Molecular Analysis of 14-3-3 Genes in Citrus sinensis and Their Responses to Different Stresses

14-3-3 proteins (14-3-3s) are among the most important phosphorylated molecules playing crucial roles in regulating plant development and defense responses to environmental constraints. No report thus far has documented the gene family of 14-3-3s in Citrus sinensis and their roles in response to stresses. In this study, nine 14-3-3 genes, designated as CitGF14s (CitGF14a through CitGF14i) were identified from the latest C. sinensis genome. Phylogenetic analysis classified them into ε-like and non-ε groups, which were supported by gene structure analysis. The nine CitGF14s were located on five chromosomes, and none had duplication. Publicly available RNA-Seq raw data and microarray databases were mined for 14-3-3 expression profiles in different organs of citrus and in response to biotic and abiotic stresses. RT-qPCR was used for further examining spatial expression patterns of CitGF14s in citrus and their temporal expressions in one-year-old C. sinensis “Xuegan” plants after being exposed to different biotic and abiotic stresses. The nine CitGF14s were expressed in eight different organs with some isoforms displayed tissue-specific expression patterns. Six of the CitGF14s positively responded to citrus canker infection (Xanthomonas axonopodis pv. citri). The CitGF14s showed expressional divergence after phytohormone application and abiotic stress treatments, suggesting that 14-3-3 proteins are ubiquitous regulators in C. sinensis. Using the yeast two-hybrid assay, CitGF14a, b, c, d, g, and h were found to interact with CitGF14i proteins to form a heterodimer, while CitGF14i interacted with itself to form a homodimer. Further analysis of CitGF14s co-expression and potential interactors established a 14-3-3s protein interaction network. The established network identified 14-3-3 genes and several candidate clients which may play an important role in developmental regulation and stress responses in this important fruit crop. This is the first study of 14-3-3s in citrus, and the established network may help further investigation of the roles of 14-3-3s in response to abiotic and biotic constraints.


Introduction
Plants are constantly exposed to different abiotic and biotic stresses, including drought, extreme temperatures, high salinity, and various pathogens. Due to their sessile nature, plants have evolved a series of mechanisms to cope with the environmental challenges. Plant 14-3-3s, encoded by genes called general regulatory factors [1,2], regulate critical biochemical processes and sophisticated signaling networks in plants though proteinprotein interactions by binding to phosphorylated protein clients [3]. 14-3-3 proteins were originally isolated from mammalian brain tissue and were named according to their elution and migration pattern on DEAD-cellulose chromatography and starch-gel electrophoresis [4]. The 14-3-3s are highly conserved proteins and exist in all eukaryotes with multiple isoforms. Yeast has two genes encoding 14-3-3s [5], animals typically have seven [6], and plants have more 14-3-3 genes: 13 in Arabidopsis [6], 17 in tobacco (Nicotiana tabacum) [7], 12 in tomato (Solanum lycopersicum) [8], and eight in rice (Oryza sativa) [9]. The evolutionary relationships of 44 14-3-3s from Arabidopsis, rice, black cottonwood, and C. sinensis were phylogenetically analyzed. Eighteen of them were clustered into ε-like groups, and 26 were clustered into non-ε groups ( Figure 1). Four CitGF14s (CitGF14a, b, c, and d) were grouped into ε-like isoforms, and the remaining five CitGF14s were grouped into non-ε isoforms. None of the CitGF14s were duplicated. Int
Gene structure analysis showed that CitGR14s contained 3 to 6 exons, interspersed by highly distinct introns ( Figure 3). Four ε-like CitGF14s showed six conserved exons interrupted by intron and UTR in different lengths. Non-ε CitGF14 genes were also interrupted by introns. CitGF14e, CitGF14g, and CitGF14i carried three introns, CitGF14f and CitGF14h possessed four introns, the remaining had six introns. Figure 1. Phylogenetic trees of 14-3-3 family genes in Citrus sinensis (CitGF14s, indicated by the star symbol), Arabidopsis (AtGRF), Populus trichocarpa (PtGRF), and Oryza sativa (OsGRF) constructed using the MEGA6.0 program with neighbor-joining method within 1000 bootstrap replicates. The green and blue shade separates the ε and non-ε groups.   The graphic representation is displayed using GSDS (http://gsds.cbi.pku.edu.cn/). The unrooted phylogenetic tree was constructed using the amino sequences of CitGF14s genes by the Neighbor-Joining method with 1000 bootstrap replicates. Exons consisted of CDS shown as orange boxes, introns are shown as thin lines, and UTRs are shown as blue boxes.

CitGF14s Sequence Alignment
Amino acid sequence alignment showed that the deduced CitGF14s from C. sinensis were highly conserved with the exception of the N-terminal and C-terminal regions. Nine α-helices were shown in green rectangular boxes (Additional file 1: Figure S1A). Two conserved signature motifs RNL(L/V)SV(G/A)YKNV and YKDSTLIMQLLRDNLTLWTS were found in α-helix 3 and α-helix 9, respectively. Protein CitGF14b, CitGF14d, CitGF14e, CitGF14f, CitGF14g, CitGF14h, and CitGF14i had rather similar three-dimensional structures based on the Swiss-model prediction (Additional file 1: Figure S1B). All these similar proteins contained two potassium channel KAT1 ligands except for CitGF14b. The remaining two proteins (CitGF14a and CitGF14c) had relatively simple structures with no ligands.

Cis-Regulatory Elements
Different cis-acting elements related to plant growth, development, and stress responses were identified in CitGR14s (Additional file 2: Table S1). The circadian element was presented in all CitGF14 promoters except for CitGF14d. The Skn-1 motif required for endosperm expression was highly conserved in six CitGF14 promoters. The ABA responsive element ABRE was presented in CitGF14b, CitGF14c, CitGF14e, and CitGF14g. Box-W1, a fungal elicitor responsive element, was a highly conserved stress-related element which was found in five of nine CitGF14s. The upstream flanking regions of CitGF14g contained 11 stress-responsive promoters, including ABRE, ARE, CGTCA-motif, ERE, GC-motif, HSE, MBS, P-box, TC-rich repeats, TCA-element, TGA-element, and GACG- The graphic representation is displayed using GSDS (http://gsds.cbi.pku.edu.cn/). The unrooted phylogenetic tree was constructed using the amino sequences of CitGF14s genes by the Neighbor-Joining method with 1000 bootstrap replicates. Exons consisted of CDS shown as orange boxes, introns are shown as thin lines, and UTRs are shown as blue boxes.

CitGF14s Sequence Alignment
Amino acid sequence alignment showed that the deduced CitGF14s from C. sinensis were highly conserved with the exception of the N-terminal and C-terminal regions. Nine α-helices were shown in green rectangular boxes (Additional file 1: Figure S1A). Two conserved signature motifs RNL(L/V)SV(G/A)YKNV and YKDSTLIMQLLRDNLTLWTS were found in α-helix 3 and α-helix 9, respectively. Protein CitGF14b, CitGF14d, CitGF14e, CitGF14f, CitGF14g, CitGF14h, and CitGF14i had rather similar three-dimensional structures based on the Swiss-model prediction (Additional file 1: Figure S1B). All these similar proteins contained two potassium channel KAT1 ligands except for CitGF14b. The remaining two proteins (CitGF14a and CitGF14c) had relatively simple structures with no ligands.

Cis-Regulatory Elements
Different cis-acting elements related to plant growth, development, and stress responses were identified in CitGR14s (Additional file 2: Table S1). The circadian element was presented in all CitGF14 promoters except for CitGF14d. The Skn-1 motif required for endosperm expression was highly conserved in six CitGF14 promoters. The ABA responsive element ABRE was presented in CitGF14b, CitGF14c, CitGF14e, and CitGF14g. Box-W1, a fungal elicitor responsive element, was a highly conserved stress-related element which was found in five of nine CitGF14s . The upstream flanking regions of CitGF14g  contained 11 stress-responsive promoters, including ABRE, ARE, CGTCA-motif, ERE, GCmotif, HSE, MBS, P-box, TC-rich repeats, TCA-element, TGA-element, and GACG-motif. Moreover, other cis-acting elements associated with biotic and abiotic stress responses, such as WUN-motif, GARE-motif, AuxRR-core, and SARE were also identified.

Tissue-Specific Expression Patterns of CitGF14s
The expression of CitGF14s in callus, leaf, flower, and fruit, which were mined from the RNA-Seq raw data of C. sinensis genome database (http://citrus.hzau.edu.cn/orange/) is presented in Figure 4A. CitGF14a and CitGF14i were strongly expressed in flowers and leaves. The expression of CitGF14e in flower was higher than in callus and then further increased in leaves and fruit. CitGF14b expression was induced in callus and flower, slightly decreased in leaves, and then increased in fruit. CitGF14h was highly induced and constantly expressed in all four tissues or organs. On the other hand, CitGF14d was down regulated in callus, leaves, and fruit. Other CitGF14s were either slightly or moderately induced depending on tissue or organ.
The spatial expressions of CitGF14s were further analyzed by RT-qPCR in C. sinensis "Xuegan" when plants were not exposed to any stresses ( Figure 4B). CitGF14s were not highly induced in roots and stems except for CitGF14e and CitGF14h in stems that had over a 1.5-fold increase. The expression of CitGF14a, c, d, h, and i in shoots and CitGF14c, e, and h in leaves was highly induced, which was largely similar to those mentioned in above RNA-Seq data with the exception of CitGF14d that was primarily down regulated in callus, leaves, and fruit based on the RNA-Seq data ( Figure 4A), but it was highly upregulated in flower as well as shoots in the RT-qPCR analysis. Furthermore, the expression of all CitGF14s was low in peel, juice, and seeds.

Responses to Infection of Citrus Canker and Citrus Greening Pathogens
Affymetrix microarrays data were mined in this study for potential roles of CitGF14 genes in response to citrus canker. Results showed that all citrus probe sets contained less than 10,000 genes, indicating the not all citrus CitGF14s had been covered by the microarray data. For example, the probe signal for CitGF14d was not detected in the microarray. Based on eight CitGF14s from the microarray data, their responses to citrus canker infection are presented in Figure 5A (first four columns). CitGF14g was induced by Xanthomonas axonopodis pv. citri (Xaa) 6 to 48 h after infection and by Xanthomonas axonopodis pv. Aurantifolii (Xac) 48 h after infection ( Figure 5A). The inoculation of Xaa and Xac respectively induced CitGF14i expression only 48 h after infection. CitGF14h was slightly induced by Xaa and Xac 48 h after infection. The other CitGF14s did not respond to the infection of the two pathogens.
The infection of Xac to C. sinensis "Xuegan" caused downregulation of CitGF14a in 2 to 6 h and then variable expression thereafter till 192 h ( Figure 5B). CitGF14b was induced 6 h after inoculation, and its expression was then reduced from 12 to 96 h but highly increased at 192 h. CitGF14d was highly induced from 48 h to 192 h. CitGF14g responded quickly, 2 h after the inoculation, and reached the highest expression level from 48 h to 96 h. There was a downregulation of CitGF14i initially, its expression increased from 6 h to 24 h, and attained the highest expression level from 48 h to 192 h. The active responses of CitGF14g and CitGF14i largely concurred with the above microarray results. The other CitGF14s showed varied levels of down or upregulation over the 192-h evaluation period.
Microarray data were also explored for potential roles of CitGF14s in response to the infection of citrus greening: Candidatus Liberibacter asiaticus (Ca. Las) ( Figure 5A, five columns from the right). Compared to healthy organs, there was slight increase in the expressions of CitGF14c and CitGF14f in leaves and CitGF14e in peel. The other genes showed little response to the infection except for CitGF14a that was down regulated in leaves and stem and CitGF14g and CitGF14i that were down regulated in peel.
"Xuegan" when plants were not exposed to any stresses ( Figure 4B). CitGF14s were not highly induced in roots and stems except for CitGF14e and CitGF14h in stems that had over a 1.5-fold increase. The expression of CitGF14a, c, d, h, and i in shoots and CitGF14c, e, and h in leaves was highly induced, which was largely similar to those mentioned in above RNA-Seq data with the exception of CitGF14d that was primarily down regulated in callus, leaves, and fruit based on the RNA-Seq data ( Figure 4A), but it was highly upregulated in flower as well as shoots in the RT-qPCR analysis. Furthermore, the expression of all CitGF14s was low in peel, juice, and seeds.  Based on eight CitGF14s from the microarray data, their responses to citrus canker infection are presented in Figure 5A (first four columns). CitGF14g was induced by Xanthomonas axonopodis pv. citri (Xaa) 6 to 48 h after infection and by Xanthomonas axonopodis pv. Aurantifolii (Xac) 48 h after infection ( Figure 5A). The inoculation of Xaa and Xac respectively induced CitGF14i expression only 48 h after infection. CitGF14h was slightly induced by Xaa and Xac 48 h after infection. The other CitGF14s did not respond to the infection of the two pathogens. The microarray data was downloaded from NCBI database. Heatmap shows hierarchical clustering of CitGF14 expression after plants were inoculated with citrus canker pathogens: Xanthomonas axonopodis pv. citri (Xaa) and Xanthomonas axonopodis pv. Aurantifolii (Xac) and citrus greening pathogen: Candidatus Liberibacter asiaticus (Ca. Las). Gene expression values were calculated based on the ratios between the infection and the mock (control). Heatmap was generated based on log2 (infection expression/mock expression) values by using HemI. The color scale represents the relative intensity level of transcript abundance. (B) RT-qPCR analysis of CitGF14s expression after C. sinensis "Xuegan" was inoculated with Xac. The relative expression was normalized using the ACTIN and GAPDH genes as references using 2 −ΔΔ Ct method. The values were based on the means of four biological replications.
The infection of Xac to C. sinensis "Xuegan" caused downregulation of CitGF14a in 2 to 6 h and then variable expression thereafter till 192 h ( Figure 5B). CitGF14b was induced The microarray data was downloaded from NCBI database. Heatmap shows hierarchical clustering of CitGF14 expression after plants were inoculated with citrus canker pathogens: Xanthomonas axonopodis pv. citri (Xaa) and Xanthomonas axonopodis pv. Aurantifolii (Xac) and citrus greening pathogen: Candidatus Liberibacter asiaticus (Ca. Las). Gene expression values were calculated based on the ratios between the infection and the mock (control). Heatmap was generated based on log2 (infection expression/mock expression) values by using HemI. The color scale represents the relative intensity level of transcript abundance. (B) RT-qPCR analysis of CitGF14s expression after C. sinensis "Xuegan" was inoculated with Xac. The relative expression was normalized using the ACTIN and GAPDH genes as references using 2 −∆∆ Ct method. The values were based on the means of four biological replications.

Responses to Plant Hormone Treatments
Foliar application of jasmonate (JA) on "Xuegan" caused more downregulation of CitGF14 genes than upregulation ( Figure 6A). The upregulation only occurred with CitGF14a, c, and f 12 h after JA application and CitGF14h from 6 h to 12 h. Genes with the most pronounced down regulation were CitGF14a at 2 h and 48 h; CitGF14d from 6 h to 24 h; CitGF14e from 6 h to 12 h; and CitGF14g and i at 24 h. CitGF14b and CitGF14d showed noticeable responses to ABA ( Figure 6A). The former was primarily down regulated from 2 h to 48 h after application, whereas the latter was completely upregulated from 2 h to 48 h after application. ABA application also induced the expression of all the other CitGF14s at different time periods except for CitGF14i that was largely down regulated.   The application of ethephon (ETH) induced variable expressions of CitGF14a, b, c, f, and h ( Figure 6A). These genes were upregulated from 6 h to 12 h after application with exception of CitGF14b and c whose expression reduced at 24 h but increased at 48 h. The other genes were primarily down regulated after ETH application. Foliar spraying of salicylic acid (SA) also induced variable expressions of CitGF14a, b, c, f, and h ( Figure 6A). CitGF14a was upregulated after 12 h of application, and CitGF14b, c, f, and h were induced from 6 h to 12 h. The other CitGF14s were downregulated or remained unchanged.

Responses to Low and High Temperatures and Wounding
The exposure of C. sinensis "Xuegan" plants to 42 • C led to the upregulations of five CitGF14 genes at different times during the treatment ( Figure 6B). CitGF14a, e, and h were either uninduced or downregulated from 0 h to 24 h but highly induced at 48 h. On the other hand, the expression levels of CitGF14d and g were higher at 2 h, decreased thereafter, and then highly increased at 48 h. CitGF14i had similar expression pattern as CitGF14d and g, but its increase started 24 h and sustained to 48 h. The other CitGF14s were either downregulated or showed little change.
Chilling treatment of "Xuegan" resulted in CitGF14 responses opposite to the high temperature treatment. CitGF14b, c, and f, which were not upregulated at high temperature were strongly induced ( Figure 6B). CitGF14b was induced at 2 h, and its expression reached the highest at 6 h, and then decreased at 12 h. CitGF14c was induced at 2 h, attained the highest from 6 h to 24, but decreased at 48 h. The expression of CitGF14f was high at 2 h and peaked at 6 h, and then gradually reduced. Additionally, CitGF14h and CitGF14i were also highly induced at 6 h. The expression of the other CitGF14s were variable, and largely downregulated.
Wounding of "Xuegan" plants induced the upregulation of CitGF14a at 12 to 48 h and CitGF14i from 2 h to 6 h ( Figure 6B). Wounding treatment also moderately induced the expression of CitGF14b at 48 h and CitGF14c, f, and h at 6 h as well as CitGF14g from 2 h to 12 h, and CitGF14i at 12 h. The other CitGF14s either remained unchanged or downregulated.

Responses to Salinity and Drought Stresses
All CitGF14 genes responded to the salt treatment ( Figure 6C). CitGF14a was highly induced from 0 h to 48 h. The expressions of CitGF14b, e, and f were higher starting from 2 h to 144 h. The other CitGF14s were upregulated varying from moderate to high except for CitGF14d that fluctuated up and down and then remained downregulation from 24 h to 144 h.
The simulated drought stress by polyethylene glycol (PEG) treatment showed a rather similar expression pattern of CitGF14s as NaCl treatment ( Figure 6C). All genes were upregulated except for CitGF14d that was initially upregulated, then slightly decreased, and finally downregulated from 24 h to 144 h. CitGF14a and b were highly induced from 2 h to 12 h, and both became moderately induced in 24 h. The expression of CitGF14a was higher again at 48 h and then decreased thereafter, but CitGF14b maintained a high expression level from 48 h to 144 h. CitGF14c and e were highly upregulated at 2 h, remained moderate in expression from 6 h to 24 h and 6 h to 12 h, respectively, and finally highly induced thereafter. CitGF14f was consistently highly expressed from 2 h to 144 h.

Interactions among CitGFs
To determine whether 14-3-3 proteins could interact with each other or with other proteins, we systematically assessed the interactions among all the 14-3-3 proteins using yeast two hybrid (Y2H) assay. GAL4 DNA-binding domain (BD) was fused to prey protein (CitGF14a through CitGF14i), GAL4 transcriptional activating domain (AD) was fused to bait protein (CitGF14a through CitGF14i and pGADT7 empty vector), and they were co-transformed into yeast cells. As shown in Figure 7, CitGF14a, b, c, d, g, and h interacted with CitGF14i proteins to form a heterodimer while CitGF14i interacted with itself to form a homodimer in yeast. Interactions were also found in other 14-3-3 protein pairs, CitGF14g interacted with b and c as prey or bait. CitGF14d showed strong interaction with a and c as prey, while when CitGF14d was bait, the interaction became weak. CitGF14h interacted with CitGF14c as bait protein, but CitGF14h weakly interacted with all target and empty vector. Thus, CitGF14h showed auto-activation as prey vector.
To determine whether 14-3-3 proteins could interact with each other or with other proteins, we systematically assessed the interactions among all the 14-3-3 proteins using yeast two hybrid (Y2H) assay. GAL4 DNA-binding domain (BD) was fused to prey protein (CitGF14a through CitGF14i), GAL4 transcriptional activating domain (AD) was fused to bait protein (CitGF14a through CitGF14i and pGADT7 empty vector), and they were co-transformed into yeast cells. As shown in Figure 7, CitGF14a, b, c, d, g, and h interacted with CitGF14i proteins to form a heterodimer while CitGF14i interacted with itself to form a homodimer in yeast. Interactions were also found in other 14-3-3 protein pairs, CitGF14g interacted with b and c as prey or bait. CitGF14d showed strong interaction with a and c as prey, while when CitGF14d was bait, the interaction became weak. CitGF14h interacted with CitGF14c as bait protein, but CitGF14h weakly interacted with all target and empty vector. Thus, CitGF14h showed auto-activation as prey vector. Figure 7. Yeast two-hybrid (Y2H) assay to test interactions among nine CitGF14 (14-3-3) proteins of Citrus sinensis. The coding sequences of the CitGF14s genes were cloned into the Y2H vectors pGADT7 (AD) and pGBKT7 (BD), and introduced into yeast cells Y2H gold. Transformants were assayed for growth on SD/-Trp-Leu (A) and SD/-Trp-Leu-His-Adex-α-gal (B) nutritional selection media. EV was empty vector pGADT7 used as a negative control. Figure 7. Yeast two-hybrid (Y2H) assay to test interactions among nine CitGF14 (14-3-3) proteins of Citrus sinensis. The coding sequences of the CitGF14s genes were cloned into the Y2H vectors pGADT7 (AD) and pGBKT7 (BD), and introduced into yeast cells Y2H gold. Transformants were assayed for growth on SD/-Trp-Leu (A) and SD/-Trp-Leu-His-Ade-x-α-gal (B) nutritional selection media. EV was empty vector pGADT7 used as a negative control.

Discussion
Plant 14-3-3 proteins play important roles in regulation of plant responses to abiotic and biotic stresses. Different number of 14-3-3 proteins have been identified and analyzed in a number of land plant species, but such information is not available in citrus. The present study for the first time documented 14-3-3 proteins in C. sinensis and evaluated 14-3-3 genes in response to various stresses.

Nine 14-3-3 Genes and Their Characteristics in C. sinensis
A total of nine CitGF14 genes were identified in C. sinensis by a genome-wide search based on conserved domains and sequence similarities from known 14-3-3s. They were divided into ε-like and non-ε groups (Figure 1), and eight of the nine were located on five chromosomes ( Figure 2). Phylogenetic results were consistent with the clustering of 14-3-3 proteins in Arabidopsis [6], rice [40], and wheat [41]. The ε group, also known as "living fossil" 14-3-3 isoforms, is considered essential to eukaryotic biology, while the proteins in the non-ε group generally play organism-specific regulatory roles [42]. Gene structure analysis also supported this classification (Figure 3). Each of ε group genes (CitGF14a through CitGF14d) has six exons, while each of non-ε group genes (CitGF14e through CitGF14i) has three exons.
The relatively small number of CitGF14 genes along with no duplicated ones were not surprising because there was no recent whole-genome duplication in C. sinensis evolution except γ event which was shared by all core eudicots [43]. Phylogentic analysis suggest that C. sinensis 14-3-3s are evolutionally close to Populus since they belonged to the same clade and they are all ancient. Amino acid sequence alignment indicated that all C. sinensis 14-3-3s showed a high level of amino-acid similarity except for the Nterminal and C-terminal regions (Additional file 1: Figure S1), which are similar to all other 14-3-3s [44]. These highly conserved 30-kDa acidic proteins are, each composed of approximately 250 amino acids, nine α-helices, and two conserved signature motifs RNL(L/V)SV(G/A)YKNV and YKDSTLIMQLLRDNLTLWTS [14,44].
The online tool predicted that C. sinensis 14-3-3 isoforms were localized in cytoplasm, chloroplast, nucleus plasma, and plasma membrane (Table 1), suggesting distinct and differential patterns of subcellular distribution. These isoforms exhibited a high celltype specificity. The specificity of cellular and subcellular localization may contribute to their diverse interactions with targets as well as differential functions in cellular activities. Arabidopsis 14-3-3/GFP fusions experiment indicated that 14-3-3 localization is both isoform specific and highly dependent upon interaction with cellular clients [45]. Additionally, this interaction can alter the subcellular localization of target proteins [46]. In soybean, SGF14 proteins can regulate the nuclear-cytoplasmic movement of GmMYB176, which can alter the expression of CHS8, a gene in isoflavonoid biosynthesis [47].

CitGF14s Were Differentially Expressed in Various Organs of Citrus
The tissue-specific pattern of gene expression can provide important clues about gene function [48]. The expression of CitGF14s in different organs ( Figure 4) may suggest that they are involved in various aspects of physiological and developmental processes. CitGF14a, e, h, and i were highly upregulated in different organs based on RNA-Seq raw data ( Figure 4A). RT-qPCR analysis also showed the highly upregulation of CitGF14a, c, d, h, and i in shoots and CitGF14c, e, and h in leaves, which were principally similar to those from the RNA-Seq analysis. A discrepancy occurred in CitGF14d between RNA-Seq ( Figure 4A) and RT-qPCR ( Figure 4B) data. RNA-Seq data showed its down regulation in almost all tested tissue and organs, but it was dramatically unregulated in flower and also in shoot in RT-qPCR analysis. Such a disagreement could be attributed to the differences in cultivars and plant growth conditions, which needs further investigation. Nevertheless, the high level of CitGF14d transcript in flowers and shoot in contrast to the minimal levels in the other organs ( Figure 4B) may suggest the specificity of CitGF14d in flower and leaf development. A similar expression pattern also occurred in another woody plant mulberry tree [49] where some MaGF14s were specifically expressed in certain organs. Additionally, soybean 14-3-3 isoforms also showed ubiquitous expression in all tissues, and different expression of SGF14 genes in embryos during seed development indicated that 14-3-3s may be involved in soybean seed development [36]. In cotton, Northern blotting and RT-qPCR analysis showed that Gh14-3-3 genes were developmentally regulated in fiber development [50,51]. In an early report, a high level of Arabidopsis 14-3-3ω mRNA occurred in flowers. On the other hand, 14-3-3κ and 14-3-3λ expression did not show much difference across tissues [52]. The expression of different CitGF14s in different organs or in the same organs of citrus may indicate the versality of CitGF14 involvement in citrus growth and development, which deserve further investigation.

CitGF14s Were Induced by Citrus Canker and Greening Infections
Citrus canker and citrus greening are two notorious bacterial pathogens and have significantly affected citrus production worldwide. In this study, both microarray and RT-qPCR data clearly showed that CitGF14s are involved in citrus responses to canker ( Figure  5). Microarray data indicated that both CitGF14g and CitGF14i were highly upregulated upon the infection of both Xaa and Xac ( Figure 5A). RT-qPCR analysis further confirmed the upregulation of both CitGF14g and CitGF14i ( Figure 5B) and found that CitGF14b, d, and f were also involved in the later period of Xac infection. These results indicate that 14-3-3 proteins may involve in the regulation of citrus canker resistance as well as functional redundancy in stress tolerance.
The responses of CitGF14s to citrus greening pathogen were not pronounced as those to citrus canker based on the microarray results ( Figure 5A). There was a slight increase in the expressions of CitGF14c and f in leaves and CitGF14e in peel, whereas CitGF14a, g, and i were downregulated in leaves, stems or fruit peel. How such up or down regulations affects the pathogen development is unknown. Considering the severity of citrus greening in the citrus industry and the role of 14-3-3 in regulation of plant responses to biotic stresses, further research on CitGF14s in regulation of citrus response to greening is warranted. Pathogen infection triggered differential expressions of 14-3-3 genes has been documented in other plants. In tomato, at least 10 14-3-3 genes are differentially expressed in response to fungal toxin fusicoccin [32]. The tomato 14-3-3 protein TFT7 can bind with MAPKKKα and SIMKK2, resulting in programmed cell death associated with immunity [53]. TFT4 is another member of tomato 14-3-3 proteins that can bind with the effector XopQ from Xanthomonas euvesicatoria (Xcv) to suppress the effector-triggered immunity [54].

CitGF14s Differentially Responded to Abiotic Stresses
Phytohormones play a central role in plant defense responses to environmental stress. RT-qPCR analysis showed that all CitGF14 genes responded to JA, ABA, ETH, and SA applications by either down or up regulation at different times ( Figure 6A). Such variable responses may suggest that 14-3-3s could function as a multiple regulator in plant hormone signaling. Four rice GF14 genes were induced by benzothiadiazole, JA, ETH, and H 2 O 2 during pathogen attach [26]. Quantification of the 20R/16R promoter-driven GUS expression in different transgenic potato plants revealed that 14-3-3 isoforms can be induced by various stimuli, such as ABA, SA, NaCl, and metal ions [55]. The Arabidopsis 14-3-3 λ isoform was reported to specifically bind the C-terminal domain of RPW8.2, resulting in the enhanced resistance to powdery mildew fungus via the SA signaling pathway [56]. In the present study, CitGF14b was highly downregulated, but CitGF14d was strongly upregulated by ABA treatment. This may suggest that citrus 14-3-3s could link to ABA in mediation of different stress responses. A study with barley 14-3-3 proteins as baits in yeast two-hybrid (Y2H) library resulted in the identification of 132 new molecular targets, including three ABA signal transduction related proteins (AREB/ABF/ABI5-like proteins) [57].
High and low temperatures as well as wounding induced variable expressions of all nine CitGF14s ( Figure 6B). CitGF14a, d, e, and g that were upregulated at high temperature became down regulated in low temperature treatment. The reverse is true for CitGF14b, c, and f. Only CitGF14h and i had both up and down regulations in high and low temperature treatments. These results may indicate that different isoforms of CitGF14s were involved in response to two opposite temperature stresses. A study with Arabidopsis showed that overexpression of 14-3-3ε was ineffective in cold tolerance, but overexpression of both ω and ε produced more cold tolerant plants [58]. Wounding also induced all CitGF14s gene expression including the upregulation of CitGF14a and i but downregulation of CitGF14d at variable times. These results concurred with a report by Lapointe et al. [59] that 14-3-3 mRNA was upregulated in poplar plants after wounding treatment.
CitGF14s exhibited similar expression patterns in response to both NaCl and PEGsimulated drought stresses ( Figure 6C). All CitGF14s genes were highly upregulated except CitGF14d that was downregulated, suggesting that all CitGF14s participated in citrus responses to NaCl and drought. These results are consistent with 14-3-3 responses to salinity and drought stresses in other plants [3,22,60]. An Arabidopsis GF14λ was introduced into cotton plants, resulting in improved drought tolerance with a "stay-green" phenotype. The stomata of the transgenic plants might be regulated by GF14λ through some transporters, such as H+-ATPase whose activities are controlled by their interaction with 14-3-3 proteins. In the present study, we noticed that CitGF14d, which was highly induced by ABA signal (Figure 6A), were downregulated under salinity and drought stresses ( Figure 6C). This result may indicate that CitGF14d acted as a negative regulator. The overexpression of GsGF14o from Glycine soja in Arabidopsis resulted in down-regulation of a droughtresponsive marker gene, the transgenic line showed reduction of stomatal development under drought treatment. Thus, the Glycine soja 14-3-3 gene GsGF14o functioned as a negative regulator of drought tolerance [61].

Prediction of Interactions among CitGF14 Proteins
Different genes with similar expression patterns, commonly known as co-expressed genes were believed to be functionally related [62]. To improve our understanding of CitGF14s' functions in C. sinensis, CitGF14s were integrated into the Network Inference for Citrus Co-Expression (NICCE) (http://citrus.adelaide.edu.au/nicce/home.aspx). A detailed 14-3-3 gene information and tabs were presented in Additional file 3: Table S2, which contained top 100 expressed genes in C. sinensis based on HRR (highest reciprocal ranks). HRR highlighted in black (bold), black, and grey colors signified statistical significance of HRR at p < 0.01, p < 0.05, and p > 0.05 levels, respectively. As described previously, the microarray data did not cover all the CitGF14 genes, we only obtained co-expressed genes from six member of CitGF14 genes in C. sinensis datasets. We found a number of genes co-expressed with three citrus canker responsible CitGF14s (CitGF14g, CitGF14h, and CitGF14i). They are leucine-rich repeat family proteins, zinc finger (C3HC4-type RING finger) family proteins, and zinc finger (DHHC type) family proteins, WD-40 repeat family proteins, lesion inducing protein-related, and peroxidase 63. These proteins are involved in a variety of functions ranging from signal transduction and transcription regulation to defense responses. For example, the leucine-rich repeat family proteins are associated with innate immunity in plants, serving as the first line of defense against pathogens. Our results may indicate that CitGF14g, CitGF14h, and CitGF14i are involved in regulation of defense response to the infection of citrus canker pathogens.
The interaction among 14-3-3 isoforms or individual isoforms with other proteins is important for understanding the biological functions of 14-3-3s. Y2H assay is a powerful tool widely used for identifying novel protein-protein interaction [63]. In the present study, CitGF14i was found to be able to interact with CitGF14a, b, c, d, g, or h to form a heterodimer and interact with itself to form a homodimer in yeast (Figure 7). Interactions were also found in other 14-3-3 protein pairs, including CitGF14g with b or c and CitGF14d with a or c. Among the isoforms, CitGF14i appears to be the most active and important one due to its interaction with six isoforms and also with itself. It is known that 14-3-3 isoforms have different affinities to individual targets; thereby, there is a possibility that regulation of specific processes could be accomplished by single 14-3-3 isoforms [16]. In this study, CitGF14i was highly up-regulated in flower and leaves or shoots in general (Figure 4), strongly induced by the infection of Xac (Figure 5B), largely down-regulated in response to the application of growth regulators, and variably expressed in abiotic treatments ( Figure 6). Intriguingly, CitGF14i is the only one localized in plasma membrane. Whether its subcellular location contributes to such active interactions is unknown. Although the function of all CitGF14s deserve further investigation, specific attention should be given to CitGF14i for its interactions with other proteins. Y2H screens complemented with 14-3-3 protein affinity purification and tandem mass spectrometry are another powerful tool for identifying protein interactions. This method identified five 14-3-3 isoforms in 7-day-old barley. Some of proteins were identified as 14-3-3 targets in both Y2H and affinity purification including 14-3-3 proteins themselves [64]. In order to uncover the 14-3-3 signaling pathway in healthy and disease, a high-throughput data in VisANT graphs (http://visant.bu.edu) was used to graph and validate 14-3-3 protein interactions [65].
Two public citrus databases provided a genome-wide approach to predict 14-3-3 protein-protein interactions (PPI) or gene co-expression. Thus, a 14-3-3s protein interaction network was generated by CitrusNet ( Figure 8). Among them, CitGF14a, e, and i play much greater roles than CitGF14g and h, which are more important than the remaining four CitGF14s. CitGF14d, CitGF14g, and CitGF14i contribute to their resistance to Xac infection, and CitGF14d could be a negative regulator in response to drought stress. With more than 150 nodes in the network, most nodes had different degrees of connection. Detailed connection among each CitGF14 and target was provided in Additional file 4: Table S3. More than 70 citrus 14-3-3 targets were identified, which participate in many molecular processes, including those involved in development, hormone, redox, signaling, stress, and transport. Polyubiquitin-A, clathrin heavy chain 1, heat shock protein 81-3, and heat shock protein 83 were protein nodes with the highest degree in CitrusNet [66], and these proteins also were found in 14-3-3 PPI network. Furthermore, casein kinase I, vesicle-fusing ATPase, serine/threonine-protein kinase TOR (TARGET OF RAPAMYCIN), histone deacetylase 6, leucine rich repeat-type serine/threonine receptor-like kinase were predicted to interact with several members of CitGF14s. TOR kinase is a client of CitGF14s. In Arabidopsis, TOR kinase was important in controlling plant growth, responding to environmental cues, and regulating cell processes [67]. In genome-wide PPI network, TOR kinase was a central part of citrus hormone cross-talk, which potentially interacted with proteins related to hormone signaling and hormone receptors [66]. Therefore, all these proteins play critical roles in CitGF14s regulating of various aspects of cellular processes. Furthermore, different members of CitGF14 interactions were found in both co-expression network and PPI network. Protein BRASSINOSTEROID INSENSITIVE 1 and leucine rich repeat-type serine/threonine receptor-like kinase appeared in PPI network, and they were also predicted as 14-3-3s co-expression genes.

Sequence Alignment and Phylogenetic Analysis of 14-3-3 Proteins
The amino acid sequences of 14-3-3 genes from different plant species were aligned using the software Clustal_X (version 1.83) with default parameters [69]. The unrooted phylogenetic trees were constructed based on alignments using MEGA 4.0 with the neighborjoining method [70]. The bootstrap test was carried out with 1000 replicates.

RNA-Seq and Microarray Data Analysis
After the identification and characterization of 14-3-3 genes in citrus, publicly available RNA-Seq raw data and microarray database were mined for 14-3-3 expression profiles in different organs of citrus and in response to abiotic and biotic stresses.
The expression of 14-3-3s in callus, leaf, flower, and fruit were analyzed using RNA-Seq raw data available in C. sinensis genome database (http://citrus.hzau.edu.cn/orange/). The expression level of 14-3-3 genes was calculated as Log2 based RPKM. The expression data were hierarchically clustered with average linkage and displayed in HemI [71].
Affymetrix microarray data obtained from the NCBI Gene Expression Omnibus (GEO) database under the series accession number GSE33003, GSE33004, and GSE10798 were mined for citrus 14-3-3 response to biotic stresses. In experiments with Ca. Las infection (GSE33003 and GSE33004), young, healthy Valencia sweet orange plants were graftinoculated with budwood from Ca. Las-infected citrus plants. The leaf, stem, and root samples were collected from three symptomatic and three healthy control trees for RNA extraction and analyzed using microarrays [72,73]. For citrus canker (Xac) or (Xaa) infection (GSE10798), adult leaves of sweet orange were infiltrated with the bacterial suspensions or water (mock control). Leaves samples were collected after bacterial infiltration for RNA extraction and hybridization on Affymetrix microarrays [74]. The microarray CEL files were normalized using Robust Multi-array Average (RMA) in R/Bioconductor (ver 2.15), and normalized data was used for identifying differential expressed genes [75,76]. The heatmap for the 14-3-3 transcripts with their expression values were performed by HemI [66].

Plant Materials and Treatments
To confirm the expression of 14-3-3 genes in different organs of C. sinensis, root, stem, leaf, and shoot samples were taken from one-year old seedling of C. sinensis. Flower, fruit peel, juice, and seed samples were collected from eight-year-old adult trees. Four biological samples per organ were frozen in liquid N and stored in −80 • C.
Further analysis of 14-3-3 genes in response to various stresses were conducted using C. sinensis "Xuegan". Seeds of "Xuegan" were germinated in a plant growth chamber at 25 • C under a photoperiod of 16-h light/8-h dark with a relative humidity of 70%. The light was provided by fluorescent white-light tubes. At the two-leaf stage, seedlings were transplanted singly into pots (15 × 15 cm) filled with Fafard Professional Potting Mix (Sun Gro Horticulture, Agawam, MA, USA). The following treatments were applied to the plants, each experiment was arranged as a complete randomized design with four replicate per treatment.
Infection with citrus canker: The bacterial strain Xac 29-1 was cultured in NA nutrient broth with 1.5% agar at 28 • C for 36 h. Cultured bacterial cells were washed twice with sterile water and then resuspended in sterile water to a final concentration OD600 = 0.3. The diluted cells were infiltrated into leaves with a needleless syringe, sterile water was injected into leaves as the control treatment. Leaf samples were collected after 0, 2, 6, 12, 48, 96, and 192 h of inoculation, respectively.
Hormone treatments: Seedlings were sprayed with 1 mM SA, 100 µM JA, 100 µM ABA, and 100 µM ETH solutions, respectively. Leaf samples were collected after 0, 2, 6, 12, 24, and 48 h of the application, respectively. Leaf samples collected from seedlings sprayed with distilled water at the corresponding time were considered the control treatment.
Low or high temperature and wounding treatments: Seedlings were incubated in a growth chamber with temperature of 4 • C or 42 • C for 48 h. For wounding treatment, each fully expanded leaf was penetrated with a needleless syringe 10 times. Leaf samples were collected at 0, 2, 6, 12, and 48 h, respectively. Seedlings grown in the chamber with the temperature at 25 • C as the control, and leaf samples were collected at the same times as those of treated seedlings.
Salt and drought treatments: Seedlings were grown in the potting mix were drenched with 200 mM NaCl or 20% PEG6000 solution until leachate appeared from the bottom of pots (about 250 mL solution was used for each treatment). Seedlings were irrigated with 250 mL water as the control treatment. Leaf samples were collected from treated and control seedlings after 0, 2, 6, 12, 48, 96, and 144 h of treatment, respectively.
All the collected leaf samples were immediately frozen in liquid N and stored at −80 • C for RNA extraction.

RNA Isolation and Expression Analysis
Total RNA was extracted from the collected samples using the RNAprep pure Plant Kit (Tiangen, Beijing, China) according to the manufacturer's instructions. One thousand nanograms of total RNA was used to synthesize first-strand cDNA with EasyScript One-Step gDNA Removal and cDNA Synthesis SuperMix (Transgen, Beijing, China). Quantitative PCR was carried out using TransStart Tip Green qPCR SuperMix (Transgen, Beijing, China) on a CFX96 Real-time System (Bio-Rad, Hercules, CA, USA) according to the manufacturer's protocol. GAPDH [77], Actin, and F-box [78] were used as reference genes to normalize the expression of the investigated genes. Gene specific primers (Additional file 5: Table S4) were designed with the Primer Premier 6 software (Premier Biosoft International, San Francisco, CA, USA). The PCR reaction mixtures were incubated at 95 • C for 30 s, followed by 40 cycles of 95 • C for 5 s and 60 • C for 30 s. The relative expression levels were determined by 2−∆∆ Ct method described by Livak and Schmittgen [79], data were analyzed using SPSS 22.0 (IBM Corporation, Somers, NY, USA) statistics software, and mean differences were separated by Tukey's HSD test at p < 0.05 level with four biological replicates. The heatmap for the 14-3-3 transcripts with their expression values were generated using HemI tool [66].

Prediction of Cis-Regulatory Elements
The sequences of 1.5 kb upstream regions from the translation initiation codon of each gene of 14-3-3s were selected and subjected to analysis by the online database PlantCare. Putative developmental and stress-responsive cis-elements in citrus 14-3-3s were identified.

Gene Co-Expression Network and Protein-Protein Interaction Prediction
To predict gene interaction on a genome-wide scale, keyword 14-3-3s were searched in publicly accessible tool NICCE (http://citrus.adelaide.edu.au/nicce/home.aspx) to predict potential targets of 14-3-3s [80]. CitrusNet and PPI networks in C. sinensis were constructed using ortholog-based and domain-based interaction methods, which contained 8195 proteins and 124,491 interactions [65]. The nine citrus 14-3-3 proteins were used as hub nodes to connect potential target proteins in CitrusNet (http://citrus.hzau.edu.cn/ orange/ppi/index.php). All 14-3-3s and client proteins were linked into an interconnected sub-network which was visualized by Cytoscape.

Analysis of CitGRFs Interactions by Y2H Assay
The full-length coding sequences of 9 citrus 14-3-3 proteins were introduced into the pGBKT7 fusion bait vector and pGADT7 fusion prey vector. The fused pGADT7-CitGF14s and pGBT9-CitGF14s recombinant vectors were then co-transformed into yeast strain Y2H gold by LiAc/SS carrier DNA/PEG method [81]. The transformants were first selected in the SD/-Trp-Leu medium and PCR testing. After that the positive colonies were transferred to the selection medium supplemented with X-α-gal but lacked Trp, Leu, His, and adenine (SD/-Trp-Leu-His-Ade). Aureobasidin A (AbA) was added to the selection plates to suppress the auto-activation of the prey vectors.

Conclusions
The present study identified nine 14-3-3 genes (CitGF14a through CitGF14i) in C. sinensis through genome-wide analysis. All the CitGF14s genes were detected in different tissues or organs but varied in abundance. Transcript levels of CitGF14s were also analyzed after plants were treated with hormones, extreme temperatures, drought, salinity, wounding, and infected with Xac 29-1 strains. Almost all CitGF14s responded to the treatments by variable levels of expression during the experiments, suggesting that CitGF14s play important roles in citrus responses to different exogenous and endogenous signals. This study also showed that most gene family members had a functional divergence of 14-3-3 proteins. Y2H assay showed that CitGF14i was the most active and important isoform due to its interaction with six other isoforms and also with itself. Additionally, CitGF14d, CitGF14g, and CitGF14i contribute to their resistance to Xac infection, and CitGF14d could be a negative regulator in response to drought stress. Finally, a citrus 14-3-3 interactome network was constructed by PPI method and microarray gene co-expression. Our study for the first time provides a comprehensive framework about 14-3-3 family genes in C. sinensis, which may lead to further investigation of their roles in citrus growth and development as well as in response to abiotic and biotic stresses.