with an increase in miRNA319 levels earlier in the sensitive variety than in the tolerant variety. Curiously, in contrast with plants grown in vitro, we did not observe a reduction in the mRNA levels of PCF6 at any time on the contrary, an accumulation of PCF6 mRNA occurred in leaves of both cultivars under the conditions tested. However, at this stage we cannot conclude whether this difference is due to the experimental conditions employed – plants grown in vitro versus in vivo grown plants – or to some other factor affecting the regulation of PCF6. However, the decrease in mRNA levels of the GAMyb gene is delayed as well in the tolerant cultivar TAMBO FEPAGRO. Furthermore, as described for GAMyb-like genes in Arabidopsis (Millar & Gubler 2005), we observed the localization of GAMyb transcript in the root tip sugarcane plants (Fig. 9). As described in rice and Arabidopsis (Gocal et al. 2001; Kaneko et al. 2004), the expression in the root tip could indicate that GAMyb gene is involved in the growth and development of the sugarcane roots. Strikingly, in root tips submitted to cold stress, there is a complete disappearance of GAMyb transcripts. This result suggests that at least one of the mechanisms leading to the inhibition of plant growth by cold stress is caused by miRNA319-mediated cleavage of its targets.

Although one of the major challenges to the expansion of areas for growing sugarcane is tolerance to low temperatures and frost, decade-long efforts to identify and breed for cold tolerance have had very little success. Effectively, previous efforts (Duncan & Cooke 1932; Irvine 1968; Cesnik, Bassinello & Oliveira 1978; Tai 1981) failed to produce new cold-tolerant cultivars. The identification of the native cultivar TAMBO FEPAGRO was fortuitous; it only happened because plants of this cultivar survived frequent frost and below-freezing temperatures in a mountain region in the South of Brazil (Caren Lamb, unpublished results). Our results show differences in miR319 regulation among sugarcane genotypes contrasting in tolerance to cold. These results suggest that differences in the intensity of the regulation of miR319 and its targets could be tested as markers for selection of cold-tolerant sugarcane cultivars.

ACKNOWLEDGMENTS

We are grateful to Drs Laureen Kido and Andrea Barros from CETENE for providing the RB931011 sugarcane cultivar. We thank Martha Sorenson (UFRJ) for language editing. We thank Milos Tanurdzic (Cold Spring Harbor Laboratory) for critical reading and language editing. F.T., C.A.R., K.L.A., A.S.H. and P.C.G.F. are indebted to the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) for fellowships. The study was partially supported by funds from the Instituto Nacional de Ciência e Tecnologia em Fixação Biológica de Nitrogênio (INCT), and by the CAPES/Cofecub Program.

Ancillary