Plant Mol Biol (2014) 86:681–689 DOI 10.1007/s11103-014-0258-x
Methylation of the Sf locus in almond is associated with S-RNase loss of function Angel Fernández i Martí · Thomas M. Gradziel · Rafel Socias i Company
Received: 25 July 2014 / Accepted: 10 October 2014 / Published online: 18 October 2014 © Springer Science+Business Media Dordrecht 2014
Abstract Self-compatibility in almond (Prunus dulcis) is attributed to the presence of the Sf haplotype, allelic to and dominant over the series of S-alleles controlling selfincompatibility. Some forms of the Sf haplotype, however, are phenotypically self-incompatible even though their nucleotide sequences are identical. DNA from leaves and styles from genetically diverse almond samples was cloned and sequenced and then analyzed for changes affecting Sf-RNase variants. Epigenetic changes in several cytosine residues were detected in a fragment of 4,700 bp of the 5′ upstream region of all self-compatible samples of the SfRNases, differentiating them from all self-incompatible samples of Sf-RNases analyzed. This is the first report of DNA methylation in a Rosaceae species and appears to be strongly associated with inactivation of the Sf allele. Results facilitate an understanding of the evolution of selfcompatibility/self-incompatibility in almond and other Prunus species, and suggest novel approaches for future crop improvement.
A. Fernández i Martí (*) · R. Socias i Company Unidad de Hortofruticultura, Centro de Investigación y Tecnología Agroalimentaria de Aragón (CITA), Av. Montañana 930, 50059 Saragossa, Spain e-mail: [email protected] R. Socias i Company e-mail: [email protected] Present Address: A. Fernández i Martí Genome Center, University of California, Davis, CA 956161, USA T. M. Gradziel Department of Plant Science, University of California, Davis, CA 95616, USA e-mail: [email protected]
Keywords Almond · Self-compatibility · DNA methylation · Epigenetics · Upstream region · S-RNase
Introduction In almond [Prunus dulcis (Mill.) D.A.Webb syn. Prunus amygdalus (L.) Batsch], as in other Prunus species, selfincompatibility (SI) is of the gametophytic (GSI) type. Control is by a single polymorphic locus containing at least two linked genes; one specifically expressed in the pistil and the other in the pollen (Kao and Tsukamoto 2004). Pollen tube growth is arrested in the style whenever the S haplotype expressed in the haploid pollen matches either of the two S haplotypes expressed in the diploid pistil tissue (de Nettancourt 1977). The pistil component of SI in Rosaceae, Solanaceae and Plantaginaceae is a polymorphic glycoprotein with ribonuclease activity called S-RNase (McClure et al. 1989; Tao et al. 1997). The candidate gene for the pollen component in almond has been identified to be an SFB (S haplotype-specific F-box protein) by Ushijima et al. (2003), showing a tight linkage with the S-RNase gene (Ikeda et al. 2005). Most almond cultivars are self-incompatible (Socias i Company 1990). The few exceptions that have been identified as self-compatible (SC) have been shown to carry a mutation at the S-locus referred to as Sf. Since SC is expressed in heterozygous SfSx cultivars, the Sf haplotype in pistils is considered to be dominant over that of functional S alleles (Socias i Company 1984) since self-fertilization is achieved. Fertilization will only be by Sf pollen as the Sx pollen will still be recognized and arrested by the pistil SxRNase. Self-pollination of SfSx genotypes will thus produce only two genotypes, SfSx and SfSf, in the progeny with no homozygous SxSx progeny (Alonso and Socias i Company 2005).
13
682
Bošković et al. (1997) showed that in almond the S alleles code for stylar proteins with RNase activity whereas Sf does not. The characterization of the Prunus T2-type RNase (Igic and Kohn 2001) with five conserved domains (C1, C2, C3, RC4 and C5) and one hyper-variable region (Sassa et al. 1997; Ushijima et al. 1998) has allowed different S alleles to be sequenced. A comparison of the different amino acid sequences has revealed a higher degree of diversity among S alleles in almond relative to other Prunus species. Amplification of the S alleles by PCR from genomic DNA using conserved and specific primers has been widely used for S genotyping and determination of cross-incompatibility groups (López et al. 2006; Kodad et al. 2008). Despite the improved knowledge concerning the genetic structure of the female and male determinants of SI, the nature of their interaction remains unclear. In addition, genes and factors outside the S-locus have been shown to play important roles in the SI system in the Rosaceae, as well as the related Solanaceae and Plantaginaceae (Goldraij et al. 2006; Kodad et al. 2009; Tsukamoto et al. 2003; Vilanova et al. 2006). Fernández i Martí et al. (2009) reported that in almond the breakdown of SI in the Sf haplotype appears to occur in the pistil. The observed S-genotype distribution for seedlings from the cross ‘Vivot’ × ‘Blanquerna’ differed significantly from the expected distribution, suggesting a modifier effects in the pistil gene (S-RNase) but not in the pollen gene (SFB). A similar model had previously been proposed by Tsukamoto et al. (2003) for a population of Petunia axillaris where breakdown of SI was due to modifier locus suppression of the S-RNases for the S13 allele. Pollen tube growth and fruit set studies, as well as NEpHGE analysis of S-RNase activity has recently shown that the Sf phenotype in almond can also be strongly affected by genetic background, ranging from self-incompatible to fully self-compatible (Kodad et al. 2009 and 2010; Fernández i Martí et al. 2010; Socias i Company et al. 2014). Quantitative real-time PCR analysis has revealed transcripts of Sf-RNase in the style tissue of a SI cultivar (‘Vivot’, S23Sf) as well as the absence of the Sf-RNase transcripts in a SC cultivar (‘Blanquerna’, S8Sf). Because of this differential transcription, the Sf-RNase allele conferring SI has been designated Sfa (active) (Kodad et al.2009), whereas the same Sf-RNase allele conferring SC has been designated Sfi (inactive) (Fernández i Martí et al. 2009). Previous sequencing confirmed that Sfa and Sfi were identical at the nucleotide level in the coding region (C1–C5) and in their 5′-flanking regions (Fernández i Martí et al. 2010). Similarly, the TATA-Box and IB-like motifs did not reveal any nucleotide difference in the cis-element in the active relative to inactive Sf types (Fernández i Martí et al. 2010).
13
Plant Mol Biol (2014) 86:681–689
Differing phenotypic expression for the same S-genotype has been observed in other Prunus species, including Japanese plum (P. salicina Lindl.) (Watari et al. 2007; Guerra et al. 2009) and sweet cherry (P. avium L.) (Wünsch et al. 2010). The ‘Ambrunés’, ‘Duroni 3′ and ‘Satonishiki’ sweet cherry cultivars with the S3S6 genotype are SI, whereas ‘Cristobalina’ with the same S3S6 genotype is SC (Wünsch and Hormaza 2004). In Japanese plum, ‘Black Diamond’, ‘Black Gold’, ‘Black Star’, and ‘Earliqueen’ possessing the Se allele are SI (Guerra 2011) while in other cultivars the Se allele is associated with SC. The absence of any nucleotide sequence difference in the Sf-RNases of the SC (Sfi) and SI (Sfa) phenotypes, suggests that either modifier genes strongly affect expression or that epigenetic changes are involved. Epigenetic changes have been shown responsible for patterns of gene silencing in eukaryotic cells without alterations in the DNA sequence (Jeanish and Bird 2003; Suzuki and Bird 2008). Thus, the same gene sequence can be either well-expressed or transcriptionally silent (Bender 2004). Since Manning et al. (2006) first demonstrated that epigenetic variants in tomato were due to spontaneous yet heritable changes in the methylation of controlling genes, DNA methylation has become the most extensively studied epigenetic mechanism. Methylation of cytosine nucleotides, usually within CG dinucleotides, is the most common form of covalent DNA modification in eukaryotes (Bird 2002). The high amount of 5-methylcytosine found in some plant species has suggested that methylation is not restricted to the CG sequence context and has led to the discovery that cytosine is also methylated in the CNG group, where N is any nucleotide (Gruenbaum et al. 1981). Methylation of cytosine residues in promoter regions generally represses the expression of the gene (Saze 2008). However, cytosine methylation modifications have only been reported to date in a few tree species (Bitonti et al. 2002; Fraga et al. 2002), with no reports in almond. The objective of this study was to determine if DNA methylation in the sequence of the Sf-RNase allele, either in the 5′ flanking region containing the TATA box, the IBlike Motif or the coding region of the Sf-RNase, was associated with the differential phenotypic expression of SC/SI observed in almond cultivars.
Materials and methods Plant material Six almond genotypes, containing both homozygotes and heterozygotes for the Sf haplotype, were included for analysis (Table 1). Genotypes were selected based on previous genetic and phenotypic characterization of Sf established by pollen tube growth and field pollination studies (Socias i Company
Plant Mol Biol (2014) 86:681–689
683
Table 1 Almond genotypes included in the methylation analysis Genotype
S-genotype Phenotype Origin
Blanquerna S8 Sf Vivot S23 Sf
SC
Genco × AS-1
SI
Local of Majorca, Spain
Soleta
S23 Sf
SC
Blanquerna × Belle d’Aurons
Ponç
S23 Sf
SI
Local of Majorca, Spain
A2-199
Sf Sf
SC
M-2-16
Sf Sf
SC
C-1322 [Tuono × Genco] ⊗
M-2-2 [Tuono × Ferragnès] ⊗
SC self-compatible, SI self-incompatible
and Felipe 1999 and 2007; Fernández i Martí et al. 2009; Kodad et al. 2009). They consist of two traditional cultivars from the island of Majorca, Spain (‘Vivot’ and ‘Ponç’), two releases from the CITA breeding program (‘Blanquerna’ and ‘Soleta’), and seedlings from breeding programs of CITA (M-2-16) and CEBAS-CSIC at Murcia, Spain (A-2-199). The Sf genes analyzed could be traced back to two different origins: the Italian region of Puglia for the standard self-compatible form exemplified by the heirloom cultivar ‘Tuono’, and the island of Majorca where Sf types showing self-incompatible phenotypes have been identified (Kodad et al. 2008; Fernández i Martí et al. 2010). Leaf samples of two cultivars, ‘Blanquerna’ and ‘Vivot’, were analyzed from two different locations (Zaragoza and Majorca) and during two different seasons (spring and summer) to assess the effect that environmental differences might have on methylation patterns. DNA methylation in gene regulation Total DNA was extracted from leaves and styles at stage D (Felipe 1977) using the CTAB method following the procedure of Doyle and Doyle (1987). After DNA extraction, the samples were further purified with phenol/chloroform in order to reduce the amount of proteins and polysaccharides. DNA samples were then quantified by NANODROP (Thermo Scientific, Wilmington, USA) and by gel electrophoresis with a λ DNA marker (Takara, Otsu, Japan). Once DNA was extracted, it was submitted to the DNA bisulphite modification treatment ‘MethylEasy’, according
Fig. 1 Nucleotide sequence using the primer pair CIPABF4/ CIPABR4 of ‘Blanquerna’ showing that the sequence from pistil DNA with bisulphite treatment is the same as that of genomic DNA from pistil without bisulphite treatment with the single exception of
to the manufacturer (Human Genetic Signatures, Brisbane, Australia) protocol. This kit has been designed to efficiently convert cytosine to uracil and to reduce DNA degradation and loss without decreasing the conversion of C to U residues. The converted DNA was then used for PCR amplification, cloning and sequencing. For PCR amplification, specific primers covering a maximum length of 250 bp available in our Sf-RNase upstream region were designed in this work (Fig. 1). Primers were designed to cover short sequences since amplification with short fragments provides more precise methylation patterns. In total, 15 different primer pairs were designed, all located in a range of 4,700 bp of the 5′ flanking region of the Sf-RNase. Primer design took into account the conversion of cytosine nucleotides to uracil after bisulphite treatment and transcribed to thymine after PCR amplification. In addition to the upstream region, the first and second intron of the Sf-RNase were also evaluated to determine if nucleotide methylation was present. The thermal cycling program was as follows: an initial denaturation at 95 °C for 1 min followed by 35 cycles of 94 °C for 30 s, annealing for 30 s of 53 °C and extension at 72 °C for 1 min, ending with a 5 min extension at 72 °C. The PCR products were cloned using the pGEM-T-Easy Vector System (Promega, Madison, WI, USA). Cloning and sequencing of the whole region of 4,700 bp in the 5′ upstream were initially performed for ‘Blanquerna’ and ‘Vivot’ and methylation changes over the whole sequence in both cultivars were carefully analyzed. Subsequently, only those samples showing a methylation change (Table 2) were used for cloning and sequencing the remaining genotypes. Five clones for each pair and genotype were additionally purified using ExoSAP-IT (USB, Cleveland, OH, USA) and sequenced using the DTCS Quick Start kit and CEQ 8000 Genetic Analysis System (Beckman Coulter, Fullerton, CA, USA). Expression analysis by real time PCR (qRT-PCR) of the stylar product Total RNA from leaves and styles of ‘Blanquerna’, ‘Vivot’, ‘Ponç’ and ‘Soleta’ was extracted following the cold
the conversion of cytosines to thymines (black arrows). The large (red) arrow shows where the methylation of the cytosine residue appeared
13
684 Table 2 Sequences of the specific primers which detected methylation patterns in the test cultivars ‘Blanquerna’ and ‘Vivot’ These primers were specifically designed for bisulphite analysis and are located in the upstream region of the almond Sf-RNase a CIPAB CITA and Prunus Amygdalus Bastch
Plant Mol Biol (2014) 86:681–689 Primera
Sequence for methylation 5′ → 3′
CIPAB-F1 CIPAB-R1 CIPAB-F2 CIPAB-R2 CIPAB-F3 CIPAB-R3 CIPAB-F4
ATTAATTATTTTATTATGATTTTATGGTTATTAATATA ACACAAACAAAAATAAAAAAAATAAATCATATTTTACTA GAAGAATTAGGAAGTATGATTATTGTTAAGTATTGAAATG ATTACTATATCAAACAATTTCTTAAAAAAATTTCTATC GTAATGAATTTTGAAATGTTGAAATAAATAATTAGAGGTT AAAAATAAAATCTAACCACACCAAACAAAATCACATTA TTTAAATAATATTAAGATAAATTTGTAAATAAAGG
CIPAB-R4
CCAAAACAATAAAACCTAACTCTTAATAATTAAAAAAT
phenol SDS method as described by Tao et al. (1999). One microgram of total RNA treated with DNase I (Invitrogen, Tokyo, Japan) was used for first strand cDNA synthesis by SuperScriptIII RT (Invitrogen, Tokyo, Japan). Quantitative real time PCR was performed with a LightCycler® 480 real-time PCR system (Roche). The PCR mixture conditions were the same as that described in Fernández i Martí et al. (2010). Three reactions were performed for each cultivar. All reactions were compared with a SyBr Green Supermix (Roche) using the housekeeping gene Actin as an internal control. Actin transcripts were quantified using two actin-specific primers designed by Watari et al. (2007) for real-time RT-PCR analysis, qpActF (5′ CAACTGGTATTGTGCTGGATTCTG 3′) and qpActR (5′ CAGCAAGGTCAAGACGAAGGAT 3′). Final concentration was 0.5 µM for the Actin primer and 0.9 µM for the SfRNase primers. Primers and probes for Sf-RNase were specifically designed by Fernández i Martí et al. (2010). Melt curve analysis was performed for SYBR® green reaction after each run to determine the specificity of the amplification of Actin. The PCR reaction was performed following the same conditions as in Fernández i Martí et al. (2010).
Results and discussion DNA methylation and sequence characterization Fifteen specific primers were designed in the Sf-RNase upstream region in order to cover a size of 4,700 bp. In this way, only short fragments for PCR amplification were obtained. From all possible combinations of primer pairs, only four gave methylated changes in the amplified fragment sequences in the test cultivars ‘Blanquerna’ and ‘Vivot’ (Table 2), and thus were selected for cloning and sequencing in the remaining genotypes. Cloning and sequencing of the PCR products obtained confirmed that the bisulphite treatment worked properly in all DNA tested. Accordingly, the treatment was able to eliminate all C bases that were not present as 5mCs, replacing the C bases with T bases in the resulting DNA. Thus, the treatment created
13
a nearly C-less sequence of mostly 3-base-DNA having predominantly A, G and T. In the partial sequence shown in Fig. 1, the presence of several cytosine residues can be observed in the upstream region of genomic DNA of ‘Blanquerna’ obtained from pistil before the bisulphite treatment. However, after the treatment, sequencing of all clones from ‘Blanquerna’ showed that almost all cytosines were converted to thymines with a single exception (red arrow). After confirming the efficacy of the treatment, every nucleotide was carefully compared between the two sequences for a length of 4,700 bp in the upstream region, as well as in the first and second intron of the Sf-RNase. Among all sequences obtained, no changes were produced in the introns of the Sf-RNase; however four cytosine residues were not able to be converted to thymine in the upstream region of the SC cultivars indicating the presence of methylation. Figure 2 shows the diagram of the whole sequence of the unconverted Sf upstream region where the four methylation points have been identified, as well as the four primer pairs overlapping these methylated cytosines. Five clones were obtained for each of the six genotypes analyzed (four SC and two SI) and for each portion of sequence where the methylated fragments were found. These epi-mutations were always located in the same position in all the corresponding clones. The length of the region flanked by each primer designed in this work was approximately 250 bp. The first methylated nucleotide was located 4,275 bp upstream to the start codon in all samples from leaf (Fig. 3) as well as pistil tissue (Fig. 4). This methylated region was stable in all five clones obtained from the SC cultivar ‘Blanquerna’. The other three cytosine epimutations were located 3,200, 2,400 and 650 bp upstream of the start codon. Once the presence of several cytosine nucleotides in the sequence of the SC cultivar ‘Blanquerna’ was confirmed, the upstream region of ‘Soleta’, another SC cultivar, was analyzed. All clones from ‘Soleta’ revealed the same sequence as ‘Blanquerna’, with the cytosine changes in the same positions when using the same primer sets. In all cases, these methylated cytosines were detected in the CG and CNG forms, where the N nucleotide was adenine. The
Plant Mol Biol (2014) 86:681–689
685
Fig. 2 Diagram showing the detection of methylated sites in the 5′ untranslational region of Sf-RNase. Primer positions overlapping the cytosine residues are indicated by black arrows. Methylated cytosines, either detected in the CG or the CNG forms, are indicated in red
Fig. 3 Sequences surrounding the four (a–d) methylated points in the six genotypes studied. For each genotype and region five clones were sequenced. All samples were from leaves collected in Zaragoza in the spring
first and second methylated changes were detected in the CG form, whereas the third and fourth methylated changes were detected in the CAG form. In leaf DNA methyl occupancy was A(M1) = 17/20, B(M2) = 18/20, C(M3) = 16/20, and D(M4) = 15/20, whereas in ‘Blanquerna’ methyl occupancy was of 18/20, in ‘Soleta’ of 14/20, in A2-199 of 19/20, and in M2-16 of 15/20, considering the four sites together (Fig. 3). In pistil DNA methyl occupancy was A(M1) = 15/20, B(M2) = 14/20, C(M3) = 15/20, and D(M4) = 15/20; whereas in ‘Blanquerna’ was of 16/20, in ‘Soleta’ of 14/20, in A2-199 of 13/20, and in M2-16 of 16/20, considering the four sites together (Fig. 4). Analysis of the DNA samples of the SI cultivars ‘Vivot’ and ‘Ponç’ showed that all cytosine residues were converted to thymine in all the clones obtained (Fig. 3). However, as these cultivars possess a second haplotype, either S8 or S23, the conversion of the cytosine residues in the Sf haplotype could have been influenced by this second haplotype. To
test this possibility, genomic clones constructed from a fosmid library for the S8 and S23 haplotypes (Fernández i Martí et al. 2010) were used for PCR amplification with the same primer combinations as those designed for Sf. After PCR amplification, no band was detected in either S8 or S23, thus showing that the primers used were specific to the Sf-RNase. To confirm these results, the 5′ upstream region of two homozygous Sf selections were cloned and sequenced. As expected, the genotypes sequenced showed that the four cytosine residues were located in the same position as for ‘Blanquerna’ and ‘Soleta’ in all clones obtained from each selection. These results confirmed the stability of the presence of the cytosine residues in homozygous cultivars possessing only the Sf haplotype. In addition, when the experiment was done in different tissues, identical sequences and methyl-cytosine positions were obtained from both pistils and leaves. Leaf DNA retained the same methylation levels as the pistil DNA (Figs. 3 and 4), indicating that the
13
686
Plant Mol Biol (2014) 86:681–689
Fig. 4 Sequences surrounding the four (a–d) methylated points in the six genotypes studied. For each genotype and region five clones were sequenced. All samples were from pistils collected in Zaragoza in the spring
epi-mutation is most likely inherited through mitosis rather than being a tissue specific methylation event (Calarco et al. 2012). It is significant that this phenomenon was observed in ‘Blanquerna’ and ‘Soleta’, whereas the two SfSf homozygotes (A2-199 and M2-16), showed different DNA methylation patterns in the four sites examined. This different behavior might be due to the origin of these homozygotes, as none is a progeny from ‘Blanquerna’ or ‘Soleta’. Although the epi-mutations are supposed to be inheritable through generations, the manner and frequency of this transmission are unknown (for review, see Niederhuth and Schmitz 2014). Consequently, the epigenetic modifications in the SfSf plants could have not been inherited in all the clones selected and sequenced. Long-term studies would be required to test this possibility. In addition, samples of ‘Blanquerna’ and ‘Vivot’ collected from two different locations (Zaragoza and Majorca) and during two different seasons (spring and summer) showed no sequence differences (Fig. 5). Consequently, cytosine methylation had not been influenced by external factors such as location or season. These results of the bisulphite treatment in samples of different locations and seasons show the stability of this modification in the plant material studied. To verify results previously obtained by Fernández i Martí et al. (2010), a RT-qPCR analysis was performed for four genotypes. As expected, transcripts of the SfaRNase gene were present in the style tissue of ‘Vivot’ and ‘Ponç’, whereas qRT-PCR showed the absence of Sfi-RNase
13
transcripts in the style of SC cultivars (Fig. 6). In addition, no transcripts in either variant of Sf-RNase were detected when using cDNA from leaves. Thus, the detected expression of the Sf-RNase gene in the pistil tissue of SI genotypes indicates that this gene is probably functional but non-functional in SC types resulting in an absence of RNase. Self-compatibility and gene silencing The differential expression of the two Sf-RNases was initially attributed to a possible mutation (Kodad et al. 2009) and the present results suggest that this apparent mutation could be an epigenetic change. Recent advances in our understanding of the role of methylation show it to be an important contributor to epigenetic variability (Bird 2002). DNA methylation has frequently been reported to occur in differentiated tissues and in specific organs (Shiba et al. 2006; Gehring and Henikoff 2007). In Brassica rapa, DNA was extracted from different tissues, such as leaves, anther tapetum, pollen and petals, and methylation levels were compared in the region of the SP11 gene responsible for pollen specificity in sporophytic SI. Methylation was found to be specific for the anther tapetum and not for other tissues (Shiba et al. 2006). In the present study, DNA was first extracted from the pistils were Sf is normally expressed. However, when DNA from leaf tissue, which is not involved in SI/SC expression was tested, the level of methylation was retained. Thus, Sf methylation in almond may be considered as genotype-specific and not tissue-specific.
Plant Mol Biol (2014) 86:681–689
687
Fig. 5 Sequences surrounding the four methylated points in the two leaf samples collected in summer in Majorca. For each genotype and region five clones were sequenced
Fig. 6 Quantitative Sf-RNase expression using real-time PCR. Relative transcript level of the pistil tissue of the Sf (‘Vivot’ and ‘Ponç’) and Sf (‘Blanquerna’ and ‘Soleta’) genes are compared to the housekeeping gene Actin
Bisulphite sequencing allows the determination of the methylation status of every cytosine residue. This information is particularly relevant for studies of methylation in developmental processes. Using this sequencing method, it was possible to establish exact methylation patterns of the Sf-RNase gene from very small amounts of genomic DNA. Cytosine methylation has been associated with gene silencing, and genes with 5-methylcytosine in their promoter region are usually transcriptionally silent (Jones and Takai 2001). Because DNA methylation does not change the DNA sequence but only the expression level, Bender
(2004) suggested that the detection of DNA methylation could be an effective approach for determining epigenetic changes when an absence of normal RNase transcripts is detected within specific tissue. Our results support this proposal since the change of DNA methylation was only found in the SC genotypes. The affected Sf allele also lacked RNase activity, demonstrating that its expression is inhibited. Transcripts of Sf-RNase were present in the style tissue of the SI genotypes, whereas qRT-PCR showed the absence of Sf-RNase transcripts in the style of the SC genotypes, confirming previous results of Fernández i Martí et al. (2010). This inhibition also supports the importance of DNA methylation to SI in almond, suggesting that DNA methylation is involved in the normal activation/inactivation of the Sf-RNase. Thus, when the Sf-RNase sequence is methylated, its expression is inhibited resulting in an SC phenotype, as occurs in ‘Blanquerna’, ‘Soleta’ and the two homozygous SC selections. When Sf-RNase sequences do not show methylated cytosines, the RNase remains active, resulting in a SI genotype and phenotype, as in ‘Vivot’ and ‘Ponç’. The original Sf haplotype would have been in an active form, thus conferring SI, as with other S-haplotypes. Epigenetic changes through DNA methylation resulted in loss of RNase function and so a novel SC phenotype in a in an otherwise SI species such as almond. While the Sf alleles examined were derived from two distinct Mediterranean regions, they may be identical by descent from the original pool of Mediterranean
13
688
germplasm. The epigenetic change appears to have taken place in the Italian region of Puglia, where most SC cultivars have been described (Socias i Company 1990). Additionally, all SC almond genotypes identified so far have shown an identical Sf-RNase (Fernández i Martí et al. 2010) and are of Mediterranean origin and so probably related to the Puglia almond population (Socias i Company et al. 2012). The molecular nature of the S-locus has been widely studied in almond using different approaches to better characterize the basis of SC/SI (López et al. 2006; Fernández i Martí et al. 2010; Kodad et al. 2010). The present results suggest that even if the S-locus confers S-specificity, epigenetic changes in the 5′-flanking region of the Sf-RNase can significantly alter its expression. Thus, these findings suggest that the methylated 5′ region of the Sf-RNase may be a crucial site for modifying S-RNase expression in the style. DNA methylation has previously been proposed as a regulatory factor to suppress gene expression (Jones and Takai 2001), suggesting that these changes are not ‘anomalies’ but rather different aspects of the same underlying system. Such regulatory factors may have evolved as independent mechanism to control Sf expression. Epigenetic modifications may thus provide novel opportunities for crop improvement, particularly in clonally propagated crops such as almond where methylations changes associated with seed propagation are avoided. Acknowledgments This research was funded by the Spanish grant AGL2010-22197-C02-01 and the Research Group A12 of Aragón. A. Fernández i Martí acknowledges a scholarship co-funded by the Spanish ‘Ministerio de Educación y Ciencia’ and the European Social Fund (FSE), under projects AGL 2004-06674-C02-01 and BES-200612621. Special thanks are due to Drs. J.M. Alonso (CITA) and F. Dicenta (CEBAS-CSIC) for providing plant samples, and to Drs. R. Tao, T. Akagi and T. Hanada (Univ. of Kyoto, Japan) for their collaboration in designing the experiment.
References Alonso JM, Socias i Company R (2005) Differential pollen tube growth in inbred self-compatible almond genotypes. Euphytica 144:207–213 Bender J (2004) DNA Methylation and epigenetics. Annu Rev Plant Biol 55:41–68 Bird AP (2002) DNA methylation patterns and epigenetic memory. Genes Dev 16:6–21 Bitonti MB, Cozza R, Chiappeta A, Gianino D, Ruffini Castiglioni A, Dewitte W, Mariotti D, Van Onckelen H, Innocenti A (2002) Distinct nuclear organization, DNA methylation pattern and cytokinin distribution mark juvenile, juvenile-like and adult vegetative apical meristems in peach (Prunus persica (L.) Batsch). J Exp Bot 53:1047–1053 Bošković R, Tobutt KR, Batlle I, Duval H (1997) Correlation of ribonuclease zymograms and incompatibility genotypes in almond. Euphytica 97:167–176
13
Plant Mol Biol (2014) 86:681–689 Calarco JP, Borges F, Donoghue MT, Van Ex F, Jullien PE, Lopes T, Gardner R, Berger F, Feijo JA, Becker JD, Martienssen R (2012) Reprogramming of DNA methylation in pollen guides epigenetic inheritance via small RNA. Cell 151:194–205 de Nettancourt D (1977) Incompatibility in angiosperms. Springer, Berlin Doyle JJ, Doyle JL (1987) A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochem Bull 19:11–15 Felipe AJ (1977) Almendro: estados fenológicos. Inf Técn Econ Agrar 27:8–9 Fernández i Martí A, Hanada T, Alonso JM, Yamane H, Tao R, Socias i Company R (2009) A modifier locus affecting the expression of the S-RNase gene could be the cause of breakdown of selfincompatibility in almond. Sex Plant Reprod 22:179–186 Fernández i Martí A, Hanada T, Alonso JM, Yamane H, Tao R, Socias i Company R (2010) The almond Sf haplotype shows a double expression despite its comprehensive genetic identity. Sci Hort 125:685–691 Fraga M, Cañal MJ, Rodriguez R (2002) Phase-change related epigenetic and physiological changes in Pinus radiata D. Don. Planta 215:672–678 Gehring M, Henikoff S (2007) DNA methylation dynamics in plant genomes. Biochim Biophys Acta 1769:276–286 Goldraij A, Kondo K, Lee CB, Hancock CN, Sivaguru M, VazquezSantana S, Kim S, Phillips TE, Cruz-Garcia F, McClure B (2006) Compartmentalization of S-RNase and HT-B degradation in selfincompatible Nicotiana. Nature 439:805–810 Gruenbaum Y, Naveh-Many T, Cedar H, Razin A (1981) Sequence specificity of methylation in higher plant DNA. Nature 292:860–862 Guerra ME (2011) Polinización y cuajado en ciruelo japonés. PhD thesis. Universidad de Extremadura Guerra ME, Rodrigo J, López-Corrales M, Wünsch A (2009) S-RNase genotyping and incompatibility group assignment by PCR and pollination experiments in Japanese plum. Plant Breed 128:304–311 Igic B, Kohn JR (2001) Evolutionary relationships among self-incompatibility RNases. Proc Natl Acad Sci USA 98:13167–13171 Ikeda K, Ushijima K, Yamane H, Tao R, Hauck NR, Sebolt AM, Iezzoni AF (2005) Linkage and physical distances between the S-haplotype S-RNase and SFB genes in sweet cherry. Sex Plant Reprod 17:289–296 Jeanish R, Bird A (2003) Epigenetic regulation of gene expression: how the genome integrates intrinsic and environmental signals. Nature Genet 33:S245–S254 Jones PA, Takai D (2001) The role of DNA methylation in mammalian epigenetics. Science 293:1068–1070 Kao TH, Tsukamoto T (2004) The molecular and genetic bases of S-RNase based self-incompatibility. Plant Cell 16:S72–S83 Kodad O, Alonso JM, Sánchez A, Oliveira MM, Socias i Company R (2008) Evaluation of genetic diversity of S-alleles in an almond germplasm collection. J Hort Sci Biotechnol 83:603–608 Kodad O, Socias i Company R, Sánchez A, Oliveira MM (2009) The expression of self-compatibility in almond may not only be due to the presence of the Sf allele. J Amer Soc Hort Sci 134:221–227 Kodad O, Alonso JM, Fernández i Martí A, Oliveira MM, Socias i Company R (2010) Molecular and physiological identification of new S-alleles associated with self-(in)compatibility in local Spanish almond cultivars. Sci Hort 123:308–311 López M, Vargas FJ, Batlle I (2006) Self-(in) compatibility almond genotypes: a review. Euphytica 150:1–16 Manning K, Mahmut T, Poole M, Hong Y, Thompson J, King G, Giovannoni J, Seymour B (2006) A naturally occurring epigenetic mutation in a gene encoding an SPB-box transcription factor inhibits tomato fruit ripening. Nature Genet 38(948):952
Plant Mol Biol (2014) 86:681–689 McClure BA, Haring V, Ebert PR, Anderson MA, Simpson RJ, Sakiyama F, Clarke AE (1989) Style self incompatibility gene products of Nicotiana alata are ribonucleases. Nature 342:955–957 Niederhuth C, Schmitz R (2014) Covering your bases: inheritance of DNA methylation in plant genomes. Mol Plant 7:472–480 Sassa H, Hirano H, Nishio T, Koba T (1997) Style-specific selfincompatibility mutation caused by deletion of the S-RNase gene in Japanese pear (Pyrus serotina). Plant J 12:223–227 Saze H (2008) Epigenetic memory transmission through mitosis and meiosis in plants. Semin Cell Dev Biol 19:527–536 Shiba H, Kakizaki T, Iwano M, Tarutani Y, Watanabe M, Isogai A, Takayama S (2006) Dominance relationships between selfincompatibility alleles controlled by DNA methylation. Nature Genet 38:297–299 Socias i Company R (1984) A genetic approach to the transmission of self-compatibility in almond. Options Méditerr CIHEAM/IAMZ 84/II:123–128 Socias i Company R (1990) Breeding self-compatible almonds. Plant Breed Rev 8:313–338 Socias i Company R, Felipe AJ (1999) ‘Blanquerna’, ‘Cambra’ y ‘Felisia’: tres nuevos cultivares autógamos de almendro. Inf Técn Econ Agrar 95V:111–117 Socias i Company R, Felipe AJ (2007) ‘Belona’ and ‘Soleta’ almonds. HortScience 42:704–706 Socias i Company R, Alonso JM, Kodad O, Gradziel TM (2012) Almond [Prunus dulcis syn. P. amygdalus]. In: Badenes ML, Byrne D (eds) Fruit breeding. Springer, Heidelberg, pp 697–728 Socias i Company R, Kodad O, Fernández i Martí A, Alonso JM (2014) Self-compatibility expression in almond by pollen tube growth. Acta Hort 1028:111–116 Suzuki M, Bird A (2008) DNA methylation landscapes: provocative insights from epigenomics. Nature 9:465–475 Tao R, Yamane H, Sassa H, Mori H, Gradziel TM, Dandekar AM, Sugiura A (1997) Identification of stylar RNases associated with gametophytic self-incompatibility in almond (Prunus dulcis). Plant Cell Physiol 38:304–311
689 Tao R, Yamane H, Sugiura A, Murayama H, Sassa H, Mori H (1999) Molecular typing of S alleles trough identification, characterization and cDNA cloning for s-RNases in sweet cherry. J Amer Soc Hort Sci 124:224–233 Tsukamoto T, Ando T, Kobukun H, Watanabe H, Sato T, Masada M, Marchesi E, Kao TH (2003) Breakdown of self-incompatibility in a natural population of Petunia axillaris caused by a modifier locus that suppresses the expression of an S-RNase gene. Sex Plant Reprod 15:255–263 Ushijima K, Sasss H, Tao R, Yamane H, Dankenkar AM, Gradziel TM, Hirano H (1998) Cloning and characterization of cDNAs encoding S-RNases from almond (Prunus dulcis): primary structural features and sequence diversity of the S-RNases in Rosaceae. Mol Gen Genet 260:261–268 Ushijima K, Sassa H, Dandekar AM, Gradziel TM, Tao R, Hirano H (2003) Structural and transcriptional analysis of the self-incompatibility locus of almond: identification of a pollen-expressed F-box gene with haplotype-specific polymorphism. Plant Cell 15:771–781 Vilanova S, Badenes ML, Burgos L, Martínez-Calvo J, Llácer G, Romero C (2006) Self-compatibility of two apricot selections is associated with two pollen part mutations of different nature. Plant Physiol 142:629–641 Watari A, Hanada T, Yamane H, Esumi T, Tao R, Yaegaki H, Yamaguchi M, Beppu B, Kataoka I (2007) A low transcriptional level of Se-RNase in the Se-haplotype confers self-compatibility in Japanese plum. J Amer Soc Hort Sci 132:396–406 Wünsch A, Hormaza JI (2004) Genetic and molecular analysis in Cristobalina sweet cherry, a spontaneous self-compatible mutant. Sex Plant Reprod 17:203–210 Wünsch A, Tao R, Hormaza JI (2010) Self-compatibility in ‘Cristobalina’ sweet cherry is not associated with duplications or modified transcription levels of S-locus genes. Plant Cell Rep 29:715–721
13
Methylation of the Sf locus in almond is associated with S-RNase loss of function Angel Fernández i Martí · Thomas M. Gradziel · Rafel Socias i Company
Received: 25 July 2014 / Accepted: 10 October 2014 / Published online: 18 October 2014 © Springer Science+Business Media Dordrecht 2014
Abstract Self-compatibility in almond (Prunus dulcis) is attributed to the presence of the Sf haplotype, allelic to and dominant over the series of S-alleles controlling selfincompatibility. Some forms of the Sf haplotype, however, are phenotypically self-incompatible even though their nucleotide sequences are identical. DNA from leaves and styles from genetically diverse almond samples was cloned and sequenced and then analyzed for changes affecting Sf-RNase variants. Epigenetic changes in several cytosine residues were detected in a fragment of 4,700 bp of the 5′ upstream region of all self-compatible samples of the SfRNases, differentiating them from all self-incompatible samples of Sf-RNases analyzed. This is the first report of DNA methylation in a Rosaceae species and appears to be strongly associated with inactivation of the Sf allele. Results facilitate an understanding of the evolution of selfcompatibility/self-incompatibility in almond and other Prunus species, and suggest novel approaches for future crop improvement.
A. Fernández i Martí (*) · R. Socias i Company Unidad de Hortofruticultura, Centro de Investigación y Tecnología Agroalimentaria de Aragón (CITA), Av. Montañana 930, 50059 Saragossa, Spain e-mail: [email protected] R. Socias i Company e-mail: [email protected] Present Address: A. Fernández i Martí Genome Center, University of California, Davis, CA 956161, USA T. M. Gradziel Department of Plant Science, University of California, Davis, CA 95616, USA e-mail: [email protected]
Keywords Almond · Self-compatibility · DNA methylation · Epigenetics · Upstream region · S-RNase
Introduction In almond [Prunus dulcis (Mill.) D.A.Webb syn. Prunus amygdalus (L.) Batsch], as in other Prunus species, selfincompatibility (SI) is of the gametophytic (GSI) type. Control is by a single polymorphic locus containing at least two linked genes; one specifically expressed in the pistil and the other in the pollen (Kao and Tsukamoto 2004). Pollen tube growth is arrested in the style whenever the S haplotype expressed in the haploid pollen matches either of the two S haplotypes expressed in the diploid pistil tissue (de Nettancourt 1977). The pistil component of SI in Rosaceae, Solanaceae and Plantaginaceae is a polymorphic glycoprotein with ribonuclease activity called S-RNase (McClure et al. 1989; Tao et al. 1997). The candidate gene for the pollen component in almond has been identified to be an SFB (S haplotype-specific F-box protein) by Ushijima et al. (2003), showing a tight linkage with the S-RNase gene (Ikeda et al. 2005). Most almond cultivars are self-incompatible (Socias i Company 1990). The few exceptions that have been identified as self-compatible (SC) have been shown to carry a mutation at the S-locus referred to as Sf. Since SC is expressed in heterozygous SfSx cultivars, the Sf haplotype in pistils is considered to be dominant over that of functional S alleles (Socias i Company 1984) since self-fertilization is achieved. Fertilization will only be by Sf pollen as the Sx pollen will still be recognized and arrested by the pistil SxRNase. Self-pollination of SfSx genotypes will thus produce only two genotypes, SfSx and SfSf, in the progeny with no homozygous SxSx progeny (Alonso and Socias i Company 2005).
13
682
Bošković et al. (1997) showed that in almond the S alleles code for stylar proteins with RNase activity whereas Sf does not. The characterization of the Prunus T2-type RNase (Igic and Kohn 2001) with five conserved domains (C1, C2, C3, RC4 and C5) and one hyper-variable region (Sassa et al. 1997; Ushijima et al. 1998) has allowed different S alleles to be sequenced. A comparison of the different amino acid sequences has revealed a higher degree of diversity among S alleles in almond relative to other Prunus species. Amplification of the S alleles by PCR from genomic DNA using conserved and specific primers has been widely used for S genotyping and determination of cross-incompatibility groups (López et al. 2006; Kodad et al. 2008). Despite the improved knowledge concerning the genetic structure of the female and male determinants of SI, the nature of their interaction remains unclear. In addition, genes and factors outside the S-locus have been shown to play important roles in the SI system in the Rosaceae, as well as the related Solanaceae and Plantaginaceae (Goldraij et al. 2006; Kodad et al. 2009; Tsukamoto et al. 2003; Vilanova et al. 2006). Fernández i Martí et al. (2009) reported that in almond the breakdown of SI in the Sf haplotype appears to occur in the pistil. The observed S-genotype distribution for seedlings from the cross ‘Vivot’ × ‘Blanquerna’ differed significantly from the expected distribution, suggesting a modifier effects in the pistil gene (S-RNase) but not in the pollen gene (SFB). A similar model had previously been proposed by Tsukamoto et al. (2003) for a population of Petunia axillaris where breakdown of SI was due to modifier locus suppression of the S-RNases for the S13 allele. Pollen tube growth and fruit set studies, as well as NEpHGE analysis of S-RNase activity has recently shown that the Sf phenotype in almond can also be strongly affected by genetic background, ranging from self-incompatible to fully self-compatible (Kodad et al. 2009 and 2010; Fernández i Martí et al. 2010; Socias i Company et al. 2014). Quantitative real-time PCR analysis has revealed transcripts of Sf-RNase in the style tissue of a SI cultivar (‘Vivot’, S23Sf) as well as the absence of the Sf-RNase transcripts in a SC cultivar (‘Blanquerna’, S8Sf). Because of this differential transcription, the Sf-RNase allele conferring SI has been designated Sfa (active) (Kodad et al.2009), whereas the same Sf-RNase allele conferring SC has been designated Sfi (inactive) (Fernández i Martí et al. 2009). Previous sequencing confirmed that Sfa and Sfi were identical at the nucleotide level in the coding region (C1–C5) and in their 5′-flanking regions (Fernández i Martí et al. 2010). Similarly, the TATA-Box and IB-like motifs did not reveal any nucleotide difference in the cis-element in the active relative to inactive Sf types (Fernández i Martí et al. 2010).
13
Plant Mol Biol (2014) 86:681–689
Differing phenotypic expression for the same S-genotype has been observed in other Prunus species, including Japanese plum (P. salicina Lindl.) (Watari et al. 2007; Guerra et al. 2009) and sweet cherry (P. avium L.) (Wünsch et al. 2010). The ‘Ambrunés’, ‘Duroni 3′ and ‘Satonishiki’ sweet cherry cultivars with the S3S6 genotype are SI, whereas ‘Cristobalina’ with the same S3S6 genotype is SC (Wünsch and Hormaza 2004). In Japanese plum, ‘Black Diamond’, ‘Black Gold’, ‘Black Star’, and ‘Earliqueen’ possessing the Se allele are SI (Guerra 2011) while in other cultivars the Se allele is associated with SC. The absence of any nucleotide sequence difference in the Sf-RNases of the SC (Sfi) and SI (Sfa) phenotypes, suggests that either modifier genes strongly affect expression or that epigenetic changes are involved. Epigenetic changes have been shown responsible for patterns of gene silencing in eukaryotic cells without alterations in the DNA sequence (Jeanish and Bird 2003; Suzuki and Bird 2008). Thus, the same gene sequence can be either well-expressed or transcriptionally silent (Bender 2004). Since Manning et al. (2006) first demonstrated that epigenetic variants in tomato were due to spontaneous yet heritable changes in the methylation of controlling genes, DNA methylation has become the most extensively studied epigenetic mechanism. Methylation of cytosine nucleotides, usually within CG dinucleotides, is the most common form of covalent DNA modification in eukaryotes (Bird 2002). The high amount of 5-methylcytosine found in some plant species has suggested that methylation is not restricted to the CG sequence context and has led to the discovery that cytosine is also methylated in the CNG group, where N is any nucleotide (Gruenbaum et al. 1981). Methylation of cytosine residues in promoter regions generally represses the expression of the gene (Saze 2008). However, cytosine methylation modifications have only been reported to date in a few tree species (Bitonti et al. 2002; Fraga et al. 2002), with no reports in almond. The objective of this study was to determine if DNA methylation in the sequence of the Sf-RNase allele, either in the 5′ flanking region containing the TATA box, the IBlike Motif or the coding region of the Sf-RNase, was associated with the differential phenotypic expression of SC/SI observed in almond cultivars.
Materials and methods Plant material Six almond genotypes, containing both homozygotes and heterozygotes for the Sf haplotype, were included for analysis (Table 1). Genotypes were selected based on previous genetic and phenotypic characterization of Sf established by pollen tube growth and field pollination studies (Socias i Company
Plant Mol Biol (2014) 86:681–689
683
Table 1 Almond genotypes included in the methylation analysis Genotype
S-genotype Phenotype Origin
Blanquerna S8 Sf Vivot S23 Sf
SC
Genco × AS-1
SI
Local of Majorca, Spain
Soleta
S23 Sf
SC
Blanquerna × Belle d’Aurons
Ponç
S23 Sf
SI
Local of Majorca, Spain
A2-199
Sf Sf
SC
M-2-16
Sf Sf
SC
C-1322 [Tuono × Genco] ⊗
M-2-2 [Tuono × Ferragnès] ⊗
SC self-compatible, SI self-incompatible
and Felipe 1999 and 2007; Fernández i Martí et al. 2009; Kodad et al. 2009). They consist of two traditional cultivars from the island of Majorca, Spain (‘Vivot’ and ‘Ponç’), two releases from the CITA breeding program (‘Blanquerna’ and ‘Soleta’), and seedlings from breeding programs of CITA (M-2-16) and CEBAS-CSIC at Murcia, Spain (A-2-199). The Sf genes analyzed could be traced back to two different origins: the Italian region of Puglia for the standard self-compatible form exemplified by the heirloom cultivar ‘Tuono’, and the island of Majorca where Sf types showing self-incompatible phenotypes have been identified (Kodad et al. 2008; Fernández i Martí et al. 2010). Leaf samples of two cultivars, ‘Blanquerna’ and ‘Vivot’, were analyzed from two different locations (Zaragoza and Majorca) and during two different seasons (spring and summer) to assess the effect that environmental differences might have on methylation patterns. DNA methylation in gene regulation Total DNA was extracted from leaves and styles at stage D (Felipe 1977) using the CTAB method following the procedure of Doyle and Doyle (1987). After DNA extraction, the samples were further purified with phenol/chloroform in order to reduce the amount of proteins and polysaccharides. DNA samples were then quantified by NANODROP (Thermo Scientific, Wilmington, USA) and by gel electrophoresis with a λ DNA marker (Takara, Otsu, Japan). Once DNA was extracted, it was submitted to the DNA bisulphite modification treatment ‘MethylEasy’, according
Fig. 1 Nucleotide sequence using the primer pair CIPABF4/ CIPABR4 of ‘Blanquerna’ showing that the sequence from pistil DNA with bisulphite treatment is the same as that of genomic DNA from pistil without bisulphite treatment with the single exception of
to the manufacturer (Human Genetic Signatures, Brisbane, Australia) protocol. This kit has been designed to efficiently convert cytosine to uracil and to reduce DNA degradation and loss without decreasing the conversion of C to U residues. The converted DNA was then used for PCR amplification, cloning and sequencing. For PCR amplification, specific primers covering a maximum length of 250 bp available in our Sf-RNase upstream region were designed in this work (Fig. 1). Primers were designed to cover short sequences since amplification with short fragments provides more precise methylation patterns. In total, 15 different primer pairs were designed, all located in a range of 4,700 bp of the 5′ flanking region of the Sf-RNase. Primer design took into account the conversion of cytosine nucleotides to uracil after bisulphite treatment and transcribed to thymine after PCR amplification. In addition to the upstream region, the first and second intron of the Sf-RNase were also evaluated to determine if nucleotide methylation was present. The thermal cycling program was as follows: an initial denaturation at 95 °C for 1 min followed by 35 cycles of 94 °C for 30 s, annealing for 30 s of 53 °C and extension at 72 °C for 1 min, ending with a 5 min extension at 72 °C. The PCR products were cloned using the pGEM-T-Easy Vector System (Promega, Madison, WI, USA). Cloning and sequencing of the whole region of 4,700 bp in the 5′ upstream were initially performed for ‘Blanquerna’ and ‘Vivot’ and methylation changes over the whole sequence in both cultivars were carefully analyzed. Subsequently, only those samples showing a methylation change (Table 2) were used for cloning and sequencing the remaining genotypes. Five clones for each pair and genotype were additionally purified using ExoSAP-IT (USB, Cleveland, OH, USA) and sequenced using the DTCS Quick Start kit and CEQ 8000 Genetic Analysis System (Beckman Coulter, Fullerton, CA, USA). Expression analysis by real time PCR (qRT-PCR) of the stylar product Total RNA from leaves and styles of ‘Blanquerna’, ‘Vivot’, ‘Ponç’ and ‘Soleta’ was extracted following the cold
the conversion of cytosines to thymines (black arrows). The large (red) arrow shows where the methylation of the cytosine residue appeared
13
684 Table 2 Sequences of the specific primers which detected methylation patterns in the test cultivars ‘Blanquerna’ and ‘Vivot’ These primers were specifically designed for bisulphite analysis and are located in the upstream region of the almond Sf-RNase a CIPAB CITA and Prunus Amygdalus Bastch
Plant Mol Biol (2014) 86:681–689 Primera
Sequence for methylation 5′ → 3′
CIPAB-F1 CIPAB-R1 CIPAB-F2 CIPAB-R2 CIPAB-F3 CIPAB-R3 CIPAB-F4
ATTAATTATTTTATTATGATTTTATGGTTATTAATATA ACACAAACAAAAATAAAAAAAATAAATCATATTTTACTA GAAGAATTAGGAAGTATGATTATTGTTAAGTATTGAAATG ATTACTATATCAAACAATTTCTTAAAAAAATTTCTATC GTAATGAATTTTGAAATGTTGAAATAAATAATTAGAGGTT AAAAATAAAATCTAACCACACCAAACAAAATCACATTA TTTAAATAATATTAAGATAAATTTGTAAATAAAGG
CIPAB-R4
CCAAAACAATAAAACCTAACTCTTAATAATTAAAAAAT
phenol SDS method as described by Tao et al. (1999). One microgram of total RNA treated with DNase I (Invitrogen, Tokyo, Japan) was used for first strand cDNA synthesis by SuperScriptIII RT (Invitrogen, Tokyo, Japan). Quantitative real time PCR was performed with a LightCycler® 480 real-time PCR system (Roche). The PCR mixture conditions were the same as that described in Fernández i Martí et al. (2010). Three reactions were performed for each cultivar. All reactions were compared with a SyBr Green Supermix (Roche) using the housekeeping gene Actin as an internal control. Actin transcripts were quantified using two actin-specific primers designed by Watari et al. (2007) for real-time RT-PCR analysis, qpActF (5′ CAACTGGTATTGTGCTGGATTCTG 3′) and qpActR (5′ CAGCAAGGTCAAGACGAAGGAT 3′). Final concentration was 0.5 µM for the Actin primer and 0.9 µM for the SfRNase primers. Primers and probes for Sf-RNase were specifically designed by Fernández i Martí et al. (2010). Melt curve analysis was performed for SYBR® green reaction after each run to determine the specificity of the amplification of Actin. The PCR reaction was performed following the same conditions as in Fernández i Martí et al. (2010).
Results and discussion DNA methylation and sequence characterization Fifteen specific primers were designed in the Sf-RNase upstream region in order to cover a size of 4,700 bp. In this way, only short fragments for PCR amplification were obtained. From all possible combinations of primer pairs, only four gave methylated changes in the amplified fragment sequences in the test cultivars ‘Blanquerna’ and ‘Vivot’ (Table 2), and thus were selected for cloning and sequencing in the remaining genotypes. Cloning and sequencing of the PCR products obtained confirmed that the bisulphite treatment worked properly in all DNA tested. Accordingly, the treatment was able to eliminate all C bases that were not present as 5mCs, replacing the C bases with T bases in the resulting DNA. Thus, the treatment created
13
a nearly C-less sequence of mostly 3-base-DNA having predominantly A, G and T. In the partial sequence shown in Fig. 1, the presence of several cytosine residues can be observed in the upstream region of genomic DNA of ‘Blanquerna’ obtained from pistil before the bisulphite treatment. However, after the treatment, sequencing of all clones from ‘Blanquerna’ showed that almost all cytosines were converted to thymines with a single exception (red arrow). After confirming the efficacy of the treatment, every nucleotide was carefully compared between the two sequences for a length of 4,700 bp in the upstream region, as well as in the first and second intron of the Sf-RNase. Among all sequences obtained, no changes were produced in the introns of the Sf-RNase; however four cytosine residues were not able to be converted to thymine in the upstream region of the SC cultivars indicating the presence of methylation. Figure 2 shows the diagram of the whole sequence of the unconverted Sf upstream region where the four methylation points have been identified, as well as the four primer pairs overlapping these methylated cytosines. Five clones were obtained for each of the six genotypes analyzed (four SC and two SI) and for each portion of sequence where the methylated fragments were found. These epi-mutations were always located in the same position in all the corresponding clones. The length of the region flanked by each primer designed in this work was approximately 250 bp. The first methylated nucleotide was located 4,275 bp upstream to the start codon in all samples from leaf (Fig. 3) as well as pistil tissue (Fig. 4). This methylated region was stable in all five clones obtained from the SC cultivar ‘Blanquerna’. The other three cytosine epimutations were located 3,200, 2,400 and 650 bp upstream of the start codon. Once the presence of several cytosine nucleotides in the sequence of the SC cultivar ‘Blanquerna’ was confirmed, the upstream region of ‘Soleta’, another SC cultivar, was analyzed. All clones from ‘Soleta’ revealed the same sequence as ‘Blanquerna’, with the cytosine changes in the same positions when using the same primer sets. In all cases, these methylated cytosines were detected in the CG and CNG forms, where the N nucleotide was adenine. The
Plant Mol Biol (2014) 86:681–689
685
Fig. 2 Diagram showing the detection of methylated sites in the 5′ untranslational region of Sf-RNase. Primer positions overlapping the cytosine residues are indicated by black arrows. Methylated cytosines, either detected in the CG or the CNG forms, are indicated in red
Fig. 3 Sequences surrounding the four (a–d) methylated points in the six genotypes studied. For each genotype and region five clones were sequenced. All samples were from leaves collected in Zaragoza in the spring
first and second methylated changes were detected in the CG form, whereas the third and fourth methylated changes were detected in the CAG form. In leaf DNA methyl occupancy was A(M1) = 17/20, B(M2) = 18/20, C(M3) = 16/20, and D(M4) = 15/20, whereas in ‘Blanquerna’ methyl occupancy was of 18/20, in ‘Soleta’ of 14/20, in A2-199 of 19/20, and in M2-16 of 15/20, considering the four sites together (Fig. 3). In pistil DNA methyl occupancy was A(M1) = 15/20, B(M2) = 14/20, C(M3) = 15/20, and D(M4) = 15/20; whereas in ‘Blanquerna’ was of 16/20, in ‘Soleta’ of 14/20, in A2-199 of 13/20, and in M2-16 of 16/20, considering the four sites together (Fig. 4). Analysis of the DNA samples of the SI cultivars ‘Vivot’ and ‘Ponç’ showed that all cytosine residues were converted to thymine in all the clones obtained (Fig. 3). However, as these cultivars possess a second haplotype, either S8 or S23, the conversion of the cytosine residues in the Sf haplotype could have been influenced by this second haplotype. To
test this possibility, genomic clones constructed from a fosmid library for the S8 and S23 haplotypes (Fernández i Martí et al. 2010) were used for PCR amplification with the same primer combinations as those designed for Sf. After PCR amplification, no band was detected in either S8 or S23, thus showing that the primers used were specific to the Sf-RNase. To confirm these results, the 5′ upstream region of two homozygous Sf selections were cloned and sequenced. As expected, the genotypes sequenced showed that the four cytosine residues were located in the same position as for ‘Blanquerna’ and ‘Soleta’ in all clones obtained from each selection. These results confirmed the stability of the presence of the cytosine residues in homozygous cultivars possessing only the Sf haplotype. In addition, when the experiment was done in different tissues, identical sequences and methyl-cytosine positions were obtained from both pistils and leaves. Leaf DNA retained the same methylation levels as the pistil DNA (Figs. 3 and 4), indicating that the
13
686
Plant Mol Biol (2014) 86:681–689
Fig. 4 Sequences surrounding the four (a–d) methylated points in the six genotypes studied. For each genotype and region five clones were sequenced. All samples were from pistils collected in Zaragoza in the spring
epi-mutation is most likely inherited through mitosis rather than being a tissue specific methylation event (Calarco et al. 2012). It is significant that this phenomenon was observed in ‘Blanquerna’ and ‘Soleta’, whereas the two SfSf homozygotes (A2-199 and M2-16), showed different DNA methylation patterns in the four sites examined. This different behavior might be due to the origin of these homozygotes, as none is a progeny from ‘Blanquerna’ or ‘Soleta’. Although the epi-mutations are supposed to be inheritable through generations, the manner and frequency of this transmission are unknown (for review, see Niederhuth and Schmitz 2014). Consequently, the epigenetic modifications in the SfSf plants could have not been inherited in all the clones selected and sequenced. Long-term studies would be required to test this possibility. In addition, samples of ‘Blanquerna’ and ‘Vivot’ collected from two different locations (Zaragoza and Majorca) and during two different seasons (spring and summer) showed no sequence differences (Fig. 5). Consequently, cytosine methylation had not been influenced by external factors such as location or season. These results of the bisulphite treatment in samples of different locations and seasons show the stability of this modification in the plant material studied. To verify results previously obtained by Fernández i Martí et al. (2010), a RT-qPCR analysis was performed for four genotypes. As expected, transcripts of the SfaRNase gene were present in the style tissue of ‘Vivot’ and ‘Ponç’, whereas qRT-PCR showed the absence of Sfi-RNase
13
transcripts in the style of SC cultivars (Fig. 6). In addition, no transcripts in either variant of Sf-RNase were detected when using cDNA from leaves. Thus, the detected expression of the Sf-RNase gene in the pistil tissue of SI genotypes indicates that this gene is probably functional but non-functional in SC types resulting in an absence of RNase. Self-compatibility and gene silencing The differential expression of the two Sf-RNases was initially attributed to a possible mutation (Kodad et al. 2009) and the present results suggest that this apparent mutation could be an epigenetic change. Recent advances in our understanding of the role of methylation show it to be an important contributor to epigenetic variability (Bird 2002). DNA methylation has frequently been reported to occur in differentiated tissues and in specific organs (Shiba et al. 2006; Gehring and Henikoff 2007). In Brassica rapa, DNA was extracted from different tissues, such as leaves, anther tapetum, pollen and petals, and methylation levels were compared in the region of the SP11 gene responsible for pollen specificity in sporophytic SI. Methylation was found to be specific for the anther tapetum and not for other tissues (Shiba et al. 2006). In the present study, DNA was first extracted from the pistils were Sf is normally expressed. However, when DNA from leaf tissue, which is not involved in SI/SC expression was tested, the level of methylation was retained. Thus, Sf methylation in almond may be considered as genotype-specific and not tissue-specific.
Plant Mol Biol (2014) 86:681–689
687
Fig. 5 Sequences surrounding the four methylated points in the two leaf samples collected in summer in Majorca. For each genotype and region five clones were sequenced
Fig. 6 Quantitative Sf-RNase expression using real-time PCR. Relative transcript level of the pistil tissue of the Sf (‘Vivot’ and ‘Ponç’) and Sf (‘Blanquerna’ and ‘Soleta’) genes are compared to the housekeeping gene Actin
Bisulphite sequencing allows the determination of the methylation status of every cytosine residue. This information is particularly relevant for studies of methylation in developmental processes. Using this sequencing method, it was possible to establish exact methylation patterns of the Sf-RNase gene from very small amounts of genomic DNA. Cytosine methylation has been associated with gene silencing, and genes with 5-methylcytosine in their promoter region are usually transcriptionally silent (Jones and Takai 2001). Because DNA methylation does not change the DNA sequence but only the expression level, Bender
(2004) suggested that the detection of DNA methylation could be an effective approach for determining epigenetic changes when an absence of normal RNase transcripts is detected within specific tissue. Our results support this proposal since the change of DNA methylation was only found in the SC genotypes. The affected Sf allele also lacked RNase activity, demonstrating that its expression is inhibited. Transcripts of Sf-RNase were present in the style tissue of the SI genotypes, whereas qRT-PCR showed the absence of Sf-RNase transcripts in the style of the SC genotypes, confirming previous results of Fernández i Martí et al. (2010). This inhibition also supports the importance of DNA methylation to SI in almond, suggesting that DNA methylation is involved in the normal activation/inactivation of the Sf-RNase. Thus, when the Sf-RNase sequence is methylated, its expression is inhibited resulting in an SC phenotype, as occurs in ‘Blanquerna’, ‘Soleta’ and the two homozygous SC selections. When Sf-RNase sequences do not show methylated cytosines, the RNase remains active, resulting in a SI genotype and phenotype, as in ‘Vivot’ and ‘Ponç’. The original Sf haplotype would have been in an active form, thus conferring SI, as with other S-haplotypes. Epigenetic changes through DNA methylation resulted in loss of RNase function and so a novel SC phenotype in a in an otherwise SI species such as almond. While the Sf alleles examined were derived from two distinct Mediterranean regions, they may be identical by descent from the original pool of Mediterranean
13
688
germplasm. The epigenetic change appears to have taken place in the Italian region of Puglia, where most SC cultivars have been described (Socias i Company 1990). Additionally, all SC almond genotypes identified so far have shown an identical Sf-RNase (Fernández i Martí et al. 2010) and are of Mediterranean origin and so probably related to the Puglia almond population (Socias i Company et al. 2012). The molecular nature of the S-locus has been widely studied in almond using different approaches to better characterize the basis of SC/SI (López et al. 2006; Fernández i Martí et al. 2010; Kodad et al. 2010). The present results suggest that even if the S-locus confers S-specificity, epigenetic changes in the 5′-flanking region of the Sf-RNase can significantly alter its expression. Thus, these findings suggest that the methylated 5′ region of the Sf-RNase may be a crucial site for modifying S-RNase expression in the style. DNA methylation has previously been proposed as a regulatory factor to suppress gene expression (Jones and Takai 2001), suggesting that these changes are not ‘anomalies’ but rather different aspects of the same underlying system. Such regulatory factors may have evolved as independent mechanism to control Sf expression. Epigenetic modifications may thus provide novel opportunities for crop improvement, particularly in clonally propagated crops such as almond where methylations changes associated with seed propagation are avoided. Acknowledgments This research was funded by the Spanish grant AGL2010-22197-C02-01 and the Research Group A12 of Aragón. A. Fernández i Martí acknowledges a scholarship co-funded by the Spanish ‘Ministerio de Educación y Ciencia’ and the European Social Fund (FSE), under projects AGL 2004-06674-C02-01 and BES-200612621. Special thanks are due to Drs. J.M. Alonso (CITA) and F. Dicenta (CEBAS-CSIC) for providing plant samples, and to Drs. R. Tao, T. Akagi and T. Hanada (Univ. of Kyoto, Japan) for their collaboration in designing the experiment.
References Alonso JM, Socias i Company R (2005) Differential pollen tube growth in inbred self-compatible almond genotypes. Euphytica 144:207–213 Bender J (2004) DNA Methylation and epigenetics. Annu Rev Plant Biol 55:41–68 Bird AP (2002) DNA methylation patterns and epigenetic memory. Genes Dev 16:6–21 Bitonti MB, Cozza R, Chiappeta A, Gianino D, Ruffini Castiglioni A, Dewitte W, Mariotti D, Van Onckelen H, Innocenti A (2002) Distinct nuclear organization, DNA methylation pattern and cytokinin distribution mark juvenile, juvenile-like and adult vegetative apical meristems in peach (Prunus persica (L.) Batsch). J Exp Bot 53:1047–1053 Bošković R, Tobutt KR, Batlle I, Duval H (1997) Correlation of ribonuclease zymograms and incompatibility genotypes in almond. Euphytica 97:167–176
13
Plant Mol Biol (2014) 86:681–689 Calarco JP, Borges F, Donoghue MT, Van Ex F, Jullien PE, Lopes T, Gardner R, Berger F, Feijo JA, Becker JD, Martienssen R (2012) Reprogramming of DNA methylation in pollen guides epigenetic inheritance via small RNA. Cell 151:194–205 de Nettancourt D (1977) Incompatibility in angiosperms. Springer, Berlin Doyle JJ, Doyle JL (1987) A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochem Bull 19:11–15 Felipe AJ (1977) Almendro: estados fenológicos. Inf Técn Econ Agrar 27:8–9 Fernández i Martí A, Hanada T, Alonso JM, Yamane H, Tao R, Socias i Company R (2009) A modifier locus affecting the expression of the S-RNase gene could be the cause of breakdown of selfincompatibility in almond. Sex Plant Reprod 22:179–186 Fernández i Martí A, Hanada T, Alonso JM, Yamane H, Tao R, Socias i Company R (2010) The almond Sf haplotype shows a double expression despite its comprehensive genetic identity. Sci Hort 125:685–691 Fraga M, Cañal MJ, Rodriguez R (2002) Phase-change related epigenetic and physiological changes in Pinus radiata D. Don. Planta 215:672–678 Gehring M, Henikoff S (2007) DNA methylation dynamics in plant genomes. Biochim Biophys Acta 1769:276–286 Goldraij A, Kondo K, Lee CB, Hancock CN, Sivaguru M, VazquezSantana S, Kim S, Phillips TE, Cruz-Garcia F, McClure B (2006) Compartmentalization of S-RNase and HT-B degradation in selfincompatible Nicotiana. Nature 439:805–810 Gruenbaum Y, Naveh-Many T, Cedar H, Razin A (1981) Sequence specificity of methylation in higher plant DNA. Nature 292:860–862 Guerra ME (2011) Polinización y cuajado en ciruelo japonés. PhD thesis. Universidad de Extremadura Guerra ME, Rodrigo J, López-Corrales M, Wünsch A (2009) S-RNase genotyping and incompatibility group assignment by PCR and pollination experiments in Japanese plum. Plant Breed 128:304–311 Igic B, Kohn JR (2001) Evolutionary relationships among self-incompatibility RNases. Proc Natl Acad Sci USA 98:13167–13171 Ikeda K, Ushijima K, Yamane H, Tao R, Hauck NR, Sebolt AM, Iezzoni AF (2005) Linkage and physical distances between the S-haplotype S-RNase and SFB genes in sweet cherry. Sex Plant Reprod 17:289–296 Jeanish R, Bird A (2003) Epigenetic regulation of gene expression: how the genome integrates intrinsic and environmental signals. Nature Genet 33:S245–S254 Jones PA, Takai D (2001) The role of DNA methylation in mammalian epigenetics. Science 293:1068–1070 Kao TH, Tsukamoto T (2004) The molecular and genetic bases of S-RNase based self-incompatibility. Plant Cell 16:S72–S83 Kodad O, Alonso JM, Sánchez A, Oliveira MM, Socias i Company R (2008) Evaluation of genetic diversity of S-alleles in an almond germplasm collection. J Hort Sci Biotechnol 83:603–608 Kodad O, Socias i Company R, Sánchez A, Oliveira MM (2009) The expression of self-compatibility in almond may not only be due to the presence of the Sf allele. J Amer Soc Hort Sci 134:221–227 Kodad O, Alonso JM, Fernández i Martí A, Oliveira MM, Socias i Company R (2010) Molecular and physiological identification of new S-alleles associated with self-(in)compatibility in local Spanish almond cultivars. Sci Hort 123:308–311 López M, Vargas FJ, Batlle I (2006) Self-(in) compatibility almond genotypes: a review. Euphytica 150:1–16 Manning K, Mahmut T, Poole M, Hong Y, Thompson J, King G, Giovannoni J, Seymour B (2006) A naturally occurring epigenetic mutation in a gene encoding an SPB-box transcription factor inhibits tomato fruit ripening. Nature Genet 38(948):952
Plant Mol Biol (2014) 86:681–689 McClure BA, Haring V, Ebert PR, Anderson MA, Simpson RJ, Sakiyama F, Clarke AE (1989) Style self incompatibility gene products of Nicotiana alata are ribonucleases. Nature 342:955–957 Niederhuth C, Schmitz R (2014) Covering your bases: inheritance of DNA methylation in plant genomes. Mol Plant 7:472–480 Sassa H, Hirano H, Nishio T, Koba T (1997) Style-specific selfincompatibility mutation caused by deletion of the S-RNase gene in Japanese pear (Pyrus serotina). Plant J 12:223–227 Saze H (2008) Epigenetic memory transmission through mitosis and meiosis in plants. Semin Cell Dev Biol 19:527–536 Shiba H, Kakizaki T, Iwano M, Tarutani Y, Watanabe M, Isogai A, Takayama S (2006) Dominance relationships between selfincompatibility alleles controlled by DNA methylation. Nature Genet 38:297–299 Socias i Company R (1984) A genetic approach to the transmission of self-compatibility in almond. Options Méditerr CIHEAM/IAMZ 84/II:123–128 Socias i Company R (1990) Breeding self-compatible almonds. Plant Breed Rev 8:313–338 Socias i Company R, Felipe AJ (1999) ‘Blanquerna’, ‘Cambra’ y ‘Felisia’: tres nuevos cultivares autógamos de almendro. Inf Técn Econ Agrar 95V:111–117 Socias i Company R, Felipe AJ (2007) ‘Belona’ and ‘Soleta’ almonds. HortScience 42:704–706 Socias i Company R, Alonso JM, Kodad O, Gradziel TM (2012) Almond [Prunus dulcis syn. P. amygdalus]. In: Badenes ML, Byrne D (eds) Fruit breeding. Springer, Heidelberg, pp 697–728 Socias i Company R, Kodad O, Fernández i Martí A, Alonso JM (2014) Self-compatibility expression in almond by pollen tube growth. Acta Hort 1028:111–116 Suzuki M, Bird A (2008) DNA methylation landscapes: provocative insights from epigenomics. Nature 9:465–475 Tao R, Yamane H, Sassa H, Mori H, Gradziel TM, Dandekar AM, Sugiura A (1997) Identification of stylar RNases associated with gametophytic self-incompatibility in almond (Prunus dulcis). Plant Cell Physiol 38:304–311
689 Tao R, Yamane H, Sugiura A, Murayama H, Sassa H, Mori H (1999) Molecular typing of S alleles trough identification, characterization and cDNA cloning for s-RNases in sweet cherry. J Amer Soc Hort Sci 124:224–233 Tsukamoto T, Ando T, Kobukun H, Watanabe H, Sato T, Masada M, Marchesi E, Kao TH (2003) Breakdown of self-incompatibility in a natural population of Petunia axillaris caused by a modifier locus that suppresses the expression of an S-RNase gene. Sex Plant Reprod 15:255–263 Ushijima K, Sasss H, Tao R, Yamane H, Dankenkar AM, Gradziel TM, Hirano H (1998) Cloning and characterization of cDNAs encoding S-RNases from almond (Prunus dulcis): primary structural features and sequence diversity of the S-RNases in Rosaceae. Mol Gen Genet 260:261–268 Ushijima K, Sassa H, Dandekar AM, Gradziel TM, Tao R, Hirano H (2003) Structural and transcriptional analysis of the self-incompatibility locus of almond: identification of a pollen-expressed F-box gene with haplotype-specific polymorphism. Plant Cell 15:771–781 Vilanova S, Badenes ML, Burgos L, Martínez-Calvo J, Llácer G, Romero C (2006) Self-compatibility of two apricot selections is associated with two pollen part mutations of different nature. Plant Physiol 142:629–641 Watari A, Hanada T, Yamane H, Esumi T, Tao R, Yaegaki H, Yamaguchi M, Beppu B, Kataoka I (2007) A low transcriptional level of Se-RNase in the Se-haplotype confers self-compatibility in Japanese plum. J Amer Soc Hort Sci 132:396–406 Wünsch A, Hormaza JI (2004) Genetic and molecular analysis in Cristobalina sweet cherry, a spontaneous self-compatible mutant. Sex Plant Reprod 17:203–210 Wünsch A, Tao R, Hormaza JI (2010) Self-compatibility in ‘Cristobalina’ sweet cherry is not associated with duplications or modified transcription levels of S-locus genes. Plant Cell Rep 29:715–721
13
