Introduction

Nectarines arose as peach mutants, and their inheritance pattern is consistent with the glabrous skin characteristic controlled by a single recessive gene (Blake 1932). Most aspects of nectarine trees, leaves, and flowers are indistinguishable from those of peach; however, peach researchers have noted differences between peaches and nectarines that extend beyond those of lack of pubescence alone. These differences include fruit size, shape, firmness, external color, aroma, flavor, and disease resistance (Wen et al. 1995a, b). The nectarine character is controlled by a recessive gene, g, which determines the glabrous character of the fruit skin, and has no close linkage with other phenotypic markers (Dirlewanger et al. 1998). The hypothesis is that nectarines are phenotypes for a minor, nonlethal mutation, which includes the gene causing pubescence of the fruit skin but includes closely linked genes that have other roles in the mesocarp of the fruit. Alternatively, it is possible that the nectarine phenotype may arise as an alteration in a single regulatory gene that controls the expression of other genes. This latter interpretation is consistent with genetic evidence and molecular characterization of the two known classes of glabrous mutations in Arabidopsis thaliana (L.) Heynh., glabrous (g11) and transparent testa glabrous (ttg), both of which are transcription factors (Maes et al. 2008; Walker et al. 1999). The organoleptic characteristics of organic acid and sugar content have been mapped to diverse linkage maps in peach and are not found associated with the glabrous G locus (Dirlewanger et al. 1999). In addition to differences in taste between peach and nectarine, the latter has been found to store better (Crisosto et al. 1999; Lurie and Crisosto 2005). This property is a valuable one for the global stone fruit industry, where shipment of fresh fruit from one continent to another is common. However, the reasons for the better storage of nectarines are not known.

One method for examining the molecular basis underlying the differences between peach and nectarine fruit is to utilize microarrays. For fruit crops, microarray platforms so far developed are mainly home-brewed, permitting robust, reproducible results to be obtained and to investigate the expression of thousands of genes at once. Two microarray platforms have been developed for peach, μPEACH1.0 and ChillPeach. The μPEACH1.0 microarray, developed by an Italian consortium, (ESTree), represents about 4,800 oligonucleotide probes corresponding to a set of unigenes expressed during the last stages of fruit development (ESTree Consortium 2005). The ChillPeach, a cDNA microarray, comprising 4,261 unigenes, was obtained from harvested, ripened, and stored fruit of two full-sibling peach progeny contrasting for chilling injury sensitivity (Ogundiwin et al. 2008). Using the two microarrays, both developed from peach fruit, increased the number of genes being investigated, since the overlap between the two platforms is only about 29 % (Granell, unpublished).

A nectarine mutant “Yuval” arose in 2002 within an “Oded” peach population from a commercial orchard in central Israel. Because the peach cultivar was of high quality, the mutation of interest was propagated. Nectarine was found to be more resistant to chilling injury (flesh browning and woolly texture) than the “Oded” peach after prolonged storage (Dagar et al. 2011). The objective of this study was to compare the molecular attributes at harvest of both fruits at the level of gene expression using transcriptomic analyses of this near-isogenic material. Thus, the comparison studied here brought us closer to elucidating the molecular basis for the pleiotropic effects of the peach-to-nectarine mutation, including the differential response to cold storage. Since it was of interest to see if gene expression at harvest could reflect the storage potential of the two fruits, we also compared the gene expression at harvest of these two fruits to that of genotypes from the peach population with different susceptibilities (sensitive and tolerant) to chilling injury, which were used to make the ChillPeach microarray. This gave us a greater spread of storage potential, since the four groups of cultivars had storage potential ranging from 2 to 7 weeks (Ogundiwin et al. 2008; Dagar et al. 2011).

Materials and methods

Plant material and treatments

The experiments were carried out with a peach [Prunus persica (L.) Batsch “Oded”] and its nectarine mutant [P. persica (L.) Batsch, var. nectarine “Yuval”] in 2008. The peach and nectarine fruits were harvested from two adjacent commercial orchards in Israel. The fruits were stored immediately at 5 °C, removed from cold storage after 3, 5, and 7 weeks, and subsequently held at 20 °C for 3 days for ripening after each removal. At the time of observation, 20 fruits from each cultivar were examined. At harvest, five representative fruits were cubed, weighed, flash frozen in liquid nitrogen, and stored at −80 °C for further analyses.

Fruit measurements

Fruit sections about 1 cm in from the peel were examined under transmitted light microscopy and photographed using a Nikon Eclipse 50i microscope and a Nikon Digital Sight DS-L1 camera system. The cell size was measured using ImageJ, a Java-based image-processing program. Ethylene production by the peach and nectarine fruits at harvest was determined by a gas chromatograph (GC; Varian 3300, Walnut Creek, CA, USA) with a flame ionization detector and an alumina column. Each fruit was individually sealed in a jar (600 mL) for 1 h at 20 °C; a 5-mL gas sample was taken with a syringe and loaded on the GC for analysis. Firmness, soluble solids content (SSC), and titratable acidity (TA) were measured following the protocol as described earlier (Zhou et al. 2000). Fruit quality including expressible juice and the flesh disorders of bleeding, browning, and wooliness after removal from storage and 3 days of shelf life at 20 °C were measured as described in Dagar et al. (2011).

Transcriptome analysis

ChillPeach microarray

ChillPeach was developed from pooling expressed sequence tags (ESTs) of peach from harvest through storage for 3 weeks. From 7,862 ESTs, 4,261 cDNA unigenes were used to construct the microarray (Ogundiwin et al. 2008). For the ChillPeach microarray analysis, total RNA from frozen mesocarp (4 g) tissue of “Oded” peach, “Yuval” nectarine, and tissues of sensitive and tolerant peach at harvest as described by Ogundiwin et al. (2008) was made using the method described by Meisel et al. (2005). The concentration and purity of the extracted RNA was assayed by ND-1000 spectrophotometer (NanoDrop Technologies Inc., USA) and its integrity was checked on agarose gels. RNA (1 μg) for microarray hybridization was amplified using the method of Van Gelder et al. (1990).

Transcriptome analyses and hybridization procedure of “Oded” peach, “Yuval” nectarine, and sensitive and tolerant peach fruit at harvest to the ChillPeach microarray slides were performed after a direct comparison design. Probe synthesis and labeling, hybridization procedure, and data analyses were carried out as described by Ogundiwin et al. (2008). To obtain differential gene expression values, three replicates of “Oded,” four replicates of “Yuval” nectarine, and three replicates of the two peach genotypes used to make the ChillPeach microarray were hybridized against the pool reference as described by Ogundiwin et al. (2008), one of which was a dye swap. To detect differentially expressed genes in treatments, data were analyzed with the Significance Analysis of Microarray (SAM) package (Tusher et al. 2001). Statistical significance was assessed using two-class (unpaired) SAM analysis for the treatments with a false discovery rate of 5 % and q value ≤0.05. The mean of the values of the differentially expressed genes was calculated for each sample as log2 values. Functional enrichment was performed using a homemade Perl script according to Tavazoie et al. (1999). Further, for each gene, the expression ratio (log2 scale) between “Yuval” nectarine and “Oded” peach was carried out. Genes were annotated by hand following the Gene Ontology (GO) categories and other databases such as NCBI (http://www.ncbi.nlm.nih.gov), Google scholar (http://scholar.google.com), EBI (http://www.ebi.ac.uk), Swiss-Prot (http://www.expasy.org/sprot), Prosite (http://www.expasy.org/prosite), BRENDA (http://www.brenda-enzymes.org), TAIR (http://www.arabidopsis.org), the Gene Index Project (http://compbio.dfci.harvard.edu/tgi/plant.html), KEEG (http://www.genome.ad.jp/kegg), PlantCyc (http://www.plantcyc.org), and Plant Transcription Factor Database 2.0 (plntfdb.bio.uni-potsdam.de/v2.0/; Riano-Pachon et al. 2007). In order to compare “Oded” peach and “Yuval” nectarine with the chilling-tolerant and chilling-sensitive peaches, one-way analysis of variance (p < 0.05) and principal component analysis (PCA) were performed utilizing replicates of all four fruit types.

μPEACH1.0 microarray

μPEACH1.0 is an oligo-based microarray constructed from ESTs expressed during different stages of ripening of climacteric peach fruit. It contains 4,806 specific 70-mer oligos (Trainotti et al. 2006). Total RNA was isolated as described above. Fifty micrograms of total RNA was treated with 10 U of RQ1 RNase-free DNAse (Promega) and 1 U of RNAguard (RNase INHIBITOR) (Amersham) for 30 min and then purified by phenol–chloroform according to the manufacturer’s instructions. The concentration and purity of the extracted RNA was assayed as above, and its integrity was checked on agarose gels. Transcriptome analyses of “Oded” peach and “Yuval” nectarine fruit were performed following a direct comparison design. Probe synthesis and labeling, hybridization procedure, and data analyses were carried out as described by Ziliotto et al. (2008) and Falara et al. (2011). Each comparison was repeated at least twice, one of which was a dye swap. Metric quality for the arrays was evaluated by the MIDAS software, included in the TM4 package (http://www.tm4.org) developed at TIGR (http://www.tigr.org) by using intensity box plots, MA plots, RI plots, and Z score histograms. To detect differentially expressed genes, data were analyzed with the SAM package (Tusher et al. 2001). Statistical significance was assessed using two-class (unpaired) SAM analysis for the comparisons with a false discovery rate of 5 %. After normalization, the ratios were transformed to their log2 values. The log2 transformation converted the expression values to an intuitive linear scale that represented twofold differences. The genes that were differentially expressed between peach and nectarine at harvest were identified. A threshold for the hybridization signal ratio, expressed as log2, was set to be higher than 1 and lower than −1 for selecting upregulated and downregulated genes, respectively. Genes were annotated as described by Bonghi et al. (2011).

Real-time quantitative reverse transcriptase-PCR analysis

The transcript abundance of selected genes that were differentially expressed between peach and its near-isogenic nectarine mutant with over twofold change and were common between the μPEACH1.0 and ChillPeach microarray results, as well as a few genes that came from the PCA across the four cultivars, were validated with quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) analyses. The expression levels for the genes were calculated relative to the Initiation Factor eIF-4-Gamma (eIF-G) gene as described by Ogundiwin et al. (2008), and the expression level of each analyzed gene transcript in the “Oded” peach was set to one and the level of the “Yuval” nectarine was calculated relative to this reference. Total RNA (1 μg) was used to synthesize full-length cDNA using the Versco™ cDNA kit (Applied Biosystems, USA). Primers and cDNA concentrations used for the reactions were predetermined as described to enable linear and high-efficiency response (http://www.abgene.com/downloads/article-SYBRoptimise.pdf). The reaction mixture contained forward and reverse primers and Power SYBR Green PCR Master Mix (Applied Biosystems, USA) in a 10-μL total sample volume. Reactions were analyzed on a Rotor-Gene 3000 PCR machine (Corbett Life Research, Australia) using 35 cycles of 95 °C for 10 s, 60 °C for 15 s, 72 °C for 20 s, and 80 °C for 10 s. Data obtained were analyzed with the Rotor-Gene 6 software. Primer sequences and amplicon lengths are given in Additional Table S1. The qBase Quantification Software (http://medgen.ugent.be/qbase/) was used and data are expressed according to the delta-delta-Ct method. Each biological sample was examined in duplicate with two to three technical replicates. Gene-specific oligonucleotide primers were designed using Primer Express® version 2.0 software (Applied Biosystems, USA).

Results and discussion

Ripeness and storage parameters

There was no significant difference in cell size or ethylene production between the peach and the nectarine fruits at harvest (Table 1). Average cell size measured in the fruit of both cultivars beginning 3 weeks before harvest was 66 μm and increasing to 80 and 81 μm at harvest in the peach and nectarine fruits, respectively. The literature discussing the cell size has been reported in Dagar et al. (2011). In addition, the significantly smaller fruit and higher SSC and TA in nectarines have also been reported and discussed in Dagar et al. (2011). Wu et al. (2003) and Cantin et al. (2009) also reported a significant positive correlation between TA and SSC in peaches and nectarines. It might be possible that this result is because of the colocalization of quantitative trait loci involved in SSC and acid contents in peach (Dirlewanger et al. 1999). The similar ethylene production and cell size of these near-isogenic fruits indicated that, at harvest, both were at the same physiological stage.

Table 1 Representative ripeness parameters at harvest of the “Oded” peaches and “Yuval” nectarines
Full size table

In contrast, the peach cultivar and the derived nectarine showed dramatic differences in the way they withstood extended periods of cold storage. These results were reported and discussed in Dagar et al. (2011). In cultivars sensitive to chilling injury, disorders appeared after 2 or 3 weeks of 5 °C storage (Zhou et al. 2001). In this regard, both cultivars were relatively resistant to chilling injury; however, nectarines were more resistant than peaches. These results are in agreement with the findings of Crisosto et al. (1999) who reported that nectarines have better storage and shipping characteristics than peaches.

In summary, peach and near-isogenic nectarine, which were at a similar physiological stage at harvest, showed differences in the way they responded to cold storage. We anticipate that those differences may already be in part reflected at the transcript level of the fruits at harvest.

Microarray analysis

Two microarray platforms, ChillPeach and μPEACH1.0, were used to compare nectarine and its peach progenitor. The overlap of the two platforms is 29 % (Granell, unpublished). A total of 2,584 ChillPeach probes (∼61 %) met the threshold for hybridization quality. A total of 222 ChillPeach unigenes were significantly different at a p value <0.05 (Additional Table S2). Of these, 69 genes were more highly expressed in nectarine than peach and 153 were more expressed in peach than nectarine.

In the μPEACH1.0 microarray, a total of 2,358 genes met the threshold of hybridization quality for the two fruits. There were 195 genes showing statistical significance at p < 0.05 between peach and nectarine (Additional Table S3). Of these, 57 genes had higher expression in nectarine than in peach and 138 were higher in peach.

A comparison of the functional categorization of the differentially expressed genes in the two platforms is shown in Table 2. It was found that, in the ChillPeach and μPEACH1.0 microarray platforms, nectarine was lower than peach in abundance of the majority of transcripts related to cell wall, lipid metabolism, RNA transcription regulation, signal transduction pathway, trafficking machinery and membrane dynamics, and transport. Exceptions to this trend were a few categories including secondary metabolism (in ChillPeach microarray platform), antioxidant system (in μPEACH1.0 microarray platform), RNA posttranscriptional regulation (in μPEACH1.0 microarray platform), and RNA translation and protein assembly (in μPEACH1.0 microarray platform), with more genes higher in nectarine compared to peach.

Table 2 Functional categorization of differentially expressed genes of ChillPeach and μPEACH1.0 microarray platforms as well as common genes between the two microarray platforms in the “Yuval” nectarine compared to its peach progenitor “Oded” at harvest (“Yuval”/“Oded”)
Full size table

Few studies in peaches and nectarines have utilized the global gene expression and transcriptome analysis approaches to elucidate chilling injury, and all of them were focused either during cold storage or during ripening (shelf life) after cold storage. Pegoraro et al. (2010) reported that, during ripening of two different peach cultivars, previously stored under cold storage (30 days at 4 °C followed by a 5-day shelf life) and controlled atmosphere, the differential expression of genes involved in transport and cell wall metabolism analyzed by quantitative PCR were lower in the peach resistant to woolliness compared to the susceptible peach. The authors suggested that prevention of chilling injury in peaches and nectarines is not completely dependent on the increase in abundance, but may also be tied to the decrease in abundance of these groups of genes.

The genes from the two platforms were further filtered to those which were more than twofold different between the two fruits (Table 3). There were a total of 49 (22 and 27 with higher and lower expressions, respectively, in nectarine compared to peach) genes in ChillPeach and 58 genes (25 and 33 with higher and lower expressions, respectively, in nectarine compared to peach) in μPEACH1.0. The pattern of lower expression of genes in nectarine than peach was maintained in this case as well. There were nine genes that were common between the two platforms with over twofold different expression, and most of these genes (seven) were lower in nectarine and higher in peach (Fig. 1). The expression of selected common genes between the two platforms was validated by qRT-PCR (Additional Table S1 and Additional Fig. 1)

Table 3 List of genes with ≥2.0-fold increase or decrease in expression in “Yuval” nectarine compared to “Oded” peach mesocarp tissue samples at harvest
Full size table
Fig. 1
figure 1

Venn diagram comparing the significantly and differentially expressed transcripts in “Yuval” compared to “Oded” with fold changes ≥2.0. The expression patterns of the common nine (two upregulated and seven downregulated in “Yuval”) statistically significant transcripts between the ChillPeach and μPEACH1.0 microarray platforms are reported

Full size image

In summary, the overlap between the two microarray platforms is small enough so that, together, the two platforms give a broader view than either one alone. In addition, the fact that, in both platforms, the shared genes have similar expression patterns further supports our approach of using these two independent microarrays to investigate differences in gene expression between the two cultivars.

Among the genes common in both microarrays over twofold higher in nectarine was sulfolipid synthase, which is involved in lipid metabolism (validated by qRT-PCR; Additional Fig. 1). Other genes twofold higher in nectarine compared to peach included glutathione S-transferase GST22 (validated by qRT-PCR; Additional Fig. 1); catalase and thioredoxin, related to antioxidant system; endo-1,4-β- d -glucanase and β-galactosidase, precursors involved in cell wall degradation; mitochondrial processing peptidase involved in protein degradation; and chalcone synthase 2, related to secondary metabolism. In contrast, among the common genes higher in peach compared to the nectarine mutant were glutamate dehydrogenase 2, related to amino acid metabolism; S-adenosylmethionine decarboxylase, related to secondary metabolism; calcineurin B-like (CBL)-interacting protein kinase 6 and serine–threonine protein kinase, involved in signal transduction pathway; and a gene encoding hypoxia-responsive protein in the stress responses (the selected genes were validated by qRT-PCR as listed in Additional Table S1 and shown in Additional Fig. 1). Other genes higher in peach compared to nectarine included: trehalose-6-phosphate phosphatase, involved in carbohydrate metabolism; xyloglucan endotransglucosylase, endopolygalacturonase, and pectinesterase, precursors involved in cell wall degradation; ceramide glucosyltransferase and omega-6 fatty acid desaturase, involved in lipid metabolism; gibberellin 2-oxidase and flavonoid 1-2 rhamnosyltransferase, related to secondary metabolism; and a number of stress response genes. Overall, 47 genes were ≥2.0-fold higher in nectarine and 60 in peach (Table 3). Hence, in the present study, we found that, at harvest, nectarine had more genes with lower expression compared to peach. This suggests that, perhaps, either a positive regulator responsible for the expression of these genes is mutated in peach to yield a nectarine phenotype or some of the transcripts expressed in nectarine are not represented since both microarray platforms were constructed using ESTs from peaches.

Most of the studies on transcriptome analysis were performed on fruit at the beginning or during the development of chilling injury. Gonzalez-Aguero et al. (2008) utilized a nylon macroarray from a ripe peach fruit cDNA library to study gene expression changes of “O’Henry” peach fruit. The authors compared healthy fruit (juicy; at 7 days ripening after harvest) with chilling-injured fruit (woolly, nonjuicy; at 7 days ripening after 2 weeks cold storage at 4 °C) and found 106 genes including cell wall metabolism- and endomembrane trafficking-related genes to be differentially expressed between juicy and woolly fruit. In addition, they also found lower expression of cobra, endopolygalacturonase, cinnamoyl-Co-A-reductase, and rab11 genes in woolly fruit compared to juicy fruit.

A more comprehensive study of changes in “O’Henry” peach fruit was performed with comparative EST transcript profiling (Vizoso et al. 2009). They identified genes differentially expressed during ripening, in response to cold storage, or combined effects of cold storage and ripening. They ordered 1,402 normalized unigenes into 13 clusters according to the gene expression patterns. Among the 114 genes in the cluster that increased in expression in fruits in cold storage, there were six genes that had been identified by Ogundiwin et al. (2008) in their study of the ESTs used to make the ChillPeach microarray. Ogundiwin et al. (2008) utilized the ChillPeach microarray to compare cold-stored fruit tissue to fruit at harvest. There were 287 genes significantly higher and 74 lower in cold-stored fruit compared to nonripe fruit at harvest. Of the 287 upregulated genes, 74 were ≥2-fold higher, and of the 74 downregulated genes, nine were <0.5-fold expression. Twelve of these genes are also present as differentially expressed in “Oded” peach and “Yuval” nectarine at harvest, and most were higher in nectarine than peach. These included β-galactosidase precusor, chalcone synthase 2, and genes encoding little protein 1 (Table 3), heat shock 70 protein, and tonoplast intrinsic protein (Table 4), while lower in nectarine were genes encoding BZIP protein and a putative ripening protein and glutamate dehydrogenase 2 gene (Table 3).

Table 4 Genes that separate fruit according to storage potential as seen in Fig. 2
Full size table

A number of studies have been published utilizing the μPEACH1.0 microarray. These studies have examined peach fruit development (Bonghi et al. 2011), fruit ripening (Trainotti et al. 2006; Ziliotto et al. 2008), and the influence of various hormones on fruit development and ripening (Trainotti et al. 2007; Ziosi et al. 2008; Rasori et al. 2010; Bonghi et al. 2011). The microarray has also been used to investigate transcript profiling of apricot during development and ripening (Manganaris et al. 2011). One study examined two cultivars of peaches ripened after storage of 4 or 6 weeks (Falara et al. 2011). They found a number of cell wall-modifying proteins (β-d-xylosidase and expansin) and stress proteins (HSPs, dehydrin, and PR-4B) whose expression increased in ripening without storage and after storage in the chilling injury-resistant cultivar. This corresponds with the findings in this study, which found a dehydrin gene (Table 3), genes encoding an HSP (Table 4), and cell wall enzymes (Table 3) more highly expressed in nectarine than in peach.

It was of interest to see if differences in gene expression shown at harvest could be used to predict storage behavior, so a correlation study between gene expression and this trait was conducted. To widen the comparison of storage behavior, we utilized also the results of gene expression at harvest of the fruits that had been used to prepare the ChillPeach microarray. A quantitative comparison was possible as the “Yuval” nectarine, “Oded” peach, and mature tolerant and sensitive peach samples (Ogundiwin et al. 2008) were hybridized against the same reference pool. These samples were pools of different sibling genotypes coming from a breeding population and were either sensitive (S; showing chilling injury after 1 week of cold storage plus shelf life) or tolerant (T; developing chilling injury after 2–3 weeks of cold storage plus shelf life). A PCA was obtained with the expression values for the 222 differentially expressed genes in the ChillPeach microarray (Additional Table S2) for the mature sensitive (M-S), mature tolerant (M-T), “Oded” peach (O), and “Yuval” nectarine (Y) representing the four types of mature samples with increasing tolerance to cold stress: M-S < M-T < O < Y. The analysis revealed that, in the PCA space, samples were arranged so that, according to the depicted diagonal of PCA1 and PCA2, they could be projected into it according to their tolerance (Fig. 2). We next selected genes contributing to the loading plots on the ends of this diagonal in the score plot PCA1/2. In Fig. 2, the pattern of expression of 41 genes (listed in Table 4) selected could be seen with empty circles repressed (DR; downregulated) with tolerance and solid circles induced (UR; upregulated) with tolerance. This indicated that, out of the 222 genes that could contribute to the differences in the physiology and postharvest behavior, 41 of them were also able to explain the way four different genotypes are distributed according to their tolerance to cold stress in storage. Selected genes of these 41 were validated by qRT-PCR as listed in Additional Table S1 and shown in Additional Fig. 1.

Fig. 2
figure 2

A PCA showing the score plots obtained using the 222 genes that showed statistical significance between “Yuval” and “Oded” but including also the results from the M-S and M-T. The position of the samples from the score plots is indicated. The genes contributing to loading plots at the ends of the diagonal in the score plot of PCA1 and PCA2 are shown with solid circles for genes higher in tolerant fruit and open circles for genes higher in sensitive fruit. The insets show the expression patterns of these genes; 30 genes repressed (DR; downregulated) with tolerance and 11 genes induced (UR; upregulated) with tolerance. The x-axis of the insets indicates the expression of the replicates of the four fruits, with highest expression of the 30 repressed with tolerance genes in the M-S and highest expression of the induced with tolerance genes in “Yuval”

Full size image

Out of these 41 genes, 11 had higher expression in the chilling-tolerant fruit than in the chilling-sensitive fruit and 30 had lower expression in the chilling-tolerant fruit (Table 4). In examining the functional categories of these genes, it was found that the antioxidant system, structure and maintenance proteins (chaperones), and transport and trafficking categories had about the same number of genes with higher and lower expressions in the chilling-tolerant fruit compared to the chilling-sensitive fruit. Other major categories with lower expression in the chilling-tolerant fruit compared to the chilling-sensitive fruit were cofactor and vitamin metabolism, signal transduction pathway, and secondary metabolism. It is interesting to note that many of these genes are not more than twofold different in expression in nectarine compared to peach and, therefore, are absent from Table 3. Three of the genes higher in the cold-tolerant fruit were over twofold higher expression in nectarine, while nine genes were less than twofold higher expression in nectarine. The remaining 30 downregulated genes in the chilling-tolerant fruit were significantly different, but not by twofold.

In the PCA results of genes separating the cold-tolerant fruit from the cold-sensitive fruit, there were three times fewer genes with higher expression in the cold-tolerant fruit than in the cold-sensitive fruit (Table 4). Despite that general pattern, more antioxidant and other protective genes were expressed over twofold higher in nectarine than in peach (Table 3). Among the antioxidant genes in the cold-tolerant side was glutathione S-transferase GST22 (validated by qRT-PCR; Additional Fig. 1). The glutathione S-transferase protein belongs to the tau subfamily and is involved in cellular detoxification due to its ability to conjugate endobiotic and xenobiotic compounds to glutathione (Nilo et al. 2010; Lo Piero et al. 2009). The higher expression levels of glutathione S-transferase in nectarine are in agreement with previous findings of Nilo et al. (2010), where the authors found that increased accumulation of a glutathione S-transferase protein in the cold-stored peach fruit is associated with the fruits’ increased capacity to withstand low-temperature stress.

Signal transduction pathways include entire physiological processes from signal reception to cellular response. There were fewer genes in this category with higher expression in nectarine than in peach and also in the PCA separating chilling-tolerant fruits from chilling-sensitive fruits (Table 4).

The roles of various kinases as important signal transducers during low-temperature stress have been demonstrated, although their signaling pathways are not fully understood. Surprisingly, we found more genes with higher expression in the chilling-sensitive fruits compared to the chilling-resistant fruits. One of the kinases was a CBL-interacting protein kinase, which is activated by a CBL in the presence of Ca2+. However, in Arabidopsis, CBL-1 has been reported as a positive regulator of salt and drought responses but a negative regulator of cold response (Cheong et al. 2003). Therefore, the higher expression of CBL-interacting kinase gene (validated by qRT-PCR; Additional Fig. 1) in chilling-sensitive peach might act as a negative regulator of low-temperature stress and thus reduce its resistance to the stress.

Chilling injury is a multigene process, and the discovery of genes that have emerged from our study may be a step in understanding this complex process. Differences that are found at harvest might indicate how well different cultivars of peaches and nectarines can be stored, and this can lead to the identification of candidate genes which will be of use to breeding chilling-resistant cultivars.