•We have reported previously that the gibberellin (GA) content in strawberry receptacle is high, peaking at specific stages, pointing to a role of this hormone in fruit development. In Arabidopsis, miR159 levels are dependent on GA concentration. This prompted us to investigate the role of two members of the miR159 family and their putative strawberry target gene, GAMYB, in relation to changes in GA content during the course of fruit development.
•The highest expression level of the two Fa-MIR159 genes was in the fruit’s receptacle tissue, with dramatic changes observed throughout development. The lowest levels of total mature miR159 (a and b) were observed during the white stage of receptacle development, which was concurrent with the highest expression of Fa-GAMYB. A functional interaction between miR159 and Fa-GAMYB has been demonstrated in receptacle tissue.
•The application of bioactive GA (i.e. GA3) to strawberry plants caused the down-regulated expression of Fa-MIR159a, but the expression of Fa-MIR159b was not affected significantly. Clear discrepancies between Fa-MIR159b and mature Fa-miR159b levels were indicative of post-transcriptional regulation of Fa-MIR159b gene expression.
•We propose that Fa-miR159a and Fa-miR159b interact with Fa-GAMYB during the course of strawberry receptacle development, and that they act in a cooperative fashion to respond, in part, to changes in GA endogenous levels.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Strawberries (Fragaria × ananassa) are a commercial crop in the Rosaceae family in the genus Fragaria. This edible fruit is highly appreciated for its organoleptic properties and health-promoting benefits (Seeram, 2008). The strawberry fruit is, in reality, a false fruit composed of the achenes (true fruits) and the receptacle, which arises from the development of the flower receptacle. During strawberry development, both the achenes and the receptacle follow separate patterns, but their developmental programmes are coordinated. Indeed, a recent study performed parallel profiling of primary and secondary metabolism in both the strawberry receptacle and the achene across several developmental stages. The results displayed a high level of metabolic synchrony and specialization during development (Fait et al., 2008), and it has been known for some time that auxin plays a major role in this developmental coordination. Auxin is synthesized in the achenes and promotes receptacle growth during the early stages of fruit development (Nitsch, 1950). Free auxin reaches its maximum level in both the achene and the receptacle during the green stage and subsequently declines as the fruit ripens (Manning, 1998). The involvement of other hormones during strawberry ripening and the obligated interaction between the achene and the receptacle are unclear. Many endogenous gibberellins (GAs) have been identified in strawberry fruit (Blake et al., 2000), but there have been no conclusive results on the role of this hormone during fruit development and ripening (Bustamante et al., 2009). Recently, we have reported that bioactive GA4 is notably abundant in the receptacle, with peak levels being observed during the white stage of development. We also identified the GID1-DELLA pathway as being operative in this tissue (Csukasi et al., 2011).
The importance of microRNA (miRNA)-directed gene regulation during plant development has been widely documented (Jones-Rhoades et al., 2006). Specifically, miRNAs regulate gene expression in animals and plants through the degradation and translational arrest of their target mRNAs. In plants, the ‘mature’ miRNAs are c. 21 nucleotides long and control such processes as stem cell formation, organ identity, leaf polarity, vascular differentiation and cell division patterns (Jones-Rhoades et al., 2006). The mature miRNAs are processed by Dicer (DCL1), a type III RNase, from the longer, noncoding RNA precursors, which are known as primary (pri) miRNAs. The Arabidopsis genome contains c. 180 miRNA loci, many of which are highly conserved between plant species (Axtell & Bartel, 2005; Rajagopalan et al., 2006). However, information on the role of miRNAs in the most important fruit crops, and their contributions to processes such as fruit development and ripening, is limited. Northern blotting and deep sequencing of short RNAs in tomato fruit identified several conserved and nonconserved miRNAs, and several mRNAs associated with fruit development were validated as tomato miRNA targets (Pilcher et al., 2007; Moxon et al., 2008). Similarly, several conserved miRNAs and grape-specific miRNAs were identified by massive sequencing in grapes and were functionally validated experimentally. Interestingly, the expression levels of the nonconserved miRNAs were slightly higher in ripe berries than in vegetative tissues (Carra et al., 2009). Using microarray technology, differential miRNA expression was observed in both conventional and micropropagated strawberries (Li et al., 2009).
In this study, we investigated the involvement of two members of the Fa-miR159 family and one Fa-GAMYB member during strawberry fruit development and ripening. These three genes showed tissue- and developmental stage-specific expression patterns, notably in the receptacle of the fruit. A functional interaction between Fa-miR159 and Fa-GAMYB was also demonstrated in this tissue. Our results point to a role of two Fa-miR159 family members in receptacle development. The Fa-miR159 signal is channelled through Fa-GAMYB, possibly as a response to changes in the endogenous GA levels in this tissue.
Materials and Methods
Plant material, growth conditions and hormone treatment
Strawberry plants (Fragaria× ananassa Duchesne) of cv Camarosa were grown under field conditions in Huelva in the southwest of Spain. Samples were harvested at five different developmental stages, namely the closed and open flowers (CF, OF) and during the green (G), white (W) and red (R) fruit stages. The vegetative tissues collected for analysis were the leaves (L) and roots (Rt). All tissues were flash frozen immediately in liquid nitrogen and stored at −80°C until use.
For the GA3 treatment, plants of the same cultivar were grown in vitro in Murashige and Skoog (MS) medium (Murashige & Skoog, 1962) supplemented with 15 g l−1 sucrose and 6% agar (w/v), and a solution of 100 μM GA3 in water was added. Control plants were grown in the same medium without GA3. Growth conditions for all the in vitro strawberry plants, as well as the Arabidopsis plants, were identical. Short days (SDs) were made up of16/8 h of light (200 μmol m−2 s−2)/dark and a 24/17°C regimen, and long days (LDs) were 8/16 h and 23/17°C.
RNA extraction and gene expression analysis
Total RNA was extracted from strawberry tissues as described previously (Manning, 1991). Total RNA from Arabidopsis was extracted from the rosette tissues using TRIzol® Reagent (http://www.invitrogen.com/) according to the manufacturer’s instructions. DNAseI (Sigma, AMPD1) was applied to the samples to eliminate DNA from RNA samples according to the manufacturer’s instructions. Reverse transcription-polymerase chain reactions (RT-PCRs) were performed with 1 μg of total RNA using the iScript cDNA Synthesis kit from Bio-Rad (http://www.bio-rad.com) according to the manufacturer’s instructions.
Gene expression analysis was performed by real-time quantitative PCR using the fluorescent intercalating dye, SYBR Green, in an iCycler detection system (Bio-Rad). Reactions were performed in triplicate. The absence of primer dimers was confirmed by melting curve examination. FaRib413 was used to normalize the data because it was identified by northern analysis to be constitutively expressed in a variety of strawberry plant tissues, including during the various fruit stages (Casado-Diaz et al., 2006). Relative expression level quantification was performed using the comparative Ct method (Pfaffl, 2001), where a value of unity was given to the sample with the minimum expression. Primers used for expression analysis are shown in Supporting Information Table S1; the minimal efficiency considered for the primers was 90%. The specificity of the primers was verified before their use for quantitative RT-PCR; no amplification of FaMIR159a was observed when FaMIR159b primers were used and vice versa; also no amplification of AtMIR159 genes was observed with strawberry primers.
For the quantitative stem-loop RT-PCR of the mature miR159, RT reactions were performed in the presence of 2 mM stem-loop miR159a RT primer (5′-GCCTCTCATGCTGACGAATTTTGAGAGGCTAGAGCTCC-3′, Roche) or miR59b RT primer (5′-GCCTCTCATGCTGACGAATTTTGAGAGGCTAGATCA-3′) as described previously with several modifications (Bombarely et al., 2010). RT reactions were incubated in a thermocycler for 30 min at 16°C, 30 min at 42°C and 5 min at 85°C. All reactions, including RT minus controls, were performed in duplicate. The RT products were diluted to a final volume of 200 ng μl−1 to avoid potential primer interference in the subsequent quantitative PCR. Next, quantitative PCR was performed on a Corbett RG6000 thermocycler in a total volume of 15 μl. The reaction included 3 μl of diluted RT product, 2 × FastStart Universal Probe Master Mix (Roche), 0.2 mM of TaqMan probe complementary to miR159a (5′ FAM-TTGAGAGGCTAGAGCTCCCTTCA-BBQ 3′, Roche) or to miR159b (5′ FAM-TTGAGAGGCTAGAGCTCCCTTCA-BBQ 3′, Roche) and 0.5 mM of each PCR primer (i.e. miR159a-F 5′-GAATTCGACCCTTTGGATTG-3′, miR159b-F 5′-GAATTCGACCCTTTGGATTG-3′ and miR159-R 5′-GCCTCTCATGCTGACGAAT-3′ (Roche)). The reactions were incubated in 0.1-ml tubes at 95°C for 10 min, followed by 40 cycles at 95°C for 15 s and at 60°C for 1 min. All reactions were performed in duplicate in three different PCR runs. The concentration of Fa-miR159 was calculated by converting the Ct value into an absolute copy number using a standard curve generated from a diluted series of synthetic miR159a or miR159b RNA oligos and RotorGene 6000 software (Corbett).
miRNA-guided cleavage validation
A modified RNA ligase-mediated rapid amplification of cDNA ends (RLM-RACE) protocol was used to map internal miRNA-directed cleavage sites of predicted cDNA targets (Llave et al., 2002). Total RNA was directly ligated to the 5′ RNA adapter without any further enzymatic pretreatment. Ligated RNA was reverse transcribed using gene-specific primers that annealed c. 300 nucleotides downstream of the predicted cleavage site of the target mRNA. PCR amplification of the first-strand cDNA was performed using a reverse gene-specific primer and a forward primer derived from the RNA adapter sequence. The 5′ RACE amplification products were gel purified, cloned and sequenced.
Plasmid construction and plant transformation
A partial cDNA clone of the miR159a primary transcript was isolated from a cDNA library prepared from the green fruits of strawberries after extensive sequencing of several thousand clones (Bombarely et al., 2010). A full-length clone was obtained using the Invitrogen 5′ RACE system for the Rapid Amplification of cDNA Ends, version 2.0, according to the manufacturer’s instructions. Full-length cDNAs of the Fa-MIR159b and Fa-GAMYB genes were obtained by PCR from cDNA prepared from the RNA extracted from fruits during different developmental stages. Gene-specific primers were based on the sequences of the Fragaria vesca expressed sequence tags (ESTs) DY667997.1 and DY669997.1, which displayed high homology to the Arabidopsis pre-miR159 and GAMYB-like genes.
For the 35S::Fa-MIR159a construct, the full-length sequence of Fa-MIR159a was PCR amplified using the primers described in Table S1, subcloned into the pGEM®-T Easy vector (Promega, http://www.promega.com/) and introduced into the NotI/AscI sites of the pENTR/D-TOPO® vector (Invitrogen). The clone was finally inserted into the pKGCS-StrepII vector using the LR reaction (Invitrogen). This vector was introduced into Agrobacterium tumefaciens (GV3101-pMP90-RK) by electroporation (Sambrook & Maniatis, 1989) and transformed into Arabidopsis using the floral dip method (Clough, 2004).
Northern blot hybridization
RNA blot analysis using 5 μg of total RNA was performed as described by Llave et al. (2000). Briefly, a radiolabelled DNA probe from the cloned 5′ RACE Fa-GAMYB product was created by a random priming reaction in the presence of 32P-dCTP. The blot was both prehybridized and hybridized at 38°C using the Perfect-Hybrid buffer (Sigma). Ethidium bromide staining of the gel before the blot transfer was used as the loading control. The hybridization signals were measured using densitometry on the RNA blot exposed to autoradiography films.
Identification of strawberry miR159 and GAMYB gene homologues
The high content of active GA4 in the receptacle of strawberry fruit (Csukasi et al., 2011) and the fact that the GA-DELLA signalling pathway regulates miR159 in Arabidopsis (Achard et al., 2004) prompted us to investigate a possible role for this miRNA during strawberry fruit development. The complete cDNA sequences corresponding to the homologues of two Arabidopsis MIR159 genes (Fa-MIR159a and Fa-MIR159b) were cloned from strawberries. Sequence analyses revealed striking differences in their nucleotide composition. Pairwise sequence alignment of the primary transcript regions corresponding to the Fa-miR159 precursors (pre-Fa-miR159a and pre-Fa-miR159b) revealed a degree of identity of 53% between them (Fig. 1a). As observed in Arabidopsis (Allen et al., 2007), different polyadenylation sites were identified in the strawberry primary transcripts of Fa-miR159a and Fa-miR159b (Supporting Information Fig. S1a). Four putative polyadenylation sites were identified in Fa-MIR159a, with a 790-nucleotide segment being the largest possible transcript. Three putative polyadenylation sites were predicted for Fa-MIR159b, with the largest transcript being 890 nucleotides in length. The putative Fa-miR159-generating precursors were inferred from secondary structure analysis. The MFOLD program (http://mfold.bioinfo.rpi.edu/) predicted a characteristic stem-loop structure for each pre-Fa-miR159 with free energy values in a range similar to those calculated previously for miRNA precursors (Fig. 1b). As reported for members of the same miRNA family in other species (Jones-Rhoades et al., 2006), the mature miRNA of both precursors was located in the identical stem arm (Fig. 1b). The canonical, mature miR159 sequences derived from each precursor differed in both their terminal 5′ and 3′ end nucleotide positions (Fig. 1d).
A phylogenetic tree was generated using the stem-loop regions of the pre-miR159 sequences of several different species (Fig. 1c). The putative fold-back pre-miR159 sequences from tomatoes (Solanum lycopersicum), grapes (Vitis vinifera) and apples (Malus domestica) were inferred after performing a BLAST search. As shown in Fig. 1(c), pre-Fa-miR159a was grouped with pre-Os-miR159 and pre-Vv-miR159, whereas pre-Fa-miR159b was clustered with pre-At-miR159a and pre-At-miR159b. The mature Fa-miR159a and Fa-miR159b sequences were similar to those of both Arabidopsis and rice (Fig. 1d).
Known miR159 targets include the Arabidopsis GAMYB-like genes MYB33 and MYB65 (Millar & Gubler, 2005), which belong to a gene family encoding transcription factors that are commonly involved in diverse developmental processes. This finding led us to search for members of the GAMYB-like gene family in EST sequences from F. vesca. The high degree of sequence similarity of the F. vesca and Fragaria × ananassa ESTs (Bombarely et al., 2010) allowed for the design of sequence-specific primers to clone the corresponding full-length cDNA from F. ananassa. Nucleotide sequence analyses revealed the presence of a unique miR159 cleavage site that exhibited extended complementarity to both Fa-miR159a and Fa-miR159b. Because the deduced amino acid sequence of the Fa-GAMYB-encoded protein possessed an R2R3 domain (Achard et al., 2004; Tsuji et al., 2006; Gong & Bewley, 2008) and box 1, 2 and 3 conserved motifs (Achard et al., 2004; Tsuji et al., 2006; Gong & Bewley, 2008), both of which are characteristic of transcription factors encoded by GAMYB-like genes, this gene was referred to as Fa-GAMYB (Fig. S1b).
The amino acid sequence of the Fa-GAMYB protein was aligned to the Arabidopsis MYB proteins encoded by genes containing a predicted miR159 binding site (Millar & Gubler, 2005). Interestingly, the Fa-GAMYB protein exhibited the highest degree of similarity to the Arabidopsis GAMYB-like gene-encoded proteins MYB33 and MYB65 (Fig. 2a), and base paired with near-perfect complementarity with Fa-miR159a and Fa-miR159b (Fig. 2b). Both MYB33 and MYB65 were sensitive to miR159-directed regulation via endonucleolytic cleavage of their mRNAs (Achard et al., 2004). In conclusion, we revealed the existence of at least two miR159 loci in the strawberry genome and a putative target gene encoding Fa-GAMYB.
Ectopic expression of Fa-miR159a restores a mir159-defective phenotype in Arabidopsis
We tested the hypothesis that the predicted Fa-MIR159a gene was competent to yield functional, mature miR159 sequences. Genetic transformation often constitutes an overwhelming proof-of-function for a gene product; however, the inherent difficulty associated with strawberry transformation and the lack of appropriate tissue-specific promoters in this species prompted us to use a heterologous system. It has been reported that the Arabidopsis mir159a and mir159b single mutants do not show any severe developmental defects, suggesting that they are functionally redundant (Allen et al., 2007). However, mir159a, mir159b double mutants present reduced apical dominance, curled leaves, shorter siliques and smaller seeds with an irregular shape as a result of the inactivation of the miR159-guided regulatory pathway (Allen et al., 2007). The Arabidopsis mir159a, mir159b double mutant, as well as the wild-type (WT) Columbia ecotype, were transformed with the full-length Fa-MIR159a sequence under the control of the 35S promoter (35S::Fa-MIR159a). The defective mir159a, mir159b phenotype was rescued by expression of Fa-MIR159a in Arabidopsis plants grown using an SD regimen (Fig. 3a). The same complementation pattern was observed in plants cultivated using LD conditions (Fig. S2). These results provided experimental evidence that Fa-MIR159a was effective at complementing both miR159a and mir159b loss-of-function genes in Arabidopsis.
We investigated the phenotypic rescue in miR159 double mutants to determine whether it was the consequence of Fa-miR159-guided negative regulation of Arabidopsis GAMYB target genes. We first confirmed the presence of the Fa-MIR159a transgene in the Arabidopsis mir159a, mir159b double mutants (Fig. 3b). The relative expression of AtMYB65, a known target of miR159 in Arabidopsis (Achard et al., 2004), was determined in both the miR159a, miR159b double mutant plants and in the Fa-MIR159a-complemented transgenic line. AtMYB65 mRNA levels were elevated in the mir159a, mir159b double mutants compared with the WT plants. This result was consistent with the lack of miR159 activity in the mutant (Fig. 3c). By contrast, ectopic expression of the orthologous Fa-MIR159a in the mutant background reduced the expression of AtMYB65 to WT levels (Fig. 3c). Furthermore, ectopic expression of Fa-MIR159a in a WT Arabidopsis background resulted in a down-regulation of the AtMYB65 transcripts to levels below those in the WT (Fig. 3c). Collectively, these results demonstrated that transcripts derived from the predicted strawberry MIR159a locus compensated for the lack of endogenous miR159 activity in Arabidopsis, thereby suggesting that Fa-MIR159a was effectively processed into a functional mature miR159a. Furthermore, transgenic restoration of a miR159-driven regulatory pathway targeting the AtMYB65 mRNA was sufficient to alleviate the developmental defects observed in the miR159a, mir159b double mutant Arabidopsis plants (Fig. 3a).
Fa-GAMYB is a validated target of Fa-miR159 in strawberry fruit
To experimentally validate a functional interaction between Fa-miR159 and its predicted target, Fa-GAMYB, we used a modified 5′ RLM-RACE assay (Llave et al., 2002). To determine whether the miR159-directed cleavage of Fa-GAMYB transcripts was tissue- and/or developmental stage-specific, we used poly-A(+) RNA purified from the G, W and R receptacles. The 5′ end sequencing of the amplified products revealed a predominant cleavage site in the middle of the Fa-miR159 complementarity region (Fig. 4a). Cleavage occurred at an identical position in all tested samples, except in the R receptacle (RR), where we identified a secondary cleavage site one nucleotide upstream of the preferential cleavage site. These results confirmed that Fa-GAMYB is an authentic target of Fa-miR159, and Fa-GAMYB was subjected to miR159-mediated down-regulation in all tissues tested.
Next, we performed an RNA blot analysis using a 3′ proximal Fa-GAMYB radiolabelled probe to assess the extent of miR159-guided negative regulation in all three samples interrogated. Our assay detected hybridization bands consistent with both the full-length Fa-GAMYB mRNA and a shorter 3′ RNA cleavage product, the result of the miR159-guided processing of full-length mRNA in the strawberry fruit receptacle at the three developmental stages sampled (Fig. 4b). Furthermore, the ratio of full-length mRNA to shorter cleavage products was comparable at the G, W and R stages, suggesting an active role for miR159 in the control of Fa-GAMYB mRNA levels during fruit development.
Fa-miR159a and Fa-miR159b have different expression patterns during strawberry fruit development, and their transcript levels are opposite that of Fa-GAMYB
To gain an insight into the outcome of the miR159-Fa-GAMYB mRNA interaction during strawberry fruit development, relative primary transcript levels of both Fa-miR159a and Fa-miR159b were measured by quantitative RT-PCR at several developmental stages (CF and OF, and in G, W and R fruits) and in two vegetative tissues (i.e. L and Rt). Fa-MIR159a exhibited a significant maximum expression in G fruits (Fig. 5a), whereas Fa-MIR159b reached its peak in R fruits (Fig. 5b).
The common ‘strawberry fruit’ is actually a false fruit composed of the achenes and the receptacle, which greatly differ in their origin, cell identities and physiological roles. To gain precision in our analysis, we investigated the expression of the two Fa-MIR159 genes separately, in the achenes or the receptacle, during the G, W and R stages. The expression patterns of both Fa-MIR159a and Fa-MIR159b were similar in the achenes during the three developmental stages (Fig. 5c,d). By contrast, Fa-MIR159a and Fa-MIR159b transcripts were more abundant in the receptacle than in the achenes and exhibited substantial differences during development. The Fa-MIR159a transcript reached its highest expression level during the G stage and subsequently decreased significantly during the W and R stages (Fig. 5c). By contrast, the expression of Fa-MIR159b in the receptacle increased gradually from the G through the W and R stages (Fig. 5d).
Because the expression of miRNA precursors does not always correlate with the steady-state levels of the mature miRNAs (Nogueira et al., 2009), we quantified the mature Fa-miR159a and Fa-miR159b transcript levels in the strawberry receptacles during the G, W and R stages using stem-loop quantitative RT-PCR. Absolute accumulation levels of the two Fa-miRNAs were calculated by converting Ct into an absolute copy number using a standard curve derived from a diluted series of synthetic Fa-miR159a or Fa-miR159b RNA oligonucleotides. The results of the quantitative RT-PCRs suggested a predominant accumulation of Fa-miR159b compared with Fa-miR159a at all stages analysed (Fig. 6a). Quantitative RT-PCR also identified an accumulation pattern of Fa-miR159a that virtually mimicked that of the Fa-MIR159a transcript, displaying maximum levels during the G stage and gradually decreasing towards the R stage. Interestingly, the Fa-miR159b accumulation pattern deviated from that observed for Fa-MIR159b. The level of mature Fa-miR159b was significantly lower in the W receptacle compared with the G and R stages, in contrast with the Fa-MIR159b transcript, which was significantly lower during the G stage. This discrepancy between Fa-MIR159b and Fa-miR159b patterns may reflect differential processing efficiency of the miR159b primary transcript dependent on the developmental stage. In summary, our results indicated that Fa-MIR159a was modestly expressed and converted into mature miR159a during strawberry receptacle development. By contrast, Fa-MIR159b was actively and preferentially transcribed and processed into mature Fa-miR159b at every stage of receptacle development, and its absolute content was 8–25-fold higher than that of Fa-miR159a, depending on the developmental stage analysed (Fig. 6a).
Next, we investigated whether the expression profile of Fa-miR159 members delineated the expression of the predicted target gene, Fa-GAMYB, in the aforementioned tissues and developmental stages. Our results from quantitative RT-PCR revealed that Fa-GAMYB mRNA was expressed in the receptacle at levels significantly higher than in the achenes (Fig. 6b). Furthermore, the highest transcript level in the receptacle was repeatedly observed during the W stage, when less miR159 transcript was detected. This observation regarding Fa-GAMYB mRNA was not so clear in the northern blotting studies of Fig. 4(b), probably because of the different sensitivity and/or specificity of these two techniques for the quantification of transcript levels. Conversely, Fa-GAMYB mRNA levels in the receptacle were substantially diminished during the G and R stages; in these stages, the total miR159 levels were higher (Fig. 6a). On the basis of our data, we can conclude that the relative accumulation of GAMYB mRNA is strongly and negatively dependent on the abundance of miR159 in the strawberry receptacle. These receptacle expression patterns are suggestive of Fa-miR159a and Fa-miR159b controlling Fa-GAMYB expression in strawberry fruit.
Because miR159 was reported to be up-regulated by GA in Arabidopsis (Achard et al., 2004), we hypothesized that this hormone might influence, at least partially, the expression of strawberry Fa-MIR159a and Fa-MIR159b. To test this possibility, in vitro strawberry plantlets were treated with bioactive GA (GA3). The direct GA3 treatment of strawberry fruits was not feasible because, similar to other fruits, they are recalcitrant to standardized external hormone application, mostly because of the thickness of the fruit cuticle. After 6 h of GA3 treatment, the relative accumulation of Fa-MIR159a was drastically reduced by nearly 25-fold relative to the level in untreated control plants (Fig. 7a). Conversely, Fa-MIR159b transcripts did not respond at such a high level to GA3 application (Fig. 7b). At this time, no significant effect on Fa-GAMYB expression was observed (Fig. 7c), but this was somehow expected, as a 24–36-h delay has been reported for a change in GAMYB expression in response to GA treatment (Tsuji et al., 2006; Gong & Bewley, 2008). Treatment effectiveness was confirmed by monitoring the effect of GA3 application on the transcript accumulation of two GA-responsive genes, FaGA3ox and FaGA2ox. After 2 h of GA3 treatment, the two genes were down- and up-regulated, respectively (Fig. 7d,e), as predicted on the basis of previous reports (Yamaguchi, 2008). In conclusion, in the strawberry plant, each miR159-generating locus responded differently to GA treatment.
miRNAs have a great impact on the biological processes of plants, mainly during development and the stress response (Jones-Rhoades et al., 2006). However, information about their function during fruit growth and ripening is limited. Although there have been reports on their function in several species (Pilcher et al., 2007; Carra et al., 2009), this is not the case in strawberries. We report that two strawberry genes, Fa-MIR159a and Fa-MIR159b, display tissue- and developmental stage-specific expression patterns during fruit ripening, with their highest expression being identified in the receptacle vs the achenes. However, analysis of the mature Fa-miR159 in the receptacle revealed that the levels of Fa-miR159b were, at every stage, much higher than those of Fa-miR159a. In addition, the lowest levels of Fa-miR159b were observed during the W stage, differing from the lowest expression of Fa-MIR159b, which was observed in the G receptacle. Similar discrepancies between pri-/pre-miRNA and mature miRNA levels have been well documented in both plants and animals (Kim et al., 2009; Voinnet, 2009; Correa et al., 2010). Three members of the miR159 family have been identified in Arabidopsis (Allen et al., 2007), and At-miR159a and At-miR159b are likely to act in a redundant manner (Allen et al., 2007). In the strawberry, there are no data to assess the possible redundancy of these two miR159 family members. Our expression studies indicate that the down-regulation of mature Fa-miR159 may contribute to the regulation of receptacle development. The receptacle is not the true fruit but, rather, is the result of the engrossment of the flower receptacle. The major phenotypic events in the transition from the G to W stage are marked by a dramatic increase in cell size and a loss of photosynthesizing capacity (Perkins-Veazie, 1995). These critical changes are the result of a programme that is accompanied by both dramatic gene expression and metabolic changes (Aharoni et al., 2002). This transition coincides with a significant decrease in mature Fa-miR159.
Clearly, miRNAs play important roles during plant development through their regulation of the expression levels of their target genes. Indeed, miR159 targets have been functionally analysed in both Arabidopsis and rice. In Arabidopsis, seven GAMYB-like genes contain potential miR159 binding sites (Rhoades et al., 2002), but only MYB33 and MYB65 were deregulated in the mir159a miR159b double mutant plants (Millar & Gubler, 2005; Allen et al., 2007). In rice, three genes encoding GAMYB-like transcription factors contained a miR159 complementary sequence (Jones-Rhoades & Bartel, 2004; Tsuji et al., 2006). Interestingly, a new miR159 target gene, which is unrelated to MYB, was identified in the tomato (Buxdorf et al., 2010). In strawberry, we provide evidence that a GAMYB-like gene, Fa-GAMYB, is the target of the mature Fa-miR159. It is important to note that the search for Fa-GAMYB homologues in F. vesca (Shulaev et al., 2011), the wild diploid relative of the cultivated strawberry (F. × ananassa), which has been proposed as a model for the genus (Shulaev et al., 2011), identified only one putative homologue.
The GAMYB and GAMYB-like genes encode a highly conserved family of R2R3 MYB domain transcription factors (Gubler et al., 1999). GAMYB was initially identified in the cereal aleurone (Gubler et al., 1995) and is strongly expressed in the anthers, especially in the tapetum (Murray et al., 2003; Aya et al., 2009). In this tissue, miR159 is co-expressed with GAMYB and finely regulates the levels of this transcription factor (Tsuji et al., 2006). It has been demonstrated that the interplay of miR159 and its target, GAMYB, is involved in the regulation of vegetative growth, flowering time, anther development, seed shape and germination (Millar & Gubler, 2005; Allen et al., 2007; Reyes & Chua, 2007). In the strawberry receptacle, Fa-MIR159a and Fa-MIR159b are co-expressed with Fa-GAMYB, and their opposite expression patterns indicate a possible role in the regulation of the levels of the transcription factor Fa-GAMYB. However, only the levels of the mature miR159b clearly correlate with the Fa-GAMYB expression pattern in the receptacle, suggesting that miR159a might have a more prominent role in other strawberry tissues. Despite many studies of strawberry fruit development, only a small number of regulatory genes relevant for this process have been identified (Aharoni et al., 2001).
The GAMYB-like genes have been implicated in GA signal transduction (Gubler et al., 1995; Murray et al., 2003; Aya et al., 2009). In Arabidopsis, both MYB33 and MYB65 have been implicated in the GA signalling pathway which regulates flowering during SD conditions (Gocal et al., 2001; Achard et al., 2004). We recently determined a critical role for GA in the development of the strawberry receptacle by demonstrating that, during the transition from the G to W stage, there is a dramatic peak of GA4 (Csukasi et al., 2011). We also reported that, during the change from the G to W receptacle, the expression of both FaGID1 and FaRGA is correlated with the activation of GA signalling (Csukasi et al., 2011). Recent studies on GAMYB’s regulation of anther development in rice have proposed that GA control occurs via GID1/DELLA (Aya et al., 2009). In this study, we report that the GA increase is coincident with the maximal expression of Fa-GAMYB; however, a functional relationship between GA and Fa-GAMYB production needs to be formally demonstrated. Our results depict a model in which mature Fa-miR159 may arise from two precursors, specifically pre-Fa-miR159a and pre-Fa-miR159b, which display different regulatory mechanisms. Although the expression of Fa-MIR159a is GA dependent, Fa-MIR159b expression is post-transcriptionally regulated. Different post-transcriptional regulations for the miR159 family members have been reported previously (Li et al., 2011). During receptacle development, this post-transcriptional regulation constitutes an alternative, GA-independent mechanism to regulate Fa-GAMYB expression mediated by Fa-miR159b. In Arabidopsis, the expression of miR159 is regulated by GA (Achard et al., 2004), whereas, in rice, miR159 expression has been shown to be GA independent (Tsuji et al., 2006). However, analyses of the effect of GA in specific members of the miR159 family have not been performed in these species. The occurrence of both GA-dependent and GA-independent control of mature Fa-miR159 is an example of specialization within the miR159 gene family in the strawberry. Their sequential operation during receptacle development secures a minimum for mature Fa-miR159 during the W stage.
This work was supported by BIO2010-15630 (V.V.) and BIO2009-12004 (C.L.) (Ministerio de Ciencia e Innovacion (MICINN), Spain). F.C. was granted a FPI fellowship from MICINN (Spain). The authors thank Dr Anthony A. Millar for providing Arabidopsis thaliana mir159a, b double-mutant seeds.