Abstract
Background
Interspecific hybridizations and admixtures were key in Citrus domestication, but very little is known about their impact at the transcriptomic level. To determine the effects of genome introgressions on gene expression, the transcriptomes of the pulp and flavedo of three pure species (citron, pure mandarin and pummelo) and four derived domesticated genetic admixtures (sour orange, sweet orange, lemon and domesticated mandarin) have been analyzed at color break.
Results
Many genes involved in relevant physiological processes for domestication, such sugar/acid metabolism and carotenoid/flavonoid synthesis, were differentially expressed among samples. In the low-sugar, highly acidic species lemon and citron, many genes involved in sugar metabolism, the TCA cycle and GABA shunt displayed a reduced expression, while the P-type ATPase CitPH5 and most subunits of the vacuolar ATPase were overexpressed. The red-colored species and admixtures were generally characterized by the overexpression in the flavedo of specific pivotal genes involved in the carotenoid biosynthesis, including phytoene synthase, ζ-carotene desaturase, β-lycopene cyclase and CCD4b, a carotenoid cleavage dioxygenase. The expression patterns of many genes involved in flavonoid modifications, especially the flavonoid and phenylpropanoid O-methyltransferases showed extreme diversity. However, the most noticeable differential expression was shown by a chalcone synthase gene, which catalyzes a key step in the biosynthesis of flavonoids. This chalcone synthase was exclusively expressed in mandarins and their admixed species, which only expressed the mandarin allele. In addition, comparisons between wild and domesticated mandarins revealed that the major differences between their transcriptomes concentrate in the admixed regions.
Conclusion
In this work we present a first study providing broad evidence that the genome introgressions that took place during citrus domestication largely shaped gene expression in their fruits.
Similar content being viewed by others
Background
Citrus are among the major fruit crops worldwide, with oranges, grapefruits, lemons and mandarins as the most economically relevant cultivars. Most of these commercial citrus are not pure species but genetic admixtures derived from interspecific hybrids that display genomic fragments from wild species, which are pure and hence free of introgressions, including citron, pummelo and mandarin [1]. Previous studies have classified mandarins based on the proportion of their genome containing introgressions: pure mandarins, which are generally considered wild species, were classified as type 1 mandarins by Wu et al. [1], and are not edible and small in size. In contrast, commercial mandarins are largely appreciated for their palatability, which has been associated with several pummelo introgressions. These admixed mandarins are classified as type 2 or type 3 mandarins, depending on whether the proportion of the genome including a pummelo haplotype is lower or higher than 10% [1, 2]. Other well-known citrus cultivars such as sweet oranges and grapefruits are also mandarin/pummelo admixtures, although they show larger and more frequent pummelo introgressions, with some genomic regions displaying two pummelo alleles [3]. As direct hybrids of mandarin x pummelo cross, sour oranges display two complete parental haplotypes. Lemons, that resulted from of a cross of sour orange and citron [4], have one complete citron haplotype, while the other one shows the admixture produced by the mandarin and pummelo ancestries [4, 5].
Genomic analyses suggest that the specific admixture patterns of each citrus cultivar largely determines its phenotype, and might imply human participation [1, 3, 6]. The generation of such admixtures is a complex process that requires crosses between pure species followed by backcrosses and/or crosses with other admixtures. This process has been related to the domestication of citrus species, together with the selection and propagation of the admixed individuals showing desirable traits. The complex relatedness network shared by mandarins, oranges and grapefruits, suggesting that they all share some recent common ancestors, supports this proposal [1].
Human-mediated selection has been the major driving force behind crop domestication, a process that is still ongoing nowadays. This selection unknowingly targeted structural and regulatory genes, producing effects that propagate through the whole transcriptome [7]. For this reason, domestication has profound effects in gene expression, affecting specific genes associated with the agronomical traits targeted by the domestication process but often altering the transcriptome at a global level too. This way, comparative transcriptomic revealed patterns of selection in domesticated tomato [8, 9], and RNA-seq performed at the population level showed how artificial selection greatly shaped the tomato transcriptome, altering the fruit sugar content and resistance to abiotic and biotic stresses [10]. Resha** of the maize transcriptome by domestication has been also analyzed by expression profiling analyses, identifying several genes that may have contributed to this process [11].
Knowledge of the genetic changes that occurred during the domestication and breeding of perennial trees is limited, although RNA sequencing analysis of wild, landrace, and improved cultivars of pear (Pyrus pyrifolia), for instance, revealed specific patterns of domestication and improvement, many of them highly associated with important fruit traits [12]. Evolutionary transcriptomics has been also used to reveal the origins of olive trees (Olea europaea) and the genomic changes associated with their domestication, showing how the domestication of this species had a moderate impact at genome level and that the domestication syndrome was mainly related to changes in gene expression, consistent with the olive tree evolutionary history [13].
Citrus fruits display a wide variability in size, shape, color and flavor, which make them attractive to the markets. Despite this broad phenotypic diversity, genomic studies have reported a highly conserved genome, both in structure and gene content. Thus, all the Citrus analyzed genomes are organized in 9 chromosomes, that show an almost perfect synteny [4, 14], as well as a very similar number of highly conserved genes [15, 16]. Therefore, the variability found in citrus must rely in other factors, and changes at the expression level might appear as some of the influential ones that could be ultimately associated to the domestication of Citrus species.
As fruit quality is a direct consequence of the ripening process, much effort has been made to analyze maturation at different levels, and recently several studies have used transcriptomic approaches to unveil the genetic mechanisms controlling citrus fruit ripening. These works focused on the study of the regulation of commercially relevant traits such as fruit taste [22]. In some acidic citrus, the vacuolar lumen in the cells of the pulp vesicles reaches pH values as low as 2, more than five points below the cytoplasmic one [23]. This steep pH gradient is promoted by the citrate vacuolar intake [24], which buffers the vacuolar lumen and allows a continuous proton intake that maintains the low vacuolar pH [25]. Despite its central role in this process, citrate biosynthesis does not appear to be directly correlated with its accumulation in citrus pulp [22, 29]. In non-acidic commercial citrus such as sweet oranges and clementines, sugar accumulation is a major trait for fruit quality. This accumulation is the result of a metabolic change from sucrose utilization to storage [30], which increases sucrose concentration in the pulp [31]. As fruit ripens, expression levels of sugar invertases (INV) drop, while sucrose phosphate synthases (SPS), sucrose phosphatases (SPP) and sucrose synthases (SuSy), all involved in sucrose synthesis, increase their expression [32], all of which result in the accumulation of sucrose and its derivatives in the pulp during the late ripening stages [33].
Color is another major agronomical trait of citrus fruits, mostly produced by the accumulation of carotenoids. During color break, peel chlorophylls are hydrolyzed [34] while carotenoid biosynthesis is promoted in the chromoplasts [35] and, in red-colored fruits, the carbon flux is redirected towards the production of β-carotene and its derivatives [36]. The differential accumulation profile of carotenoids and apocarotenoids generates the broad range of colors observed in Citrus species, a subject thoroughly studied given the commercial relevance of their fruits [reviewed in [37]. Specifically, the bright red color found in many citrus is produced by the accumulation of C30 apocarotenoids such as β-citraurin, while other carotenoids such as violaxanthin also contribute to the final color [38]. In contrast, yellow and non-colored fruits such as pummelos, citrons and lemons display lower carotenoid contents [39,40,41]. In the pulp vesicles of most citrus species, the accumulation of carotenoid derivatives follows a similar process, although it starts earlier and reaches lower values overall [42, 43].
Citrus well-known health benefits [44] result from the large number of bioactive compounds, like flavonoids, found in their fruits. Recent studies have highlighted the vast flavonoid diversity existing in Citrus species, especially that of polymethoxylated flavonoids and O-glycosylated flavonoids on the fruit flavedo, where their concentration is higher [19]. Moreover, flavonoid profiles in different citrus are extremely variable between species and admixtures, even allowing their clustering based on such profiles that greatly resembles the phylogenetic tree of the genus Citrus [45]. This variability might be the result of the expansion of the genic family of flavonoid O-methyltransferases that characterize citrus genomes [46].
In this work, we have used the RNA-Seq technology to address how domestication shaped gene expression in pure Citrus species, analyzing transcriptomic changes caused by hybridization and admixtures, to generate the attractive commercial varieties we enjoy nowadays. While fruit ripening has been thoroughly studied in citrus, most works have focused on comparative studies of somatic mutants or closely related cultivars [18, 47, 48], but no genus-wide analysis has been performed so far. We have performed a comprehensive transcriptomic analysis of flavedo and pulp of fruits of seven different citrus cultivars at the time of color break. Three of them are pure species and belong to the main citrus taxonomic groups (citrons, pummelos and mandarins), while the remaining four are commercial varieties with varying admixture levels. Using a novel approach that involves the analysis of genomic unbalance and allele-specific expression, we provide new insights of the effects of citrus hybridizations and domestication on gene expression during ripening.
Results
Physiological characterization of ripening fruits
Fruit acidity and sugar content of the seven selected species were measured during the ripening process. Major changes in juice acidity were only observed in the two mandarins, whose acidity decreased considerably as ripening progressed (Fig. 1). Citron, lemon and sour orange were characterized by a titratable acid (TA) content above 5% during the whole period; conversely, sweet orange and pummelo showed a constant, low level of TA content (Fig. 1a). Sugar content (measured in Brix degrees, or °Brix) increased considerably in the two mandarins, while it remained invariably high in pummelo, very low in lemon and citron, and in intermediate values in the two oranges (Fig. 1b). The color break and the fruit color at the time of sampling is shown in Fig. 1c.
Phenoty** of the selected accessions across three months in six timepoints. a Average titratable acid content per sample across three replicates. Vertical bars represent the standard error for each measurement. b Average Brix degrees per sample. Vertical bars represent the standard error for each measurement. c Average color per sample at six measuring times, indicated by the sampling date. Color was calculated from the L*a*b values provided by the Minolta colorimeter. The color values between measurements were interpolated. Diamonds mark the color and date at which samples were collected and processed for RNA-seq sequencing. CHP: C. maxima, DIA: C. medica, EUR: C. limon, SCM: C. reticulata, SSO: C. aurantium, SWO: C. sinensis, WLM: C. deliciosa
RNA-seq read map** and sample clustering
RNA-Seq analysis yielded an average of 24 million reads per sample, 90% of which mapped to the 25,434 annotated transcripts from the clementine reference genome. A principal component analysis was performed based on transformed read counts (see methods), from pulp and flavedo samples. As expected, the three replicates from each sample clustered together in all cases, supporting the reproducibility of the results (Fig. S1). Pure species were the most different samples in both tissues, as indicated by their dispersion in the PCA plot (Fig. 2). In contrast, admixed samples, and most notably sweet and sour oranges, were scattered in between the ancestral pure species.
Principal component analysis of expression data. Each replicate is shown as an independent dot; flavedo and pulp samples were analyzed independently and are displayed in a) and b), respectively. Only the principal components PC1 and PC2 are shown. CHP: C. maxima, DIA: C. medica, EUR: C. limon, SCM: C. reticulata, SSO: C. aurantium, SWO: C. sinensis, WLM: C. deliciosa
Hierarchical clustering of pulp samples grouped together lemon and citron, apart from the other species. A second cluster only included pummelo, and the third one grouped the two mandarins (pure and admixed), and the two oranges (sour and sweet). Similar results were obtained for flavedo samples (Fig. S1).
Differentially expressed genes during fruit ripening
In order to find differentially expressed genes among samples, all versus all pairwise comparisons were performed, as well as between groups of samples based on their phenotype. The last ones yielded the most significant and informative results, so we focused our study on them. For example, the comparison of pulp from citron and lemon, highly acidic and with low sugar content, against the remaining species, retrieved 6681 DEGs, while 1329 DEGs were found when compared flavedo from red and yellow fruits (at least one two-fold expression change, s-value < 0.01). Roughly 20% of the DEGs corresponded to uncharacterized loci and genes of unknown function, matching the uncharacterized proportion of the total genic space of Citrus clementina. Four differentially expressed genes and the housekee** gene CitUBC1 were selected to perform RT-qPCR confirmation of the differential expression analysis results. Overall, the log fold changes obtained from the RNA-seq data matched those calculated via the ΔΔCt method (Table S1), with a Pearson correlation coefficient between RNA-seq fold changes and ΔΔCt of 0.91 and 0.89 in pulp and flavedo samples, respectively.
Differentially expressed genes associated with relevant metabolic pathways involved in fruit ripening were selected for further analyses: glycolysis / gluconeogenesis, tricarboxylic acid (TCA) cycle, vacuolar proton intake and secondary metabolism including carotenoid and flavonoid biosynthesis: a brief list of DEGs possibly involved in Citrus ripening were reported in Table S2. As citrus fruit taste mostly depends on citrate and sugar accumulation and the vacuolar pH of the pulp, we found interesting to compare the acid low-sugar fruits lemon and citron versus the remaining species (Fig. 1), and found that many genes involved in sugar metabolism and the TCA cycle consistently showed differential expression patterns. This way, most of the genes involved in these processes displayed lower expression levels in lemon and citron when compared to the remaining samples, including those genes involved in the TCA and the γ-aminobutyric acid (GABA) cycles, while most subunits of the vacuolar ATPase were overexpressed in these 2 species (Fig. 3). Besides these pathways, some specific genes previously related to sugar and citrate accumulation were analyzed. For example, two isoforms of CitPH5, a P-type ATPase involved in vacuolar acidification [22], were more expressed in the most acidic species, citron, lemon and sour orange, although similar expression levels were observed in pummelo in one of these genes (Fig. 4a). The expression of specific transcription factors regulating sugar and acid accumulation in other fruits [49,50,51] was also analyzed, and their levels in the different pulp correlated with sugar accumulation and titratable acidity (Fig. 4b). Functional annotation and enrichment analysis of the DEGs found in citron and lemon pulp samples revealed a significant enrichment in terms related with organic acid metabolism and ion transport via ATP hydrolysis (Table S3). We could not find these patterns in the flavedo samples, where the number of DEGs was considerably lower when compared to the pulp ones.
Differentially expressed genes involved in sugar metabolism. DEGs found in flavedo (a) and pulp (b) are shown independently. Each bar represents the expression log2 fold change comparing lemon and citron against the remaining samples (only genes with a log2 fold change > 1, s-value < 0.01, are shown). SuSy: sucrose synthase, SPS: sucrose-phosphate synthase, INV − Ac: acid invertase, INV − AN: alkaline/neutral invertase, HK: hexokinase, G6PI: glucose-6-phosphate isomerase, PFK: 6-phosphofructokinase, ALDO: fructose-bisphosphate aldolase, TPI: triosephosphate isomerase, G3PD: glyceraldehyde-3-phosphate dehydrogenase, PGK: phosphoglycerate kinase, PGM: phosphoglycerate mutase, ENO: enolase, PK: pyruvate kinase, PDH: pyruvate dehydrogenase, CS: citrate synthase, ACO: aconitate hydratase, IDH: isocitrate dehydrogenase, OGD: 2-oxoglutarate dehydrogenase, SDH: succinate dehydrogenase, FH: fumarate hydratase, MD: malate dehydrogenase, VATP: V-type proton ATPase subunit, GDH: glutamate dehydrogenase, GDC: glutamate decarboxylase, GABAT: γ-aminobutyrate transaminase
Expression levels of selected genes in each species. The selected genes are involved in a) pulp acidity, b) sugar accumulation or c) carotenoid accumulation. Dot colors represent different samples as in Fig. 2. CHP: C. maxima, DIA: C. medica, EUR: C. limon, SCM: C. reticulata, SSO: C. aurantium, SWO: C. sinensis, WLM: C. deliciosa
A two-way comparison of red (sweet and sour orange, pure and admixed mandarin) against yellow (lemon, citron and pummelo) colored varieties revealed only a few differentially expressed genes (Fig. S2). Indeed, comparisons between sample pairs yielded several differentially expressed genes of the carotenoid biosynthetic pathway, albeit not consistent across all the samples with the same fruit color (Fig. 4c). One of these genes encodes a phytoene synthase (PSY, LOC18039146), one of the first enzymes in the carotenogenic pathway, which was significantly downregulated in citron and lemon. The citrus ζ-carotene desaturase gene (ZDS, LOC112098231) was more expressed in red fruits than in yellow ones, although not all comparisons showed significant differences. Another gene coding for a key enzyme involved in the synthesis of β-β-carotenes, a β-lycopene cyclase (LCYb2, LOC18034834), was consistently overexpressed in all the red fruits compared to the yellow ones. Several genes coding for carotenoid cleavage dioxygenases (CCDs), involved in carotenoid accumulation and color setting, were also differentially expressed. One of them, CCD4a (LOC18043465), displayed lower expression levels in mandarins and oranges, while CCD4b (LOC18043103) was more expressed in sweet and sour oranges, lemon and pure mandarin. The zeaxanthin epoxidase gene (ZEP LOC18036737), involved in carotenoid degradation, was overexpressed in lemon and citron. It should be noted that these comparisons were performed on flavedo samples since color break takes place earlier in the pulp, although pulp samples presented similar expression patterns for most of these genes.
Finally, the flavonoid biosynthetic pathway was studied in order to address the huge variability of flavonoid derivatives found in Citrus flavedo [19], that corresponds with the high number of genes involved in the production and regulation of these secondary metabolites [46]. Indeed, flavonoid diversity correlates with the extreme diversity of expression patterns observed in a large number of genes involved in flavonoid modifications (Fig. S3), especially the flavonoid and phenylpropanoid O-methyltransferases (FOMTs) families. Many of these genes had no expression in at least one species, while showing considerably high expression levels in others (Fig. S3). The most noticeable differential expression was displayed by a chalcone synthase gene CHS that was exclusively expressed at high levels in mandarin and their admixed species, lemon, sweet orange, sour orange and commercial mandarin (Fig. 5a). We will refer to this locus as CHSm due to their mandarin-linked expression, which was confirmed in both pulp and flavedo samples.
CHSm expression and promoter. a Expression levels of CHSm in pulp and flavedo deduced from RNA-seq data. b Expression levels of CHSm in flavedo calculated from RT-qPCR in the samples used for RNA-seq and other Citrus species. The color code represents the number of mandarin alleles of CHSm according to the admixture patterns published by Wu et al. (2018): green for zero mandarin alleles, blue for one allele and orange for two alleles. c Multiple sequence alignment of a fragment of the promoter region of CHSm. The CAAT and MYB motifs, and the region discriminating the structural alleles are highlighted. CHP: C. maxima, DIA: C. medica, EUR: C. limon, SCM: C. reticulata, SSO: C. aurantium, SWO: C. sinensis, WLM: C. deliciosa
Segmental ancestry patterns and differential gene expression
To assess the effects of the different introgressions in the global gene expression and domestication, the number of DEGs between the two palatable admixtures (sweet orange and commercial mandarin), and the pure species that originated them (pummelo and the pure mandarin) were analyzed. Using the admixed regions provided by Wu et al. [1], each gene was assigned to a specific admixture pattern based on their chromosome location, and the distribution of DEGs across the genome was studied.
The number of genes differentially expressed between wild and domesticated mandarin was considerably higher on the pummelo/mandarin admixed regions of the genome than in the rest of it. This was especially evident at the end of chromosome 6 and the beginning of chromosomes 3 and 8 (Fig. 6). Among these DEGs, some could be directly related to ripening, such as the malate dehydrogenase MDH (LOC18034657), which was significantly more expressed in wild mandarin pulps, a flavanone rhamnosyltransferase (LOC18035124) highly expressed in pummelo and commercial mandarin, or the already mentioned CCD4b. Several transcription factors were also expressed at different levels between the two mandarins: 5 genes in the case of pulp samples, and 7 in the flavedo ones. In pulp, three out of the five differentially expressed transcription factors were related to vernalization, based on their similarity to the vernalization gene VRN1 (LOC18035586, LOC18034082 and LOC18034239); two of them were also differentially expressed in the flavedo. In contrast, in the pummelo - commercial mandarin comparison, more DEGs were found along the whole genomic space, but the pummelo/mandarin admixed regions were not particularly enriched on these.
Distribution of DEGs between commercial mandarin and its progenitor species. The number of DEGs along the genome is represented in four different comparisons. Bar height represents the number of DEGs in a given windows, with color representing the admixture pattern of these genes (blue: admixed, orange: non-admixed mandarin). Bars with two colors represent genomic windows spanning two different admixed regions and include genes with both ancestries. A: Flavedo of wild and domesticated mandarin. B: Pulp of wild and domesticated mandarin. C: Flavedo of pummelo and domesticated mandarin. D: Pulp of pummelo and domesticated mandarin. E: admixture patterns of the commercial mandarin genome and distribution of runs of homozygosity (in black) and genic abundance (in red)
In sweet orange, the number of DEGs in the admixed regions was not remarkably different when comparing this cultivar with its ancestors, but this contrasts with the observation made in non-admixed regions of the genome. Specifically, the main pummelo/pummelo region of the sweet orange genome, located at the end of the chromosome 2, was amongst the regions with a higher amount of DEGs against mandarin (Fig. S4), while it was not particularly enriched when compared with pummelo. Conversely, the mandarin/mandarin regions of the sweet orange genome, and especially those at the end of the chromosomes 3 and 9, displayed a large amount of DEGs against pummelo, while showing considerably less when compared with mandarin (Fig. S4).
The number of DEGs was also studied across the existing runs of homozygosity on the commercial mandarin and sweet orange genomes. The runs of homozygosity, here defined as genomic windows with a heterozygosity below 0.1%, were scarce in sweet orange, with only 1% of its genome being composed by runs of homozygosity. In the case of the commercial mandarin, this percentage ascended to 11.4%. In none of the two species DEGs were particularly enriched in the runs of homozygosity.
Allele-specific expression in Citrus ripening genes
In order to find allele-specific expression in the admixed genomic regions, the level of expression of each allele was assessed independently. Allele expression imbalance was calculated based on species-specific markers, diagnostic SNPs that allow phasing along the genome. This strategy can properly phase the two alleles in admixed species, but only in the genomic regions displaying the 2 ancestral haplotypes. As the extent of admixed regions differs greatly between species (from a mere 10% of the genome in commercial mandarin to the whole genome in lemon and sour orange), the total number of diagnostic SNPs was not directly comparable. However, the number of diagnostic SNPs per kb laid between 6 and 10 SNPs in the admixed region of every species (Table 1).
The phasing information was used to identify allele specific reads, so it was possible to quantify expression at the allele level, and therefore, to detect genes displaying allele-specific expression (ASE). In commercial mandarin, for instance, 380 genes displayed ASE in any of the two tissues: 99 were specific from flavedo, 160 from pulp, and 121 were common to both tissues. In lemon these numbers were considerably higher, with 1324 genes displaying ASE only in flavedo, 2021 only in pulp and 937 showing this behavior in both tissues. Despite the differences in the raw number of ASE genes, when considering the admixture proportions of each species, the number of genes displaying ASE was comparable with the exception of lemon, which displays a considerably higher number of ASE genes. For example, the number of ASE genes in the citron/mandarin admixed regions of lemon, which represent around 50% of the genomic space, was as high as that found in the pummelo/mandarin regions in the hybrid sour orange (Fig. S5). This was especially pronounced in pulp samples, that displayed more ASE genes than the flavedo ones in all the analyzed species. In the lemon pulp about 12% of the total genic content (2958 genes) displayed ASE, while it was around 8% in the case of sour orange (2232 genes) and the other two admixtures, where similar proportions were found.
Several genes showing allele-specific differential expression have relevant roles in the ripening process and might have been selected during the domestication process. For instance, the citron allele of one hexokinase gene (HK, LOC18035909) was preferentially expressed in lemon pulp, while in sweet orange the pummelo allele was the one overexpressed. The pummelo allele of a GABA transporter (LOC18035651), which might play a role in the GABA shunt, was preferentially expressed in commercial mandarin and sweet orange, while both the pummelo and mandarin alleles have similar expression levels in sour orange. Other examples include the abovementioned phytoene synthase PSY gene, for which the mandarin allele was preferentially expressed in sweet orange and commercial mandarin, and the carotenoid cleavage dioxygenase 4 CCD4a: in sweet and sour orange flavedo the pummelo allele was almost exclusively expressed, while in lemon the citron and pummelo alleles were expressed at comparable levels. Another relevant example is that of a chalcone synthase CHSm, in which only the mandarin allele was expressed in lemon, sweet orange and sour orange (Table S4), that was studied in detail.
Promoter structure of the chalcone synthase CHSm
As described above, in the pure species, CHSm was only expressed in mandarin. In the admixed commercial varieties, only the mandarin allele was expressed, and the pummelo and citron alleles were hardly expressed. To study putative causes of this reduced expression in citron and pummelo, the promoter region of CHSm was studied. A 500 bp sequence upstream of the CHSm locus was retrieved from the C. clementina reference genome, and orthologs in other available citrus genome sequences, including two different lemon haplotypes recently published [2, 4, 15, 16, 52,53,54,55], see Methods. By aligning the 11 sequences, we found a number of differential SNPs that allow the discrimination of the species-specific alleles; a region of almost 100 bp with different sequences between pure species was also found (Fig. 5c, complete alignment in Fig. S6), with three different structural alleles. A motif search using the NewPLACE database revealed the existence of some sets of motifs that were uniquely found in the mandarin allele and shared by the mandarin admixtures but are absent in the other species (Table S5), most of which arise due to specific SNPs (Fig. 5c). The nucleotide sequence was confirmed in citron, pummelo and mandarin via Sanger sequencing (see Table S6 for primer details). The expression of CHSm in different Citrus species was verified via RT-qPCR, revealing that its high expression was exclusive to mandarins, while it was intermediate in the admixtures harboring a single mandarin allele and very low in the rest of pure species (Fig. 5b).
Discussion
Admixture patterns and their role in sha** gene expression
Recent genomic studies of commercial citrus species have shown extensive introgressions in their genomes from three pure species: citrons, pummelos and mandarins. The principal component analysis (PCA) based on genomic sequence variability depicted the three pure species in the vertices of a triangle, with the admixtures scattered somewhere between them [1]. A phenotype-based PCA study, using 146 citrus botanical traits produced similar results [56]. In our study, the PCA based on gene expression also generated a highly similar distribution (Fig. 2). It is worth noting that completely different types of the data gave rise to the same distribution, linking the genomic, transcriptomic and phenotypical levels. This observation suggests that indeed admixture patterns might have a relevant role in regulating gene expression that would result in the wide phenotypic variability found in citrus fruits.
It has been reported, that in plant interspecific hybrids, gene expression can reach extreme values when compared to both parentals, in a process called transgressive gene expression [57]. This phenomenon is explained by inter-loci epistatic relations and complementarity between loci, among others [72], and furthermore, VRN1 genes are located on the abovementioned admixed region of the beginning of chromosome 8. This could explain why this admixed region is conserved among most if not all commercial mandarins, while also being extensively linked with domestication [1, 54, 66]. Other relevant genes found in this admixed region also include a GABA transporter GAT1 (LOC18035651), which is involved in the GABA intake from the apoplast [73] and could be related to citrate accumulation in citrus fruits [27]. Hence, the preferential expression of the pummelo GAT1 allele in commercial mandarin pulp might be linked with the domestication process.
Comparison of cultivated vs. wild mandarins showed that the differentially expressed genes were more frequently located on the small fraction of P/M admixed regions of the genome. On the other hand, in sweet orange, that has a mostly admixed genome, the major differences at the expression level were found in the scarce non-admixed regions. This might suggest that the specific admixture pattern of each commercial citrus cultivar has a global effect at the transcriptomic level, which appears to be related to the results obtained in the principal component analysis (Fig. 2). Indeed, interspecific hybridizations represent a major force in the domestication of many tree species such as olives, apple or grapevine [74,75,76], as well as in some annual crops such as tomato or maize [8, 77]. Given the ease of vegetative propagation of Citrus and the apomictic nature of many commercial cultivars, the domestication process of these species might have started with the initial generation of specific admixture patterns by hybridization between a reduced number of individuals, thus generating a considerable degree of inbreeding [1]. From there, cultivars of interest were selected, maintaining specific segmental ancestry patterns that provided desirable traits [1, 72], and further producing new cultivars via selection of somatic mutants.
Determinants on sugar and acidic contents among species
Gene expression analysis of acid fruits showed a large number of DEGs involved in carbohydrate metabolism in citron and lemon pulp (Fig. 3). Some of these genes coded for enzymes involved in hexose mobilization, like sucrose synthases (SuSy), sucrose phosphate synthases (SPS), sucrose phosphatases (SPP) and invertases (INV-Ac and INV-AN). Most of the sugars found in citrus fruits are actually synthetized elsewhere, and then translocated into them [78], and sugar sink strength has been linked to the hydrolysis of sucrose by several enzymes including SuSy [78, 79]. In other fruits, such as pear or apple, sugar accumulation has been also linked with hexokinase activity [80, 81]. In our study, sugar content remained invariably low across the whole ripening process in lemon and citron samples, which correlates with the reduced expression of the SuSy and hexokinase genes. Moreover, the latter also displayed ASE in lemon pulp, where the citron allele is preferentially expressed.
Interestingly, the expression of some specific transcription factors correlated with the accumulation of sugar in the pulp from some samples. For example, the transcription factor TRAB1 (LOC18033660) had the lowest expression level in lemon pulp (Fig. 4b). TRAB1 is a homolog of the Malus domestica gene MdAREB2, whose expression is related to an increased sucrose uptake, that causes a larger accumulation of soluble sugars, which in turn alters the fruit sink strength [50]. The reduced TRAB1 expression in lemon fruits could partially explain the lower sugar accumulation, although it cannot explain the differences in citron, which might depend upon other mechanisms.
Another remarkable observation is the consistently lower expression in lemon and citron of several genes involved in carbohydrates and organic acids metabolism, while the expression of genes coding for several V-ATPase subunits was significantly increased (Fig. 3). Moreover, many differentially expressed genes in citron and lemon pulp are involved in organic acids metabolism and ATP-dependent molecular transport of several molecules (Table S3). Recently, the P-type ATPase CitPH5, which is located on the tonoplast membrane, was postulated as a relevant factor in determining the vacuolar proton gradient [22], while other studies report that V-ATPases can fulfill the same role, complementing each other [82, 83]. We found that one of the isoforms of CitPH5 was indeed overexpressed in the three most acidic accessions, citron, lemon and sour orange (Fig. 4a), in agreement with previous results [22]; however, another CitPH5 isoform was also highly expressed in pummelo, whose acidity is very low. The consistent overexpression of V-ATPases in citron and lemon pulp, coupled with the higher CitPH5 P-ATPase activity, suggest that the two mechanisms might be working in these species, maintaining a high acidity throughout the whole ripening process (Fig. 1).
The study in detail of transcription factors associated with the regulation of citric acid content showed a MYB transcription factor, CrMYB73, that was overexpressed in citron, lemon and, to a lesser extent, in sour orange (Fig. 4b). CrMYB73 is highly similar to the Petunia hybrida PhPH4 gene, and is involved in the accumulation of citric acid in mandarin pulp over time [22]. Hence, our results suggest a relevant role of CrMYB73 in the accumulation of acid in citron, lemon and sour orange.
Despite the evidence of some allelic imbalance affecting rate-limiting glycolytic enzymes in lemon pulp, our work indicate that the differential expression of genes involved in many steps of sugar metabolism might be the main cause of the differences in acid contents in fruit. We present evidence supporting that CitPH5 is relevant for citrus acidity, since it is certainly overexpressed in the three most acidic samples: citron, lemon and sour orange (Fig. 4a). However, our data also suggest the possibility of an additional mechanism besides CitPH5, that would involve the accumulation of citrate in the pulp of citron and lemon by an increased V-ATPase activity (Fig. 3). This is likely the result of the complex interplay among the transcription factors analyzed above, and very possibly many others not studied in this work. The overall reduction in sugar accumulation in these fruits would also contribute to increase their sourness.
Citrus pigmentation from a genus-wide perspective
We observed several changes in the expression of multiple key genes involved in carotenoid biosynthesis (Fig. S2), suggesting that citrus color might not depend on a single master gene, but instead on the additive effects of several genes (Fig. 4c). For example, a PSY gene, whose expression in some citrus has been shown to correlate with total carotenoid content [85, 86], was downregulated in citron and lemon fruits. The total carotenoid content of these fruits has been thoroughly characterized in previous studies [87, 88] showing that the carotenoid content of yellow fruits such as lemons, citrons and pummelos are lower than those of red-peeled varieties such as mandarins or oranges. We did not find the reduction of PSY expression in pummelo, which also produces yellow fruits, suggesting that a different mechanism might be involved in the low carotenoid content described in this species.
Another differentially expressed gene coding for a ζ-carotene desaturase was overexpressed in the flavedo of sweet orange and the commercial mandarin, both producing red fruits. ZDS is essential for the red coloration of tomatoes, as it is required for the production of lycopene and β-carotene derivatives [89]; ZDS has been also associated with carotenoid biosynthesis in carrot [90]. In Citrus, the sweet orange Pinalate mutant, that produces yellow fruits, was initially thought to be a ZDS defective mutant [91], linking ZDS activity with the red coloration of sweet oranges. Further studies revealed that the defective gene was a ζ-carotene isomerase, and not a desaturase [92]. The high expression levels of ZDS gene that we found in sweet oranges and commercial mandarin might suggest that ZDS might after all be involved in the red pigmentation of oranges and mandarins.
One of the main branching points in the carotenoid biosynthesis pathway is the lycopene cyclization, carried out by the lycopene β-cyclase LCYb, which funnels the carbon flux towards the β-β-carotene production [36]. Actually, LCYb gene expression has been shown to increase during color break of mandarin and orange fruits, suggesting its role in this process [87, 93]. In this work we show that the LCYb2 gene was consistently overexpressed in the flavedo of all red fruits, when compared to the yellow ones from citron, pummelo and lemon, both in the pairwise and in the red against yellow fruits comparisons (Fig. S2). The role of LCYb2 directing the carbon flux of the carotenoid pathway towards β-carotene and its derivatives has already been suggested in Citrus [36, 37] and other species such as carrot [94]. Overall, our results suggest that LCYb2 activity might be involved with the fruit red coloration in different species from the genus Citrus, as its expression is consistently higher in red-colored peels.
The gene coding for a zeaxanthin epoxidase also presented differential expression patterns, being overexpressed in citron and lemon when compared with the remaining species, although only in pummelo, sweet orange and sour orange were statistically significant. In Arabidopsis, ZEP defective mutants accumulate β-carotene, β-cryptoxanthin and zeaxanthin due to a metabolic blockage in carotenoid degradation [95]; a similar observation was made in potato, where reduced ZEP expression resulted into the accumulation of zeaxanthin [96]. In maize, specific ZEP alleles have been identified as reliable predictors of total carotenoid content, highlighting their crucial role in this process [97]. According to our results, the increased expression of the ZEP gene we found in citron and lemon could be related with the lower carotenoid accumulation previously reported in the flavedo of these species [35].
We also found significant alterations in the expression of genes coding for carotenoid cleavage dioxygenases, including CCD4b, that has been postulated as the major enzyme involved in the production of the predominant red carotenoids in mandarins and oranges by cleaving β-carotene, β-cryptoxanthin and zeaxanthin into C30-apocarotenoids [98, 99]. Among our samples, CCD4b was significantly overexpressed in some but not all the red-colored flavedo samples. Commercial mandarin presented a low CCD4b expression, with levels comparable to those of the yellow citron and pummelo, while its expression in the lemon flavedo was similar to that of red fruits. CCD4a, a paralog of CCD4b [99], was differentially expressed among red and yellow species: it was hardly expressed in mandarins, displayed an intermediate expression in oranges and was highly expressed in the yellow fruits from pummelo, citron and lemon. Moreover, the ASE analysis performed in the present work showed that, in the flavedo of the admixed cultivars sweet and sour orange, the pummelo allele was expressed significantly higher than the mandarin one (over 90% of the expression came from the pummelo allele in both cases), while in lemon both citron and pummelo alleles were expressed at comparable levels. This may explain the intermediate expression level in sweet and sour orange, which might be produced by the combination of a highly expressed pummelo allele and an almost non-expressed mandarin allele. CCD4a has been considerably less studied in Citrus since its expression in mandarin and orange peel is low [98], but recent studies have reported that it is actually expressed in the flavedo of yellow fruits [46], might be related this expression variability. It is well known that gene family expansion paves the way for neofunctionalization by providing with extra copies of genes belonging to the family, therefore allowing for a better adaptation to new environments [123].
It is interesting to note that the FOMT gene family is specifically expanded in mandarins [46], which have gone through an intensive domestication process. The increased CHSm expression we report could in principle generate greater amounts of flavonoid precursors, which could be further modified by the expanded FOMT family to produce a broader range of compounds. Considering that the FOMT expansion is by far the largest in mandarins [72], it appears plausible that the mandarin CHSm allele activity, by generating a higher abundance of substrate, could be one of the factors leading to the expansion of the FOMT families, possibly during the mandarin early domestication or sometime in between the evolutionary and domestication processes.
Conclusions
In this work we have performed a genus wide RNA-seq analysis in ripening fruits from wild and domesticated citrus species, in order to assess the impact on gene expression of the complex evolutionary and domestication histories of the commercial citrus varieties in gene expression during ripening. We report for the first time the effects of the segmental ancestry of specific citrus species in the expression patterns of the genes contained within, highlighting the importance of introgressions during the early domestication of the genus Citrus. The broad scope of our work provides an evolutionary viewpoint on Citrus ripening transcriptomics, a perspective of special interest in a genus where most commercial cultivars are the product of interspecific hybridizations.
Methods
Plant material
Plant material for the RNA-seq analysis was provided by the germplasm resources at the Instituto Valenciano de Investigaciones Agrarias (IVIA). Three pure species, Sun Chu Sha Kat mandarin (Citrus reticulata), Chandler pummelo (Citrus maxima) and Diamante citron (Citrus medica), and four admixtures, Seville sour orange (Citrus aurantium), Salustiana sweet orange (Citrus sinensis), Willowleaf mandarin (Citrus deliciosa) and Eureka lemon (Citrus limon) were analyzed.
Sun Chu Sha Kat is a type 1 mandarin that produces small acidic fruits with bright red peels. Chandler pummelo produces large round fruits with pale yellow peel, pink pulp and sweet taste. Diamante citron also produces large fruits, but with an ovate shape, and its pulp is yellow and much more acidic. These three cultivar names are abbreviated as SCM, CHP and DIA in the figures of this manuscript.
The Seville sour orange is a direct hybrid between a mandarin and a pummelo [2], displaying the two complete parental haplotypes. Sour orange fruits are extremely acidic, with bright red peel and medium size. Salustiana sweet orange produces medium-sized round fruits with low acidity and a red rind; it is a pummelo/mandarin admixture displaying M/M, P/M and P/P regions in its genome, corresponding to 14, 83 and 3% of the total genomic space respectively [2]. Willowleaf mandarin is a type 2 mandarin, which produces edible pale-red fruits of intermediate size. As a type 2 mandarin, it has pummelo introgressions, in this case accounting for roughly 9% of the genome, while the rest displays two mandarin haplotypes [2]. The Eureka lemon is a tri-specific hybrid in which 55% of its genome has citron/mandarin ancestry while the other 45% is citron/pummelo [1]. It produces medium-sized fruits with a yellow peel and extremely acidic pulp. These cultivars have been respectively abbreviated as SSO, SWO, WLM and EUR in the figures.
Fruit samples of Ichang papeda (Citrus ichangensis), the Nagami kumquat (Fortunella margarita) and Poncirus trifoliata were also collected from the IVIA germplasm bank. Citrus ichangensis produces small inedible fruits containing acrid oil droplets in their pulp. Kumquat fruits are very small, with a red rind and highly acidic. Finally, Poncirus produces medium yellow fruits, also inedible. The abbreviations used for these species are ICH, FOR and PON, respectively. Accession numbers of each cultivar and species used in this work are shown in Table S7.
Phenotypical data collection
Fresh fruit samples were collected every three weeks, from mid-September to January. Peel color was measured on field using a hand colorimeter Konica Minolta CR400. For each sample analyzed, color was measured in four different fruits performing three technical replicates on each. Fruits were then collected and processed the same day.
Fruits were squeezed and the titratable acid content of the juice was measured by titration with 0.1 M sodium hydroxide and a phenolphthalein indicator. Juice total soluble sugar content was measured in Brix degrees using a table refractometer ATAGO PR-1. Brix and acidity were analyzed on the pooled juice performing three technical replicates.
RNA extraction, library preparation and sequencing
For each species, three biological replicates were collected at different dates to match each species color break, in order to collect samples in a comparable physiological state. Flavedo and pulp tissues were manually separated from each fruit and treated independently. Tissues were grinded frozen and total RNA was extracted using the acid phenol extraction coupled with lithium chloride precipitation as described in Ecker 1987 [124]. RNA-seq library preparation and sequencing were carried out by Novogene Company. Briefly, RNA samples were enriched in mRNA using oligo (dT) beads and the mRNA was randomly fragmented. cDNAs were then synthesized from mRNA using random hexamers, followed by adapter ligation, size selection and PCR enrichment. Samples were sequenced in a NovaSeq 6000 platform, delivering 150 bp pair ended reads with an insert size of approximately 250 bp.
RNA-seq read map** and DEG analysis
Illumina reads were mapped against the Citrus clementina reference genome [2] using STAR 2.7.2 [125]. C. clementina genome annotation was downloaded from the NCBI and reads mapped to each genomic feature were counted using featureCounts 2.0 [126]. Read counts were normalized using a variance stabilizing transformation implemented in R [127]; these pseudocounts were used for the sample clustering for quality control and downstream analysis. Differential gene expression analyses were performed using the R package DESeq2 1.26 [128] following the author’s recommendations. Pulp and flavedo data were analyzed independently, performing pairwise comparisons among every species pair, as well as pairwise comparisons of grouped samples of citron and lemon against the rest and citron, lemon and pummelo against the rest. Differentially expressed genes (DEG, log2 fold change expression > 1, false sign or smaller rate < 0.01) were detected using the model implemented in apeglm [129]. Genes annotated into admixed regions [1] were used to assess the admixture effect in gene expression.
KEGG enrichment analysis
A GO enrichment analysis was carried out for the comparison of citron and lemon pulp against the other analyzed samples. GO enrichment was performed using the R package clusterProfiler, KEGG data was accessed using AnnotationHub [130, 131]. To account for multiple hypothesis testing, p-values were corrected using the Benjamini-Hochberg method (FDR < 0.05).
Confirmation of RNA-seq data by RT-qPCR
To validate the RNA-seq analyses and to assess the CHSm expression in other species, one-step RT-qPCR of a set of genes was carried out. Reverse transcription was performed by incubating the RNA samples with the reverse transcriptase MultiScribe (Invitrogen) at 48 °C, 30 min. RNAse activity was inhibited using RNAse Inhibitor (Applied Biosystems). Real-time qPCR was performed using the LightCycler FastStart DNA Master Plus SYBR Green I kit in a LightCycler 2.0 Instrument. Two technical replicates were performed for each reaction. Amplification specificity was verified by the presence of a single peak in the melting curve analysis. Oligonucleotides used for each reaction can be found in the Table S6.
Relative quantification of the gene expression was expressed as a log 2-fold change expression compared with a housekee** gene, CitUBC1 [132], using the ΔΔCt method.
DNA extraction, sequencing and map**
To find diagnostic SNPs and further validate the genomic structure of the studied genes, whole genome sequencing data was used. For the already published data, raw reads were retrieved from the Sequence Read Archive database (the SRA accession numbers are available in Table S7). The Diamante citron genome was sequenced in this work using Illumina whole genome sequencing. In short, high molecular weight DNA was extracted using an in-house protocol. Whole genome sequencing library preparation and sequencing were carried out by the Centro Nacional de Análisis Genómico (CNAG). Briefly, libraries were constructed using the Illumina TruSeq DNA Sample Prep protocol, selecting for an insert size of 500 bp. Paired-end sequencing was performed on a HiSeq 2000 instrument.
Allele-specific expression
Allele-specific expression for each gene was studied with the following workflow: first, the two phases of every gene were established based on DNA sequencing using diagnostic heterozygous SNPs. Then, the mapped RNA-seq reads were scanned and those displaying different alleles of each diagnostic SNP were counted independently, which allowed the expression quantification at the allele level.
To achieve this, the Illumina genomic DNA reads (accession numbers available at Table S7) were mapped to the clementine reference genome using bwa mem [Analysis of the chalcone synthase promoter region The upstream region of a chalcone synthase (CHSm, LOC18042808) was extracted from the C. clementina reference genome based on its genomic coordinates. The ortholog regions in the reference genomes of other Citrus species and related genera were obtained by similarity search using BLASTN 2.7.1 [136], comparing this sequence against publicly available assembled genomes of other Citrus and related species. The other genomes used correspond to citron (C. medica), pummelo (C. maxima), wild mandarin (C. reticulata), satsuma mandarin (Citrus unshiu), sweet orange (C. sinensis), two different lemon haplotypes (C. limon), Ichang papeda (C. ichangensis), the Hong Kong kumquat (F. hindsii) and, as an outgroup of the genus Citrus, P. trifoliata [2, 4, 15, 16, 52,53,54,55]. The ortholog sequence was verified in pure mandarin, citron and pummelo by conventional PCR using the forward and reverse primers described in Table S6. Each PCR product was sequenced using Sanger sequencing, and the sequences obtained were aligned using MAFFT [137]. The genomic structure of the region was also manually curated with the Integrative Genome Viewer (IGV) browser software [138] with the same DNA-based alignments used for allele phasing. After determining the sequence of the promoter region, the existence of species-specific promoter motifs was estimated using the online motif-detection tool NewPLACE [139].
Availability of data and materials
All data used and generated in this work is publicly available. RNA-seq raw reads were deposited under BioProject PRJNA785525, and the whole genome sequencing of Citrus medica was deposited under BioProject PRJNA784926. Prior resequencing data used in this work can be accessed under BioProjects PRJNA414519 and PRJEB5963. See Table S7 for more detailed information.
References
Wu GA, Terol J, Ibanez V, López-García A, Pérez-Román E, Borredá C, et al. Genomics of the origin and evolution of Citrus. Nature. 2018;554:311–6.
Wu GA, Prochnik S, Jenkins J, Salse J, Hellsten U, Murat F, et al. Sequencing of diverse mandarin, pummelo and orange genomes reveals complex history of admixture during citrus domestication. Nat Biotechnol. 2014;32:656–62.
Oueslati A, Salhi-Hannachi A, Luro F, Vignes H, Mournet P, Ollitrault P. Genoty** by sequencing reveals the interspecific C. maxima / C. reticulata admixture along the genomes of modern citrus varieties of mandarins, tangors, tangelos, orangelos and grapefruits. PLoS One. 2017;12:e0185618.
Guardo MD, Moretto M, Moser M, Catalano C, Troggio M, Deng Z, et al. The haplotype-resolved reference genome of lemon (Citrus Limon L. Burm f). Tree Genet Genomes. 2021;17:46.
Curk F, Ollitrault F, Garcia-Lor A, Luro F, Navarro L, Ollitrault P. Phylogenetic origin of limes and lemons revealed by cytoplasmic and nuclear markers. Ann Bot. 2016;117:565–83.
Curk F, Ancillo G, Garcia-Lor A, Luro F, Perrier X, Jacquemoud-Collet J-P, et al. Next generation haploty** to decipher nuclear genomic interspecific admixture in Citrus species: analysis of chromosome 2. BMC Genet. 2014;15:152.
Olsen KM, Wendel JF. A Bountiful harvest: genomic insights into crop domestication phenotypes. Annu Rev Plant Biol. 2013;64:47–70.
Koenig D, Jiménez-Gómez JM, Kimura S, Fulop D, Chitwood DH, Headland LR, et al. Comparative transcriptomics reveals patterns of selection in domesticated and wild tomato. Proc Natl Acad Sci U S A. 2013;110:E2655–62.
Sauvage C, Rau A, Aichholz C, Chadoeuf J, Sarah G, Ruiz M, et al. Domestication rewired gene expression and nucleotide diversity patterns in tomato. Plant J Cell Mol Biol. 2017;91:631–45.
Liu D, Yang L, Zhang J, Zhu G, Lü H, Lü Y, et al. Domestication and breeding changed tomato fruit transcriptome. J Integr Agric. 2020;19:120–32.
Swanson-Wagner R, Briskine R, Schaefer R, Hufford MB, Ross-Ibarra J, Myers CL, et al. Resha** of the maize transcriptome by domestication. Proc Natl Acad Sci. 2012;109:11878–83.
Li X, Liu L, Ming M, Hu H, Zhang M, Fan J, et al. Comparative Transcriptomic analysis provides insight into the domestication and improvement of pear (P. pyrifolia) fruit. Plant Physiol. 2019;180:435–52.
Gros-Balthazard M, Besnard G, Sarah G, Holtz Y, Leclercq J, Santoni S, et al. Evolutionary transcriptomics reveals the origins of olives and the genomic changes associated with their domestication. Plant J. 2019;100:143–57.
He L, Zhao H, He J, Yang Z, Guan B, Chen K, et al. Extraordinarily conserved chromosomal synteny of Citrus species revealed by chromosome-specific painting. Plant J. 2020;103:2225–35.
Shimizu T, Tanizawa Y, Mochizuki T, Nagasaki H, Yoshioka T, Toyoda A, et al. Draft sequencing of the heterozygous diploid genome of Satsuma (Citrus unshiu Marc.) using a hybrid assembly approach. Front Genet. 2017;8:180.
Peng Z, Bredeson JV, Wu GA, Shu S, Rawat N, Du D, et al. A chromosome-scale reference genome of trifoliate orange (Poncirus trifoliata) provides insights into disease resistance, cold tolerance and genome evolution in Citrus. Plant J. 2020;104:1215–32.
Lu X, Cao X, Li F, Li J, **ong J, Long G, et al. Comparative transcriptome analysis reveals a global insight into molecular processes regulating citrate accumulation in sweet orange (Citrus sinensis). Physiol Plant. 2016;158:463–82.
Lu S, Zhang Y, Zhu K, Yang W, Ye J, Chai L, et al. The citrus transcription factor CsMADS6 modulates carotenoid metabolism by directly regulating carotenogenic genes. Plant Physiol. 2018;176:2657–76.
Wang S, Yang C, Tu H, Zhou J, Liu X, Cheng Y, et al. Characterization and metabolic diversity of flavonoids in Citrus species. Sci Rep. 2017;7:10549.
Da Conceicao Neta ER, Johanningsmeier SD, McFeeters RF, Neta ERDC, Johanningsmeier SD, McFeeters RF. The chemistry and physiology of sour taste - a review. J Food Sci. 2007;72:33–8.
Hussain SB, Shi CY, Guo LX, Kamran HM, Sadka A, Liu YZ. Recent advances in the regulation of citric acid metabolism in Citrus fruit. Crit Rev Plant Sci. 2017;36:241–56.
Strazzer P, Spelt CE, Li S, Bliek M, Federici CT, Roose ML, et al. Hyperacidification of Citrus fruits by a vacuolar proton-pum** P-ATPase complex. Nat Commun. 2019;10:1–11.
Müller ML, Irkens-Kiesecker U, Rubinstein B, Taiz L. On the mechanism of hyperacidification in lemon: comparison of the vacuolar H+-ATPase activities of fruits and epicotyls. J Biol Chem. 1996;271:1916–24.
Brune A, Gonzalez P, Goren R, Zehavi U, Echeverria E. Citrate uptake into tonoplast vesicles from acid lime (citrus aurantifolia) juice cells. J Membr Biol. 1998;166:197–203.
Shimada T, Nakano R, Shulaev V, Sadka A, Blumwald E. Vacuolar citrate/H+ symporter of citrus juice cells. Planta. 2006;224:472–80.
Chen M, **e X, Lin Q, Chen J, Grierson D, Yin X, et al. Differential expression of organic acid degradation-related genes during fruit development of navel oranges (Citrus sinensis) in two habitats. Plant Mol Biol Report. 2013;31:1131–40.
Cercós M, Soler G, Iglesias DJ, Gadea J, Forment J, Talón M. Global analysis of gene expression during development and ripening of Citrus fruit flesh. A proposed mechanism for citric acid utilization. Plant Mol Biol. 2006;62:513–27.
Hu X-M, Shi C-Y, Liu X, ** L-F, Liu Y-Z, Peng S-A. Genome-wide identification of citrus ATP-citrate lyase genes and their transcript analysis in fruits reveals their possible role in citrate utilization. Mol Gen Genomics. 2015;290:29–38.
Guo LX, Shi CY, Liu X, Ning DY, **g LF, Yang H, et al. Citrate accumulation-related gene expression and/or enzyme activity analysis combined with metabolomics provide a novel insight for an orange mutant. Sci Rep. 2016;6:29343.
Tadeo FR, Cercós M, Colmenero-Flores JM, Iglesias DJ, Naranjo MA, Ríos G, et al. Molecular physiology of development and quality of Citrus. Adv Bot Res. 2008;47:147–223.
Komatsu A, Moriguchi T, Koyama K, Omura M, Akihama T. Analysis of sucrose synthase genes in citrus suggests different roles and phylogenetic relationships. J Exp Bot. 2002;53:61–71.
Katz E, Hwan Boo K, Youn Kim H, Eigenheer RA, Phinney BS, Shulaev V, et al. Label-free shotgun proteomics and metabolite analysis reveal a significant metabolic shift during citrus fruit development. J Exp Bot. 2011;62:5367–84.
Hussain SB, Guo L-X, Shi C-Y, Khan MA, Bai Y-X, Du W, et al. Assessment of sugar and sugar accumulation-related gene expression profiles reveal new insight into the formation of low sugar accumulation trait in a sweet orange (Citrus sinensis) bud mutant. Mol Biol Rep. 2020;47:2781–91.
Jacob-Wilk D, Holland D, Goldschmidt EE, Riov J, Eyal Y. Chlorophyll breakdown by chlorophyllase: isolation and functional expression of the Chlase1 gene from ethylene-treated Citrus fruit and its regulation during development. Plant J. 1999;20:653–61.
Kato M, Ikoma Y, Matsumoto H, Sugiura M, Hyodo H, Yano M. Accumulation of carotenoids and expression of carotenoid biosynthetic genes during maturation in Citrus fruit. Plant Physiol. 2004;134:824–37.
Zhang L, Ma G, Shirai Y, Kato M, Yamawaki K, Ikoma Y, et al. Expression and functional analysis of two lycopene β-cyclases from citrus fruits. Planta. 2012;236:1315–25.
Rodrigo MJ, Alquézar B, Alós E, Lado J, Zacarías L. Biochemical bases and molecular regulation of pigmentation in the peel of Citrus fruit. Sci Hortic. 2013;163:46–62.
Oberholster R, Cowan AK, Molnár P, Gy T. Biochemical basis of color as an aesthetic quality in Citrus sinensis. J Agric Food Chem. 2001;49:303–7.
Xu C-J, Fraser PD, Wang W-J, Bramley PM. Differences in the carotenoid content of ordinary Citrus and lycopene-accumulating mutants. J Agric Food Chem. 2006;54:5474–81.
Wang Y-C, Chuang Y-C, Hsu H-W. The flavonoid, carotenoid and pectin content in peels of citrus cultivated in Taiwan. Food Chem. 2008;106:277–84.
Ikoma Y, Matsumoto H, Kato M. The characteristics of carotenoid biosynthesis in Citrus fruit. Jpn Agric Res Q JARQ. 2014;48:9–16.
Lu PJ, Wang CY, Yin TT, Zhong SL, Grierson D, Chen KS, et al. Cytological and molecular characterization of carotenoid accumulation in normal and high-lycopene mutant oranges. Sci Rep. 2017;7:761.
Lux PE, Carle R, Zacarías L, Rodrigo MJ, Schweiggert RM, Steingass CB. Genuine carotenoid profiles in sweet Orange [ Citrus sinensis (L.) Osbeck cv. Navel] Peel and pulp at different maturity stages. J Agric Food Chem. 2019;67:13164–75.
Cirmi S, Maugeri A, Ferlazzo N, Gangemi S, Calapai G, Schumacher U, et al. Anticancer potential of Citrus juices and their extracts: A systematic review of both preclinical and clinical studies. Front Pharmacol. 2017;8:420.
Zhao Z, He S, Hu Y, Yang Y, Jiao B, Fang Q, et al. Fruit flavonoid variation between and within four cultivated Citrus species evaluated by UPLC-PDA system. Sci Hortic. 2017;224:93–101.
Gonzalez-Ibeas D, Ibanez V, Perez-Roman E, Borredá C, Terol J, Talon M. Sha** the biology of citrus: I. Genomic determinants of evolution. Plant Genome. 2021;14:e20104.
Huang D, Zhao Y, Cao M, Qiao L, Zheng Z-L. Integrated systems biology analysis of Transcriptomes reveals candidate genes for acidity control in develo** fruits of sweet Orange (Citrus sinensis L. Osbeck). Front Plant Sci. 2016;7:486.
Hui WJ, Jun LJ, Ling CK, Wen LH, He J, Guan B, et al. Comparative transcriptome and proteome profiling of two Citrus sinensis cultivars during fruit development and ripening. BMC Genomics. 2017;18:1–13.
Bastías A, Yañez M, Osorio S, Arbona V, Gómez-Cadenas A, Fernie AR, et al. The transcription factor AREB1 regulates primary metabolic pathways in tomato fruits. J Exp Bot. 2014;65:2351–63.
Ma Q-J, Sun M-H, Lu J, Liu Y-J, Hu D-G, Hao Y-J. Transcription factor AREB2 is involved in soluble sugar accumulation by activating sugar transporter and amylase genes. Plant Physiol. 2017;174:2348–62.
Liu J, Chen X, Wang S, Wang Y, Ouyang Y, Yao Y, et al. MeABL5, an ABA insensitive 5-like basic Leucine zipper transcription factor, positively regulates MeCWINV3 in cassava (Manihot esculenta Crantz). Front Plant Sci. 2019;10:772.
Xu Q, Chen LL, Ruan X, Chen D, Zhu A, Chen C, et al. The draft genome of sweet orange (Citrus sinensis). Nat Genet. 2013;45:59–66.
Wang X, Xu Y, Zhang S, Cao L, Huang Y, Cheng J, et al. Genomic analyses of primitive, wild and cultivated citrus provide insights into asexual reproduction. Nat Genet. 2017;49:765–72.
Wang L, He F, Huang Y, He J, Yang S, Zeng J, et al. Genome of wild mandarin and domestication history of mandarin. Mol Plant. 2018;11:1024–37.
Zhu C, Zheng X, Huang Y, Ye J, Chen P, Zhang C, et al. Genome sequencing and CRISPR/Cas9 gene editing of an early flowering Mini-Citrus (Fortunella hindsii). Plant Biotechnol J. 2019;17:2199–210.
Barrett HC, Rhodes AM. A numerical taxonomic study of affinity relationships in cultivated Citrus and its close relatives. Syst Bot. 1976;1:105–36.
Dickinson HG, Hiscock SJ, Crane PR, Rieseberg LH, Widmer A, Arntz AM, et al. The genetic architecture necessary for transgressive segregation is common in both natural and domesticated populations. Philos Trans R Soc Lond Ser B Biol Sci. 2003;358:1141–7.
Mao D, Liu T, Xu C, Li X, **ng Y. Epistasis and complementary gene action adequately account for the genetic bases of transgressive segregation of kilo-grain weight in rice. Euphytica. 2011;180:261–71.
Kitazumi A, Pabuayon ICM, Ohyanagi H, Fujita M, Osti B, Shenton MR, et al. Potential of Oryza officinalis to augment the cold tolerance genetic mechanisms of Oryza sativa by network complementation. Sci Rep. 2018;8:16346.
Shao L, **ng F, Xu C, Zhang Q, Che J, Wang X, et al. Patterns of genome-wide allele-specific expression in hybrid rice and the implications on the genetic basis of heterosis. Proc Natl Acad Sci. 2019;116:5653–8.
Cai M, Lin J, Li Z, Lin Z, Ma Y, Wang Y, et al. Allele specific expression of Dof genes responding to hormones and abiotic stresses in sugarcane. PLoS One. 2020;15:e0227716.
Botet R, Keurentjes JJB. The role of transcriptional regulation in hybrid vigor. Front Plant Sci. 2020;11:410.
Albert E, Duboscq R, Latreille M, Santoni S, Beukers M, Bouchet J-P, et al. Allele-specific expression and genetic determinants of transcriptomic variations in response to mild water deficit in tomato. Plant J. 2018;96:635–50.
Springer NM, Stupar RM. Allele-specific expression patterns reveal biases and embryo-specific parent-of-origin effects in hybrid maize. Plant Cell. 2007;19:2391–402.
Zinati Z, Sazegari S, Amin H, Tahmasebi A. Mining transcriptome data to identify genes and pathways related to lemon taste using supervised and unsupervised data learning methods. Hortic Environ Biotechnol. 2021;62:593–603.
Minamikawa MF, Nonaka K, Kaminuma E, Kajiya-Kanegae H, Onogi A, Goto S, et al. Genome-wide association study and genomic prediction in citrus: potential of genomics-assisted breeding for fruit quality traits. Sci Rep. 2017;7:4721.
Frydman A, Liberman R, Huhman DV, Carmeli-Weissberg M, Sapir-Mir M, Ophir R, et al. The molecular and enzymatic basis of bitter/non-bitter flavor of citrus fruit: evolution of branch-forming rhamnosyltransferases under domestication. Plant J. 2013;73:166–78.
Chen J, Yuan Z, Zhang H, Li W, Shi M, Peng Z, et al. Cit1,2RhaT and two novel CitdGlcTs participate in flavor-related flavonoid metabolism during citrus fruit development. J Exp Bot. 2019;70:2759–71.
Hewitt SL, Hendrickson CA, Dhingra A. Evidence for the involvement of Vernalization-related genes in the regulation of cold-induced ripening in ‘D’Anjou’ and ‘Bartlett’ pear fruit. Sci Rep. 2020;10:8478.
Ahmad B, Zhang S, Yao J, Rahman MU, Hanif M, Zhu Y, et al. Genomic organization of the B3-domain transcription factor family in grapevine (Vitis vinifera L.) and expression during seed development in seedless and seeded cultivars. Int J Mol Sci. 2019;20:4553.
Liu Z, Ge X-X, Wu X-M, Xu Q, Atkinson RG, Guo W-W. Genome-wide analysis of the citrus B3 superfamily and their association with somatic embryogenesis. BMC Genomics. 2020;21:305.
Gonzalez-Ibeas D, Ibanez V, Perez-Roman E, Borredá C, Terol J, Talon M. Sha** the biology of citrus: II. Genomic determinants of domestication. Plant Genome. 2021;14:e20133.
Shelp BJ, Mullen RT, Waller JC. Compartmentation of GABA metabolism raises intriguing questions. Trends Plant Sci. 2012;17:57–9.
Myles S, Boyko AR, Owens CL, Brown PJ, Grassi F, Aradhya MK, et al. Genetic structure and domestication history of the grape. Proc Natl Acad Sci. 2011;108:3530–5.
Duan N, Bai Y, Sun H, Wang N, Ma Y, Li M, et al. Genome re-sequencing reveals the history of apple and supports a two-stage model for fruit enlargement. Nat Commun. 2017;8:249.
Julca I, Marcet-Houben M, Cruz F, Gómez-Garrido J, Gaut BS, Díez CM, et al. Genomic evidence for recurrent genetic admixture during the domestication of Mediterranean olive trees (Olea europaea L.). BMC Biol. 2020;18:148.
Brandenburg J-T, Mary-Huard T, Rigaill G, Hearne SJ, Corti H, Joets J, et al. Independent introductions and admixtures have contributed to adaptation of European maize and its American counterparts. PLoS Genet. 2017;13:e1006666.
Sadka A, Shlizerman L, Kamara I, Blumwald E. Primary metabolism in Citrus fruit as affected by its unique structure. Front Plant Sci. 2019;10:1167.
Ntoukakis VE, Negm FB, Lovatt CJ. Sink activity in Washington navel orange fruit borne on leafy and leafless inflorescences. Citrus Res Technol. 2017;38:1.
Li M, Li D, Feng F, Zhang S, Ma F, Cheng L. Proteomic analysis reveals dynamic regulation of fruit development and sugar and acid accumulation in apple. J Exp Bot. 2016;67:5145–57.
Zhao B, Qi K, Yi X, Chen G, Liu X, Qi X, et al. Identification of hexokinase family members in pear (Pyrus × bretschneideri) and functional exploration of PbHXK1 in modulating sugar content and plant growth. Gene. 2019;711:143932.
Shi C-Y, Song R-Q, Hu X-M, Liu X, ** L-F, Liu Y-Z. Citrus PH5-like H+-ATPase genes: identification and transcript analysis to investigate their possible relationship with citrate accumulation in fruits. Front Plant Sci. 2015;6:135.
Shi C-Y, Hussain SB, Guo L-X, Yang H, Ning D-Y, Liu Y-Z. Genome-wide identification and transcript analysis of vacuolar-ATPase genes in citrus reveal their possible involvement in citrate accumulation. Phytochemistry. 2018;155:147–54.
Li S, Liu X, **e X, Sun C, Grierson D, Yin X, et al. CrMYB73, a PH-like gene, contributes to citric acid accumulation in citrus fruit. Sci Hortic. 2015;197:212–7.
Tao N, Hu Z, Liu Q, Xu J, Cheng Y, Guo L, et al. Expression of phytoene synthase gene (Psy) is enhanced during fruit ripening of Cara Cara navel orange (Citrus sinensis Osbeck). Plant Cell Rep. 2007;26:837–43.
Zhang J, Tao N, Xu Q, Zhou W, Cao H, Xu J, et al. Functional characterization of Citrus PSY gene in Hongkong kumquat (Fortunella hindsii Swingle). Plant Cell Rep. 2009;28:1737.
Alquézar B, Rodrigo MJ, Zacarías L. Carotenoid biosynthesis and their regulation in Citrus fruits. Tree For Sci Biotechnol. 2008;2:23–35.
Zhu K, Wu Q, Huang Y, Ye J, Xu Q, Deng X. Genome-wide characterization of cis-acting elements in the promoters of key carotenoid pathway genes from the Main species of genus Citrus. Hortic Plant J. 2020;6:385–95.
McQuinn RP, Gapper NE, Gray AG, Zhong S, Tohge T, Fei Z, et al. Manipulation of ZDS in tomato exposes carotenoid- and ABA-specific effects on fruit development and ripening. Plant Biotechnol J. 2020;18(11):2210–24.
Flores-Ortiz C, Alvarez LM, Undurraga A, Arias D, Durán F, Wegener G, et al. Differential role of the two ζ-carotene desaturase paralogs in carrot (Daucus carota): ZDS1 is a functional gene essential for plant development and carotenoid synthesis. Plant Sci Int J Exp Plant Biol. 2020;291:110327.
Rodrigo M-J, Marcos JF, Alférez F, Mallent MD, Zacarías L. Characterization of Pinalate, a novel Citrus sinensis mutant with a fruit-specific alteration that results in yellow pigmentation and decreased ABA content. J Exp Bot. 2003;54:727–38.
Rodrigo MJ, Lado J, Alós E, Alquézar B, Dery O, Hirschberg J, et al. A mutant allele of ζ-carotene isomerase (Z-ISO) is associated with the yellow pigmentation of the “Pinalate” sweet orange mutant and reveals new insights into its role in fruit carotenogenesis. BMC Plant Biol. 2019;19:465.
Terol J, Nueda MJ, Ventimilla D, Tadeo F, Talon M. Transcriptomic analysis of Citrus Clementina mandarin fruits maturation reveals a MADS-box transcription factor that might be involved in the regulation of earliness. BMC Plant Biol. 2019;19:47.
Moreno JC, Pizarro L, Fuentes P, Handford M, Cifuentes V, Stange C. Levels of lycopene β-Cyclase 1 modulate carotenoid gene expression and accumulation in Daucus carota. PLoS One. 2013;8:e58144.
Gonzalez-Jorge S, Mehrshahi P, Magallanes-Lundback M, Lipka AE, Angelovici R, Gore MA, et al. ZEAXANTHIN EPOXIDASE activity potentiates carotenoid degradation in maturing Seed1[OPEN]. Plant Physiol. 2016;171:1837–51.
Wolters A-MA, Uitdewilligen JGAML, Kloosterman BA, Hutten RCB, Visser RGF, van Eck HJ. Identification of alleles of carotenoid pathway genes important for zeaxanthin accumulation in potato tubers. Plant Mol Biol. 2010;73:659–71.
Owens BF, Lipka AE, Magallanes-Lundback M, Tiede T, Diepenbrock CH, Kandianis CB, et al. A Foundation for Provitamin a biofortification of maize: genome-wide association and genomic prediction models of carotenoid levels. Genetics. 2014;198:1699–716.
Rodrigo MJ, Alquézar B, Alós E, Medina V, Carmona L, Bruno M, et al. A novel carotenoid cleavage activity involved in the biosynthesis of Citrus fruit-specific apocarotenoid pigments. J Exp Bot. 2013;64:4461–78.
Zheng X, Zhu K, Sun Q, Zhang W, Wang X, Cao H, et al. Natural variation in CCD4 promoter underpins species-specific evolution of red coloration in Citrus Peel. Mol Plant. 2019;12:1294–307.
Zheng X, **e Z, Zhu K, Xu Q, Deng X, Pan Z. Isolation and characterization of carotenoid cleavage dioxygenase 4 genes from different citrus species. Mol Gen Genomics. 2015;290:1589–603.
Yoshioka S, Aida R, Yamamizo C, Shibata M, Ohmiya A. The carotenoidcleavagedioxygenase4(CmCCD4a) gene family encodes a key regulator of petal color mutation in chrysanthemum. Euphytica. 2012;184:377–87.
Kishimoto S, Oda-Yamamizo C, Ohmiya A. Regulation of carotenoid pigmentation in corollas of Petunia. Plant Mol Biol Report. 2018;36:632–42.
Phadungsawat B, Watanabe K, Mizuno S, Kanekatsu M, Suzuki S. Expression of CCD4 gene involved in carotenoid degradation in yellow-flowered Petunia × hybrida. Sci Hortic. 2020;261:108916.
Baldermann S, Kato M, Kurosawa M, Kurobayashi Y, Fujita A, Fleischmann P, et al. Functional characterization of a carotenoid cleavage dioxygenase 1 and its relation to the carotenoid accumulation and volatile emission during the floral development of Osmanthus fragrans Lour. J Exp Bot. 2010;61:2967–77.
Liu Q, Xu J, Liu Y, Zhao X, Deng X, Guo L, et al. A novel bud mutation that confers abnormal patterns of lycopene accumulation in sweet orange fruit (Citrus sinensis L. Osbeck). J Exp Bot. 2007;58:4161–71.
Alquézar B, Rodrigo MJ, Zacarías L. Regulation of carotenoid biosynthesis during fruit maturation in the red-fleshed orange mutant Cara Cara. Phytochemistry. 2008;69:1997–2007.
Alós E, Roca M, Iglesias DJ, Mínguez-Mosquera MI, Damasceno CMB, Thannhauser TW, et al. An evaluation of the basis and consequences of a stay-green mutation in the navel negra Citrus mutant using Transcriptomic and proteomic profiling and metabolite analysis. Plant Physiol. 2008;147:1300–15.
Lana G, Zacarias-Garcia J, Distefano G, Gentile A, Rodrigo MJ, Zacarias L. Transcriptional analysis of carotenoids accumulation and metabolism in a pink-fleshed lemon mutant. Genes. 2020;11:1294.
Chen Q, Wang D, Tan C, Hu Y, Sundararajan B, Zhou Z. Profiling of flavonoid and antioxidant activity of fruit tissues from 27 Chinese local Citrus cultivars. Plants. 2020;9(2):196.
Nogata Y, Sakamoto K, Shiratsuchi H, Ishii T, Yano M, Ohta H. Flavonoid composition of fruit tissues of Citrus species. Biosci Biotechnol Biochem. 2006;70:178–92.
Dao TTH, Linthorst HJM, Verpoorte R. Chalcone synthase and its functions in plant resistance. Phytochem Rev. 2011;10:397–412.
Chaudhary PR, Bang H, Jayaprakasha GK, Patil BS. Variation in key flavonoid biosynthetic enzymes and phytochemicals in ‘Rio red’ grapefruit (Citrus paradisi Macf.) during fruit development. J Agric Food Chem. 2016;64:9022–32.
Wang Z, Yu Q, Shen W, El Mohtar CA, Zhao X, Gmitter FG. Functional study of CHS gene family members in citrus revealed a novel CHS gene affecting the production of flavonoids. BMC Plant Biol. 2018;18:189.
Tai D, Tian J, Zhang J, Song T, Yao Y. A Malus crabapple Chalcone synthase gene, McCHS, regulates red petal color and flavonoid biosynthesis. PLoS One. 2014;9:e110570.
Liu C, Long J, Zhu K, Liu L, Yang W, Zhang H, et al. Characterization of a Citrus R2R3-MYB transcription factor that regulates the Flavonol and Hydroxycinnamic acid biosynthesis. Sci Rep. 2016;6:25352.
Chen S, Zhang Y, Pan X, Li Y, Cui L, Wu F, et al. Isolation and expression analysis of NtCHS6, a new chalcone synthase gene from Nicotiana tabacum. J Integr Agric. 2017;16:1443–50.
Wu X, Zhang S, Liu X, Shang J, Zhang A, Zhu Z, et al. Chalcone synthase (CHS) family members analysis from eggplant (Solanum melongena L.) in the flavonoid biosynthetic pathway and expression patterns in response to heat stress. PLoS One. 2020;15:e0226537.
Yang L, Zhang JC, Qu JT, He G, Yu HQ, Li WC, et al. Expression response of chalcone synthase gene to inducing conditions and its effect on flavonoids accumulation in two medicinal species of Anoectochilus. Sci Rep. 2019;9:20171.
Ma M, Zhong M, Zhang Q, Zhao W, Wang M, Luo C. Phylogenetic implications and functional disparity in the Chalcone synthase gene family of common Seagrass Zostera marina. Front Mar Sci. 2021;8:1639.
Itoh N, Iwata C, Toda H. Molecular cloning and characterization of a flavonoid-O-methyltransferase with broad substrate specificity and regioselectivity from Citrus depressa. BMC Plant Biol. 2016;16:180.
Liu X, Wang Y, Chen Y, Xu S, Gong Q, Zhao C, et al. Characterization of a flavonoid 3′/5′/7-O-methyltransferase from Citrus reticulata and evaluation of the in vitro cytotoxicity of its methylated products. Molecules. 2020;25(4):858.
Liu X, Luo Y, Wu H, ** W, Yu J, Zhang Q, et al. Systematic analysis of O-methyltransferase gene family and identification of potential members involved in the formation of O-methylated flavonoids in Citrus. Gene. 2016;575:458–72.
Flagel LE, Wendel JF. Gene duplication and evolutionary novelty in plants. New Phytol. 2009;183:557–64.
Ecker JR, Davis RW. Plant defense genes are regulated by ethylene. Proc Natl Acad Sci U S A. 1987;84:5202–6.
Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29:15–21.
Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinforma Oxf Engl. 2014;30:923–30.
Anders S, Huber W. Differential expression analysis for sequence count data. Genome Biol. 2010;11:R106.
Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15:550.
Zhu A, Ibrahim JG, Love MI. Heavy-tailed prior distributions for sequence count data: removing the noise and preserving large differences. Bioinformatics. 2019;35:2084–92.
Yu G, Wang L-G, Han Y, He Q-Y. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics J Integr Biol. 2012;16:284–7.
Morgan M, Shepherd L. AnnotationHub: Client to access AnnotationHub resources. 2019; R package version 3.2.2. https://bioconductor.org/packages/release/bioc/html/AnnotationHub.html.
Merelo P, Agustí J, Arbona V, Costa ML, Estornell LH, Gómez-Cadenas A, et al. Cell Wall remodeling in abscission zone cells during ethylene-promoted fruit abscission in Citrus. Front Plant Sci. 2017;8:126.
Li H. Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM. Ar**v13033997 Q-Bio. 2013.
Van der Auwera GA, Carneiro MO, Hartl C, Poplin R, del Angel G, Levy-Moonshine A, et al. From FastQ data to high confidence variant calls: the genome analysis toolkit best practices pipeline. Curr Protoc Bioinformatics. 2013;43:11.10.1-11.10.33.
Lindenbaum P, Redon R. Bioalcidae, samjs and vcffilterjs: object-oriented formatters and filters for bioinformatics files. Bioinformatics. 2018;34:1224–5.
Camacho C, Coulouris G, Avagyan V, Ma N, Papadopoulos J, Bealer K, et al. BLAST+: architecture and applications. BMC Bioinformatics. 2009;10:421.
Katoh K, Standley DM. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol Evol. 2013;30:772–80.
Robinson JT, Thorvaldsdóttir H, Winckler W, Guttman M, Lander ES, Getz G, et al. Integrative genomics viewer. Nat Biotechnol. 2011;29:24–6.
Higo K, Ugawa Y, Iwamoto M, Korenaga T. Plant cis-acting regulatory DNA elements (PLACE) database: 1999. Nucleic Acids Res. 1999;27:297–300.
Acknowledgements
Not applicable.
Funding
Financial support was provided by the Ministerio de Ciencia, Innovación y Universidades (Spain) trough grant # RTI2018–097790-R-100 and #PROMETEO/2020/027. Funds were used for the design of the study and collection, analysis, and interpretation of data and in writing the manuscript, as well as in the open access payment.
Author information
Authors and Affiliations
Contributions
MT and JT contributed with the conception and design of this work. CB and EP collected and phenotyped the samples. CB performed the main bioinformatic analyses with contributions by JT and EP. JT conceptualized the allele-specific expression analyses. CB, MT and JT wrote and revised the manuscript. JT acquired the funding needed for this project. MT and JT supervised and coordinated the project. All authors have read and approved the final manuscript.
Corresponding author
Ethics declarations
Ethics approval and consent to participate
Study complied with local and national regulations for using plants.
Consent for publication
Not applicable.
Competing interests
The authors declare that they have no competing interests.
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Additional file 1: Figure S1.
Hierarchical clustering analysis. Hierarchical clustering of a) flavedo and b) pulp samples according to RNA-seq data. CHP: C. maxima, DIA: C. medica, EUR: C. limon, SCM: C. reticulata, SSO: C. aurantium, SWO: C. sinensis, WLM: C. deliciosa. Figure S2. Differentially expressed genes involved in carotenoid biosynthetic pathway. DEGs found in flavedo (a) and pulp (b) are shown independently. Each bar represents the expression log2 fold change comparing red samples (wild and domesticated mandarin, sweet and sour orange) against the yellow samples (lemon, citron and pummelo). Only genes with a log2 fold change > 1, s-value < 0.01, are shown. ZDS: zeta-carotene desaturase, LUT5: beta-ring hydroxylase, LCYb: beta-lycopene cyclase, CHYB: carotenoid beta-ring hydroxylase, CCD4a: carotenoid cleavage dioxygenase 4a, AAH: abscisic acid hydroxylase. Figure S3. Flavonoid-related gene expression across samples. Expression levels of genes involved in flavonoid modifications in flavedo tissues per sample and gene. Color intensity represent expression levels based on normalized read counts. Black rectangles mark differentially expressed genes between at least two samples; red rectangles denote flavonoid O-methyltransferases. CHP: C. maxima, DIA: C. medica, EUR: C. limon, SCM: C. reticulata, SSO: C. aurantium, SWO: C. sinensis, WLM: C. deliciosa. Figure S4. Distribution of DEGs between sweet orange and its progenitor species. The number of DEGs along the genome is represented in four different comparisons. Bar height represents the number of DEGs in a given windows, with color representing the admixture pattern of these genes (blue: admixed, orange: non-admixed mandarin, green: non-admixed pummelo). Bars with two colors represent genomic windows spanning two different admixed regions and include genes with both ancestries. A: Flavedo of sweet orange and wild mandarin. B: Pulp of sweet orange and wild mandarin. C: Flavedo of pummelo and sweet orange. D: Pulp of pummelo and sweet orange. E: admixture patterns of the sweet orange genome and distribution of runs of homozygosity (in black) and genic abundance (in red). Figure S5. Allele-specific expressed genes across tissues and cultivars. Total ASE genes preferentially expressing the citron, pummelo and mandarin in each sample, shown in green, orange and yellow, respectively. For each sample, regions with different ancestry were considered independently: M/P corresponds to mandarin/pummelo regions, C/P to citron/pummelo regions and C/M to citron/mandarin regions. The total number of ASE genes is shown on top of each bar. Flavedo (a) and pulp (b) samples are shown independently. CHP: C. maxima, DIA: C. medica, EUR: C. limon, SCM: C. reticulata, SSO: C. aurantium, SWO: C. sinensis, WLM: C. deliciosa. Figure S6. CHSm promoter sequence. Nucleotide sequence of 500 bp upstream from the gene start in 11 publicly available assembled genomes.
Additional file 2: Table S1.
Results of the RNA-seq validation via RT-qPCR. Table S2. NCBI locus identifiers for the major genes discussed in this work. Table S3. GO Enrichment Analysis on the differentially expressed genes between citron and lemon and the remaining samples. Table S4. Allele-specific expression genes in metabolic pathways associated with ripening. Table S5. Regulatory motifs of the CHSm promoter. Table S6. Primers used for the validation of RNA-seq data and amplifying the CHSm promoter. Table S7. Sample information and origin.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
About this article
Cite this article
Borredá, C., Perez-Roman, E., Talon, M. et al. Comparative transcriptomics of wild and commercial Citrus during early ripening reveals how domestication shaped fruit gene expression. BMC Plant Biol 22, 123 (2022). https://doi.org/10.1186/s12870-022-03509-9
Received:
Accepted:
Published:
DOI: https://doi.org/10.1186/s12870-022-03509-9