Transcriptome analysis unravels spatiotemporal modulation of phytohormone-pathway expression underlying gibberellin-induced parthenocarpic fruit set in San Pedro-type fig (Ficus carica L.)

Transcriptome analysis unravels spatiotemporal modulation of phytohormone-pathway expression... Background: Gibberellin (GA) treatments can induce parthenocarpy in the main crop of San Pedro-type figs, the native non-parthenocarpic fruit, however, the underlying mechanism is still largely unclear. Results: In our study, GA was applied to San Pedro-type fig main crop at anthesis. Sharply increased GA content 3 3 was detected in both female flowers and receptacle, along with significantly decreased indole-3-acetic acid (IAA), zeatin and abscisic acid (ABA) levels in female flowers, and increased zeatin peak intensity and earlier ABA peak in receptacles. Transcriptome comparison between control and treatment groups identified more differentially expressed genes (DEGs) in receptacles than in female flowers 2 and 4 days after treatment (DAT); 10 DAT, the number of DEGs became similar in the two tissues. Synchronized changing trends of phytohormone-associated DEGs were observed in female flowers and receptacles with fruit development. Modulation of ethylene and GA signaling and auxin metabolism by exogenous GA occurred mainly 2 DAT, whereas changes in auxin, cytokinin and ABA signaling occurred mainly 10 DAT. Auxin-, ethylene- and ABA-metabolism and response pathways were largely regulated in the two tissues, mostly 2 and 10 DAT. The major components altering fig phytohormone metabolic and response patterns included downregulated GA2ox, BAS1, NCED and ACO, and upregulated ABA 8′-h and AUX/IAA. Conclusions: Thus GA-induced parthenocarpy in fig is co-modulated by the female flowers and receptacle, and repression of ABA and ethylene biosynthesis and GA catabolism might be the main forces deflecting abscission and producing fig parthenocarpy. Keywords: Ficus carica, Gibberellin treatment, Parthenocarpy, Transcriptome analysis, Plant hormone * Correspondence: hqma@cau.edu.cn Lijuan Chai and Peng Chai contributed equally to this work. College of Horticulture, China Agricultural University, Beijing, People’s Republic of China Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. 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. Chai et al. BMC Plant Biology (2018) 18:100 Page 2 of 15 Background of GA metabolism [17]. SlIAA9, a member of the Aux/ Gibberellins (GAs) are involved in almost every aspect IAA family, is an auxin-signaling repressor that prevents of plant growth and development. In plant reproductive tomato ovary development [18]. Moreover, fruit set in to- processes, GAs regulate floral initiation [1], fruit set and mato induced by auxin analogue 2,4-D was significantly growth [2], and seed maturation and germination [3, 4]. suppressed by GA-biosynthesis inhibitors [7]. Similar to Among the abundant GAs identified to date, only GA , the situation in Arabidopsis and tomato, auxin is the GA ,GA and GA have been recognized as bioactive major inducer of fruit set in Capsicum annuum, depend- 3 4 7 [5], and only GA ,GA and/or GA have been success- ing in part on enhanced GA biosynthesis [19]. Ethylene is 3 4 7 fully applied for parthenocarpic fruit induction [6]. also considered to be a key regulator in coordinating fruit Parthenocarpy, a highly valuable agronomic character- set, and ethylene reduces fruit set by suppressing GA me- istic, especially in edible fruit crops, means that fruit set tabolism [10]. and development can be uncoupled from pollination Ficus carica L. (Moraceae) bears urn-shaped fruit with and fertilization and thus the fruit is seedless, such as in an enclosed inflorescence structure termed syconium. fig, banana, persimmon, pineapple and pear [6]. There are two major sex types in fig trees: caprifig (male Parthenocarpic fruit can also be triggered by application fig) and female fig; the latter is further classified into of exogenous plant hormones, termed artificial par- three types, common, Smyrna and San Pedro, based on thenocarpy. Auxin and GAs are the most frequently cropping characteristics [20]. Parthenocarpic or used hormones in parthenocarpic fruit induction [7, 8]. pollinated non-parthenocarpic fig fruit show a typical Other plant hormones, including cytokinin [9], ethylene double-sigmoid growth curve, with two rapid [10] and abscisic acid (ABA) [11], also regulate or par- fruit-size-increment phases, i.e., stages I and III, sepa- ticipate in parthenocarpic fruit set and growth. rated by a lag phase (stage II) [21]. Every year, San Pedro Parthenocarpy requires the unique and coordinated action fig produces a parthenocarpic first crop called breba, of different phytohormones, whereas non-parthenocarpic and a non-parthenocarpic main crop which constitutes fruit will senescence and abscise if fertilization is not the main yield [20]. Studies on non-parthenocarpic achieved. Auxin, GA and cytokinin are generally recognized Smyrna fig cv. Calimyrna revealed that parthenocarpic as the major regulators of fruit set. Elevated levels of these fruit set can be induced through auxin, GA and cytoki- hormones in tomato ovaries have been associated with fruit nin treatments [22, 23]. Analysis of auxin and GA setand early fruitgrowth [12]. Auxin and GA, and the cross- content in the two crops of San Pedro fig cv. King dem- talk between them, largely modulate pollination-dependent onstrated higher auxin content in the parthenocarpic and parthenocarpic fruit set in tomato [13]. Transcriptome breba than the pollinated non-parthenocarpic main crop, comparisons between pollinated and parthenocarpic cucum- whereas GA was higher in the latter than in the ber fruit have shown crosstalk of auxin, GA and cytokinin at former [24]. Our previous studies revealed that compared the transcriptional level during parthenocarpic fruit set [14]. with San Pedro parthenocarpic breba, expression of GA- Exogenous GA application significantly modulated the and auxin-biosynthesis genes is repressed in its expression patterns of plant hormone metabolism and non-parthenocarpic main crop female flowers, whereas signaling genes in seedless grape, as determined by transcrip- ABA- and ethylene-biosynthesis genes are enhanced [25]. tome analysis [15]. Previous studies in several plants have Though exogenous hormone treatment can induce fruit demonstrated that parthenocarpy induced by auxin and set and development of non-parthenocarpic fig, our un- cytokinin requires downstream GA biosynthesis. Application derstanding of the underlying mechanisms is limited. of exogenous auxin and cytokinin in grape increased the ex- In this study, the effect of exogenous GA treatment on pression of GA-biosynthesis genes, such as VvGA3ox1,and the main crop of San Pedro fig was studied, and hormone repressed GA-catabolism genes, such as VvGA2ox3 and contents of the female flowers and receptacle were VvGA2ox4, suggesting that auxin- and cytokinin-induced assayed. A corresponding global transcriptome compari- parthenocarpic fruit set requires enhanced GA biosynthesis son demonstrated a significant number of gene ontology [9]. Parthenocarpy induction in tomato by cytokinin depends (GO) terms and Kyoto Encyclopedia of Genes and in part on modulation of GA and indole-3-acetic acid (IAA) Genomes (KEGG) pathways with spatiotemporal regula- metabolism, as reflected by the upregulated expression tion characteristics. The expression patterns of annotated levels of GA-biosynthesis genes such as GPS, GA20ox and plant hormone metabolism and signal-transduction genes GA3ox, and the IAA-biosynthesis gene ToFZY, and down- were analyzed in depth. GA treatment led mainly to the regulation of the GA-inactivation gene GA2ox [16]. regulation of auxin, ethylene and ABA biosynthesis and Auxin-induced parthenocarpic tomato fruit set is partially signaling in fig. Differential expression of plant hormone mediated by GAs, as indicated by the dependence of par- genes revealed synchronized change trends with fruit thenocarpic fruit formation in the unfertilized entire to- development between female flowers and receptacles mato mutant with a defective SlIAA9 gene on regulation following GA treatment. Our work provides new Chai et al. BMC Plant Biology (2018) 18:100 Page 3 of 15 information on GA-induced fruit set in San Pedro-type fig by were assessed by Nanodrop 2000 (Thermo Scientific, unraveling key pathways and genes in the expression-regulation USA) and electrophoresis in a 1% agarose gel. RNA in- network, which could help elucidate the molecular mechanisms tegrity number, analyzed by Agilent 2100 Bioanalyzer, underlying the diversified parthenocarpic nature of fig species. was > 8.0. Methods Library preparation and paired-end transcriptome Plant materials and exogenous GA treatment sequencing A San Pedro-type fig cultivar (Ficus carica L. ‘Asteran’) The mRNA of each sample’s three biological replicates was used in this study. The trees were planted in a com- were respectively enriched using cellulose oligo (dT) mercial orchard in Beijing with regular management. magnetic beads (Invitrogen, USA), then fragmented into Main crop fruits at pre-anthesis (15–20 mm), anthesis ca. 200-bp fragments. The fragments were transcribed (20–25 mm) and post-anthesis (> 25 mm) stages were and double-stranded cDNA was synthesized, then end selected to evaluate the time window of GA treatment; repair, 3′-end single-nucleotide A (adenine) addition and 200 μl of 25 mg/l GA solution was injected through the ligation of adaptors were performed according to the ostiole into the fruit cavity with a 1-ml syringe. Solution manufacturer’s instructions. The resultant fragments without GA was injected into the control group. For were enriched by PCR and purified using 2% Certified each treatment and control group, 100 fruits were Low Range Ultra Agarose (Bio-Rad) to create the final injected and fruit-set rates were calculated. cDNA library. The quantitative assay was conducted Growth curves of control and GA -treated fruit were using Picogreen fluorescent dye (Invitrogen, USA) with a established with 30 tagged fruits using a digital slide cali- TBS-380 fluorimeter (Invitrogen, USA). After bridge per. Three biological replicates were collected from pools PCR amplification on Illumina cBot using Truseq PE of at least 30 fruits 2, 4 and 10 days after treatment Cluster Kit v3-cBot-HS, paired-end (2 × 151 bp) sequen- (DAT). For transcriptome sequencing and plant hormone cing of the cDNA library products was carried out using analysis, main crop fruits at anthesis, 22 ± 1 mm in trans- the Illumina HiSeq4000 platform. verse diameter, were used. Female flowers and receptacle were carefully separated, frozen in liquid nitrogen on site, Illumina read processing and functional annotation pulverized in the lab, and stored at − 80 °C for further use. Clean reads were generated by removing reads with adaptors and more than 10% unknown nucleotides, and Phytohormone content assays low-quality reads (the rate of reads with quality value The levels of IAA, GA ,GA , zeatin and ABA were moni- ≤10 was more than 20%) from the raw data. Then clean 3 4 tored by reverse-phase HPLC–MS/MS [26]. In brief, about reads were mapped to our reported reference sequences 50 mg of the frozen and pulverized plant material was deposited at DDBJ/EMBL/GenBank under the accession placed in 500 μl extraction solvent (isopropanol:H O:con- number GDKC00000000 [25] using Bowtiealigner centrated HCl, 2:1:0.002, v/v), then shaken (100 rpm) at (http://bowtie-bio.sourceforge.net/index.shtml)[28] with 4 °C for 30 min. Dichloromethane (1 ml) was added, no more than two mismatches allowed. The number of followed by shaking (100 rpm) at 4 °C for 30 min. After mapped clean reads for each transcript was normalized centrifugation (10,000 × g) at 4 °C for 5 min, the solution to FPKM (fragments per kilobase of exon model per mil- was divided into two phases, and the lower phase (ap- lion mapped reads) by RSEM (http://deweylab.biostat.- proximately 1 ml) was collected. The solvent mixture was wisc.edu/rsem/). Genes that were differentially expressed concentrated to dryness using a Termovap sample con- between GA -treated samples and their corresponding centrator (N-EVAP, Organomation, USA), then redis- controls were analyzed by edgeR (http://www.biocon- solved in 100 μl methanol. The extracted hormone ductor.org/packages/2.12/bioc/html/edgeR.html)[29], solutions were run in an AB SCIEX (USA) Triple Quad™ and “false discovery rate (FDR) < 0.01 and absolute value 5500 LC–MS/MS system. For each sample, three of log (FPKM /FPKM ) ≥ 1” were set as the 2 treatment control independent and parallel extractions were carried out. thresholds to determine significant differences in gene Presented values are means of all replicates ± SD. Signifi- expression. For GO annotation of all transcripts (http:// cance analysis was performed by SPSS 17.0. www.geneontology.org/), Goatools (https://github.com/ tanghaibao/Goatools) was used. Pathway-enrichment RNA extraction analysis was carried out based on the KEGG pathway data- Total RNA from female flowers and receptacles taken 2, base by KOBAS (http://kobas.cbi.pku.edu.cn/)[30]. Both 4 and 10 DAT and the corresponding controls was iso- GO and KEGG enrichment analyses were based on Fisher’s lated using the CTAB method [27] and treated with Exact Test [31] with multiple testing correction of FDR RNase-free DNase I (TaKaRa, China) according to the [32], and corrected P-value ≤0.05 was selected as the manufacturer’s instructions. RNA quality and quantity threshold for significance. Chai et al. BMC Plant Biology (2018) 18:100 Page 4 of 15 Quantitative real-time PCR (qRT-PCR) verification its level 2 DAT (Fig. 1c); thereafter, IAA decreased to Thirteen phytohormone-related genes were selected for very low levels in both control and treated receptacles. validation by qRT-PCR and the specific primers are Exogenous GA treatment dramatically decreased zeatin shown in Additional file 1: Table S1. Isolation of total peak intensity in the female flowers 4 DAT, whereas it RNA was as described previously and RNase-free DNase specifically enhanced the zeatin peak to 2.4-fold that in I (D2270A, TaKaRa) treatment was performed to digest control receptacles 2 DAT (Fig. 1c); this action was op- DNA in the sample. For the first-strand cDNA synthesis, posite that in female flowers. ABA content was higher in 1000 ng of total RNA was reverse-transcribed with a GA -treated female flowers and receptacles than in their synthesis kit (D6210A, TaKaRa). Ten-fold-diluted cDNA control counterparts 2 and 4 DAT, whereas its level was template was used for the qRT-PCR assay performed by repressed in both tissues 10 DAT, to a level that was less the SYBR Premix Ex Taq Kit (DRR420A, TaKaRa) on a than half of the control. Thus, GA caused a shift in the 7500 Real-Time PCR System (Applied Biosystems, USA) ABA peak to 6 days earlier than its natural peak in fe- according to the manufacturer’s protocol. Relative quan- male flowers and receptacle (Fig. 1c). −ΔΔCt titative method (2 ) was used to calculate the level of expression of the tested genes [33]. All of the reac- Transcriptome sequencing analysis and qRT-PCR tions were performed in three replicates using actin as validation the internal gene [34]. The transcriptomes of female flowers and receptacles from control and GA -treated fruit were sequenced 2, 4 Results and 10 DAT. After removing adaptors and low-quality Developmental characteristics and endogenous hormone sequences, each library generated 29.7 to 40.5 million changes upon GA treatment clean reads (Additional file 1: Table S2). Sequencing To optimize the GA treatment time window at stage I, quality compliance was indicated by an error ratio of fruits at pre-anthesis (15–20 mm in diameter), anthesis single bases of less than 0.0097%; Q20 and Q30 (the (20–25 mm) and post-anthesis (> 25 mm) stages were percentage of bases with base quality greater than 20 or 30 grouped and treated with GA .GA treatment at anthe- 3 3 among the total bases) were around 98.6 and 96.0%, re- sis gave a fruit-set rate of about 72% 42 DAT, much spectively. The number of G and C bases was about 47.0% higher than the pre- and post-anthesis stage treatments for each gene-expression library. Alignment of the clean (40 and 10%, respectively) (Fig. 1a). Anthesis-stage fruit reads to the reference sequence showed over 24.7 million with transverse diameter of 22 ± 1 mm were therefore total mapped reads, with the ratio varying from 82.11 to selected for further plant hormone and transcriptome 83.89% of the total clean reads (Additional file 1:Table S2). experiments. The growth curve of GA -induced par- Thirteen genes related to phytohormone metabolism thenocarpic fruit is shown in Fig. 1b; the typical or signal transduction were selected for validation. double-sigmoid curve pattern was in agreement with Positive correlations were found between the RNA se- previous reports [35]. However, fruit abscission was hap- quencing (RNA-Seq) and qRT-PCR data in all pairwise pened in control group at the beginning of stage II, comparison groups (Additional file 1: Figure S1 and S2), which was called “premature drop”. confirming the consistency, validity and representative- Exogenous GA treatment directly increased GA con- ness of our RNA-Seq data. centration in the female flowers, yielding about 26.8-, 38.8- and 28.7-fold its amount in controls 2, 4 and 10 DAT, respectively (Fig. 1c). Exogenous GA was also im- Differences in gene expression pattern between control mediately transported into the receptacle, with 15-fold and GA -treated fruit control GA levels in the treatment group 2 DAT. De- Comparative transcriptome analysis was performed to creasing trends of GA content were seen from 2 to 10 identify the differentially expressed genes (DEGs) in DAT in both control and GA -treated female flowers GA -treated samples compared to controls. Filtered by a 3 3 and receptacles, indicating the fig tissues’ strong capacity minimum 2-fold FPKM difference with FDR < 0.01, for to deplete this hormone (Fig. 1c). GA was present in fig female flower samples, pairwise comparisons revealed tissues at a low and stable level, and was insensitive to 685, 294 and 1811 DEGs at 2, 4 and 10 DAT, respect- GA treatment. The declining phase of native IAA in ively, with 400 (58.4%), 103 (35.0%) and 701 (38.7%) up- stage I female flowers was greatly accelerated by GA regulated and 255 (41.6%), 191 (65.0%) and 1110 (61.3%) treatment 2 and 4 DAT, but 10 DAT, IAA degradation downregulated in the GA -treated samples relative to ceased in the treated flowers and its concentration accu- controls; for receptacle samples, 908, 911 and 1672 mulated to almost 2-fold that in the controls. At this genes were differentially expressed at 2, 4 and 10 DAT, stage, the receptacle contained much less IAA than the respectively, with 436 (48.0%), 509 (55.9%) and 755 female flowers, and GA treatment only slightly changed (45.2%) upregulated and 472 (52.0%), 402 (44.1%) and 3 Chai et al. BMC Plant Biology (2018) 18:100 Page 5 of 15 Fig. 1 Physiological features of San Pedro-type fig main crop fruit after gibberellin (GA ) application. a Effect of application time window on fruit- set ratio. b Growth curves of control and GA -treated fruits (GA was applied at anthesis). c GA ,GA , indole acetic acid (IAA), zeatin and ABA 3 3 3 4 concentration in control and GA -treated female flowers and receptacle. Error bars indicate standard deviation. Significance analysis was performed by SPSS 17.0. *P ≤ 0.05, **P ≤ 0.01; FW, fresh weight 917 (54.8%) downregulated in the treated samples rela- the shared enriched terms in the two different tissues 2, 4 tive to controls (Additional file 1: Figure S3). and 10 DAT were: binding (GO:0005488) and catalytic ac- Significantly enriched terms in the GO and KEGG da- tivity (GO:0003824) terms in the molecular function cat- tabases were analyzed to identify the principle biological egory, cell part (GO:0044464) and cell (GO:0005623) functions of the DEGs. In female flowers, 38, 37 and 45 terms in the cellular component group, and metabolic significant GO terms of biological process, cellular com- process (GO:0008152), cellular process (GO:0009987) and ponent and molecular function were enriched 2, 4 and single-organism process (GO:0044699) terms in the bio- 10 DAT, respectively, and in the receptacle, 40, 39 and logical process group. 41 GO terms were enriched (Additional file 1: Table S3). DEGs identified 2, 4 and 10 DAT were assigned to 4, 1 In general, the enrichment patterns of up- and downregu- and 13 significantly different KEGG pathways in female lated genes were similar in female flowers and receptacles; flowers and 9, 4 and 12 pathways in the receptacle, Chai et al. BMC Plant Biology (2018) 18:100 Page 6 of 15 respectively (Additional file 1: Table S4). The largest With respect to GA-signal transduction, only one DEG number of significantly changed pathways was found 10 encoding DELLA protein was identified, which was more DAT in both tissues. Among the hormone-related path- highly expressed in GA -treated fruit than in controls; its ways, tryptophan metabolism (ko00380) was significantly upregulation was significant in GA -treated female flowers enriched in female flowers 10 DAT, and plant 2 DAT (Fig. 2d). These results suggest that exogenous hormone-signal transduction (ko04075) was significantly GA treatment mainly influences GA-biosynthesis and ca- enriched in receptacles 2 and 10 DAT. tabolism pathways in both female flowers and receptacle, but not the signal-transduction process. Expression pattern of genes involved in plant hormone Auxin metabolism and signal-transduction pathways In GA -treated fig, auxin-pathway genes constituted Plant hormone-related genes differentially expressed in the largest group of DEGs among growth- and female flowers and/or receptacles are summarized in development-promoting phytohormones (Additional Additional file 1: Tables S5 and S6. Briefly, most of the file 1: Table S5). The expression level of indole-3-ace- genes belonged to the ethylene-metabolism and response taldehyde dehydrogenase (IAld) remained stable in pathways, followed by auxin, ABA and GA; relatively female flowers and receptacle 2 and 4 DAT, whereas fewer DEGs were related to cytokinin. A FPKM value > 10 10 DAT, it was upregulated in the two control tissues, for at least one sample was set as the threshold to select to significantly higher levels than in their GA -treated genes for further analysis. The expression profiles of the counterparts (Fig. 3). Two auxin-responsive Gretchen rest of the hormone-related DEGs, with FPKM values of Hagen 3 (GH3) family genes revealed different expression all samples < 10, are shown in Additional file 1:Figure S4. patterns in the female flowers and similar patterns in the receptacles. In female flowers, comp25108_c0 showed Gibberellin continuously increasing expression during development in We identified 10 GA biosynthesis and catabolism genes control tissues, more than 2-fold higher than in the encoding GA 20-oxidase (GA20ox), GA 3-oxidase treated tissues 10 DAT; in contrast, the other GH3 (GA3ox) and GA 2-oxidase (GA2ox) (Additional file 1: (comp29694_c0) decreased from 2 to 10 DAT in female Table S5). The expression level of the only GA20ox identi- flowers, and showed no significant difference between fied in this study was dramatically higher in female flowers control and treated flowers. In receptacles, both GH3 than receptacles, and it was significantly upregulated in genes exhibited over 2-fold upregulation 2 DAT (Fig. 3). GA -treated female flowers 2 DAT compared to controls, Three IAA-amino acid hydrolase (IAH) genes demon- then downregulated until 10 DAT in both control and strated overall moderate changes in expression during de- treated samples (Fig. 2a). Two GA3ox genes showed dif- velopment—increasing from 2 to 4 DAT and then ferent expression patterns: comp22862_c0 was weakly decreasing in control female flowers and receptacle; how- expressed in all studied samples 2 and 4 DAT and showed ever, GA treatment significantly elevated IAH expression sharp upregulation in control female flowers and recep- 2 DAT, followed by a continuous decrease, resulting in tacle 10 DAT, to significantly higher levels compared to similar expression levels between control and treated tis- their GA -treated counterparts; comp12680_c0 showed sues 4 and 10 DAT. This implies that GA only transiently the highest expression level 2 DAT and then a decrease modulates IAH expression in female flowers and recepta- during development, and was generally downregulated in cles (Fig. 3). the two different tissues after GA treatment (Fig. 2b). With respect to the auxin-response pathway (Additional Among all GA2ox genes, comp16885_c0 was most file 1: Table S5), auxin-influx carrier (AUX1)demon- highly expressed, but its expression was repressed 3.3- strated higher expression in GA -treated fruit female and 41.1-fold in GA -treated female flowers and recep- flowers than in controls on all sampling days, and in the tacle compared to controls at 10 DAT, respectively. treated receptacle, upregulation was observed 2 DAT, Another GA2ox gene (comp26664_c0) also exhibited whereas downregulation was identified 10 DAT (Fig. 3). significantly decreased expression in GA -treated vs. Expression patterns of four small auxin-up RNA (SAUR) control samples at 10 DAT. However, comp32778_c0 ex- genes could be separated into two groups: DEGs pression was upregulated after treatment at all sampling comp31670_c0 and comp33019_c0 were significantly stages in both female flowers and receptacle, and re- repressed in GA -treated fruit female flowers and recep- markably higher expression was found in the receptacle tacle 10 DAT, whereas comp16559_c0 and comp36657_c0 vs. female flowers. In addition, comp18866_c0 only were generally upregulated in the treatment group on all showed a significant difference in expression between three sampling days relative to their control counterparts control and treated receptacles 2 DAT, with the latter (Fig. 3). The negative auxin-response regulator AUX/IAA being 7.5-fold lower than the former (Fig. 2c). revealed divergent expression patterns between female Chai et al. BMC Plant Biology (2018) 18:100 Page 7 of 15 Fig. 2 Expression patterns of GA-metabolism and signaling genes. a GA20ox. b GA3ox. c GA2ox. d DELLA. F, female flowers; R, receptacle. *Fold change of FPKM ≥2 and FDR ≤ 0.001 flowers and receptacle and between control and treatment expression in GA -treated fruit, whereas 10 DAT, it was groups. In general, AUX/IAA presented higher expression 14.7- and 8.6-fold higher in control vs. GA -treated female in GA -treated samples than in controls (Fig. 3). flowers and receptacle, respectively (Fig. 4). Note that in rice, the activity of cis-zeatin is comparable to that of Cytokinin trans-zeatin and can upregulate cytokinin-inducible genes, Differential expression of cytokinin metabolism and whereas overexpression of cis-ZOG delays rice leaf senes- signal-transduction genes was identified (Fig. 4,Additional cence [36]. file 1:Table S5). Isopentenyl transferase (IPT)expression In cytokinin-signal transduction, the cytokinin receptor was higher in GA -treated female flowers than controls 2 cytokinin response1(CRE1) was downregulated in both fe- DAT, but lower 4 and 10 DAT. On the other hand, in the male flowers and receptacle after GA treatment, espe- receptacle, IPT was downregulated 2 and 4 DAT, and cially 10 DAT (Fig. 4). The negative A-type Arabidopsis moderately upregulated 10 DAT. Compared to controls, response regulators (ARRs) remained stable with a low ex- cytokinin oxidase/dehydrogenase (CKX) was upregulated pression level in GA -treated fruit at all tested stages, in GA -treated samples, and its expression was remark- whereas a very significant increase was identified in con- ably higher in female flowers vs. receptacle. Cis-zeatin trols 10 DAT, presenting 16- and 31-fold upregulation in O-glucosyltransferase (CIS-ZOG) showed very low expres- female flowers and receptacle, respectively, compared to sion in all samples 2 and 4 DAT, and maintained this low the corresponding GA -treated tissues (Fig. 4). 3 Chai et al. BMC Plant Biology (2018) 18:100 Page 8 of 15 Fig. 3 Expression patterns of auxin-metabolism and signaling genes. Grids with 11 different colors represent FPKM values of 0–10, 10–20, 20–40, 40–60, 60–80, 80–100, 100–200, 200–300, 300–400, 400–500, and over 500, respectively. For each gene, the upper and lower rows refer to the results of control and GA treatment, respectively. 2, 4 and 10 = days after treatment. *Fold change of FPKM ≥2 and FDR ≤ 0.001 Fig. 4 Expression patterns of cytokinin-metabolism and signaling genes. F, female flowers; R, receptacle. *Fold change of FPKM ≥2 and FDR ≤ 0.001 Chai et al. BMC Plant Biology (2018) 18:100 Page 9 of 15 ABA they demonstrated higher expression after treatment than All identified ABA-synthesis, catabolism and response in controls. The other two SAM genes—comp30786_c0 and DEGs are shown in Additional file 1: Table S6. NECD comp7483_c0—increased during the control samples’ devel- (9-cisepoxycarotenoid dioxygenase) is the first commit- opment, but showed decreasing trends after GA treat- ted and rate-limiting enzyme in ABA biosynthesis [37]. ment, leading to 6.8- and 7.7-fold downregulation 10 DAT, Two NCED genes—comp17110_c0 and comp22655_c0— respectively, in the GA -treated receptacle (Fig. 6). One were enhanced in the both tissues 2 and 4 DAT with gene annotated as 1-aminocyclopropane-1-carboxylate GA (Fig. 5a). In controls, their expression was relatively (ACC) synthase (ACS) presented higher expression in stable during development in female flowers, and rose GA -treated samples than in controls. ACC oxidase (ACO) first and then fell at the 10 DAT in the receptacle; how- genes, except comp23635_c1, had low expression in all ever, decreasing trends were found in both female tested control and GA -treated samples 2 and 4 DAT, flowers and receptacle of GA -treated fruit. The other whereas dramatic upregulation was detected in control fruit two NCED—comp19377_c0 and comp26438_c0—gener- female flowers and receptacle 10 DAT, resulting in signifi- ally increased in the two tissues of control fruit during cantly higher expression in the two control vs. treated tis- development, especially from 4 to 10 DAT, whereas their sues (2.5- to 42-fold) (Fig. 6). levels changed moderately in GA -treated tissues; ultim- In the ethylene-response pathway, ETR transcripts in- ately, their expression was significantly lower in treat- creased during development, showing the highest ex- ment vs. control tissues 10 DAT (Fig. 5a). ABA2, pression levels 10 DAT in controls. For GA -treated encoding another ABA-biosynthesis enzyme, was down- samples, their expression levels fell first and then rose, regulated in GA -treated vs. control tissues, but with no with a peak at 2 or 10 DAT (Fig. 6). The expression significant difference in the female flowers (Fig. 5b). The levels of ETR and EBF1/2 were significantly upregulated process catalyzed by ABA 8′-hydroxylase (ABA 8′-h) is in GA -treated female flowers and receptacle 2 DAT. In considered to be the main pathway for ABA catabolism general, ERF transcripts revealed an increasing trend in [37]. ABA 8′-h expression remained relatively stable in control fruit female flowers and receptacle, whereas they the controls, whereas remarkable upregulation was presented divergent expression patterns after GA treat- found in the GA -treated samples 10 DAT—4.9- and ment: comp16781_c0 and comp16998_c0 had the highest 3.3-fold higher in female flowers and receptacle, respect- expression 2 DAT in female flowers and receptacle; ively, than in controls (Fig. 5c). comp28870_c0 and comp17108_c0 displayed upward In the ABA-signaling pathway, PYR/PYL expression was trends during development similar to their expression in slightly lower in GA -treated samples than in controls 2 the controls (Fig. 6). ERFs were generally upregulated in and 4 DAT, but significantly higher 10 DAT (Fig. 5d). the two treatment tissues 2 DAT, whereas 10 DAT, their Protein phosphatase 2C (PP2C) usually represses ABA re- expression levels were slightly lower in GA -treated vs. sponses. PP2C expression was not significantly different control receptacles. between the control and treated fruit 2 and 4 DAT, whereas it was dramatically downregulated in GA -treated Discussion female flowers and receptacle 10 DAT (Fig. 5e). ABRE Our aim was to investigate the role of GA in partheno- binding factor (ABF) functions as a positive regulator of carpy induction in fig by comparing control and ABA signaling. The ABF genes’ expression was consistent GA -treated fruit’s female flowers and receptacles. We with an increasing tendency from 2 to 10 DAT in both focused our study on 2, 4 and 10 days after exogenous control and GA -treated samples, and their expression GA application in a San Pedro-type fig cultivar’s main 3 3 was more or less downregulated in all studied samples crop, which is genetically non-parthenocarpic. By Illu- after GA treatment (Fig. 5f). mina RNA-Seq transcriptome analysis, 1.3- and 3.1-fold more DEGs were identified in the receptacle vs. female Ethylene flowers at 2 and 4 DAT, respectively, whereas the num- A large number of ethylene biosynthesis- and signaling-related ber of DEGs was a little higher in female flowers than in DEGs were found in female flowers and/or recepta- the receptacle 10 DAT. This implied that aside from the cles after GA treatment (Additional file 1:Table S6). reproductive female flowers, which mainly influence S-adenosylmethionine synthetase (SAM) displayed the fruit development [38], the vegetative receptacle tissue highest expression 2 DAT in control and GA -treated might also have some regulatory effect on partheno- female flowers, followed by a gradual decrease, with no sig- carpic fruit set and growth. GA, auxin, cytokinin, ABA nificant differences between control and treatment (Fig. 6). and ethylene metabolism- and response-related genes Three SAM genes—comp28816_c0, comp11492_c0 and showed differential expression between control and comp28987_c0—peaked 4 DAT and showed their lowest GA -treated fruit, and phytohormone content assay re- expression 10 DAT in control and GA -treated receptacles; vealed a new balance of endogenous phytohormone 3 Chai et al. BMC Plant Biology (2018) 18:100 Page 10 of 15 Fig. 5 Expression patterns of ABA-metabolism and signaling genes. a NCED. b ABA2. c ABA 8′-h. d PYR/PYL. e PP2C. f ABF. F, female flowers; R, receptacle. *Fold change of FPKM ≥2 and FDR ≤ 0.001 levels in both female flowers and receptacle following Exogenous GA application reset the transcriptional exogenous GA treatment (Fig. 1c). This indicated that expression of plant hormone-related genes. In female parthenocarpy induction by GA in fig is the result of the flowers, 4, 1 and 15 phytohormone-metabolism genes coordinated action of different phytohormones. and 8, 2 and 12 response-related genes were significantly Chai et al. BMC Plant Biology (2018) 18:100 Page 11 of 15 Fig. 6 Expression patterns of ethylene-biosynthesis and signaling genes. Grids with 11 different colors represent FPKM values 0–10, 10–20, 20–40, 40–60, 60–80, 80–100, 100–200, 200–300, 300–400, 400–500 and over 500, respectively. For each gene, upper and lower rows refer to the results of control and GA treatment, respectively. 2, 4 and 10 = days after treatment. *Fold change of FPKM ≥2 and FDR ≤ 0.001 differently expressed 2, 4 and 10 DAT, respectively. In the tissue and developmental stage [39, 41]. Exogenous receptacle, 19, 4 and 18 phytohormone-metabolism and GA application before or after bloom in grapevine 10, 2 and 12 signal-transduction DEGs were identified 2, 4 leads to downregulation of the GA-biosynthetic genes and 10 DAT, respectively. The higher representation of GA20ox and GA3ox and upregulation of the GA-catabolic plant hormone transcripts’ differential expression in re- gene GA2ox [15, 42, 43]; this has also been observed in ceptacles compared to female flowers may further suggest Arabidopsis [39] and tobacco [44]. Conversely, in the present an indelible role for receptacle tissue in controlling fruit study, GA20ox was upregulated in female flowers 2 DAT and set and development. Concrete spatiotemporal transcrip- GA2ox was repressed in both female flowers and receptacle tional changes in phytohormone-related genes induced by 10 DAT. This difference could be explained by the change in GA are summarized in Fig. 7. There were more common endogenous GA content: in both fig and grapevine, GA DEGs with the same changing trends in female flowers content increases significantly and immediately after treat- and receptacles 10 DAT than 2 and 4 DAT, indicating that ment and then decreases; however, for grapevine, GA con- GA treatment induces synchronized patterns of changing tent drops to levels equal to or significantly less than the gene expression between the two tissues as fruit develop- control 3 DAT [43, 45], whereas in fig, it is still significantly ment progresses. In addition, in both female flowers and higher than the control 10 DAT (Fig. 1c). receptacles 10 DAT, the number of DEGs related to ethyl- Auxin is one of the main regulators of fruit set and devel- ene metabolism was highest, followed by those related to opment. Similar to [42], auxin-biosynthesis genes, e.g. GA and ABA metabolism; for phytohormone response, YUCCA, did not show significant regulation after GA the highest number of DEGs were related to auxin. treatment in fig. The main differentially expressed auxin-metabolism genes identified in the present study Phytohormone-metabolism response to exogenous GA were IAH and GH3. The former encodes enzymes that pro- Previous studies have shown that transcription of duce free IAA from IAA–amino acid conjugates, whereas GA-metabolism genes is regulated by the content of the latter catalyzes auxin conjugation [46]. IAH and GH3 bioactive GA [2, 39]. GA oxidases modulate the GA-biosynthetic were mainly significantly upregulated 2 DAT; upregulation pathway mainly through feedback-loop mechanisms of GH3 has also been reported previously [42], indicating [40] and expression regulation that depends on the that bioactive auxin levels are co-mediated by IAH and Chai et al. BMC Plant Biology (2018) 18:100 Page 12 of 15 Fig. 7 Summary of main differentially expressed phytohormone-related genes identified in female flowers and receptacle after GA treatment. Different colored circles represent different phytohormones and circle size indicates relative number of DEGs—the higher the number, the bigger the circle. Genes in red (green) color are upregulated (downregulated). Gray highlight represents genes with same changing trend in both female flowers and receptacle following GA application GH3 in fig. Cytokinin induces parthenocarpy by altering expression of the ethylene-biosynthesis genes ACOs de- the expression of GA-biosynthesis genes [9, 16]. Onlyafew creases during parthenocarpic fruit set and development metabolic genes’ transcriptional levels were regulated by or after GA treatment [13]. GA treatment in our dataset. For female flowers, cytokinin-metabolism transcripts were only mediated by GA treatment 10 DAT, whereas for the receptacle, exogen- Phytohormone signaling in response to exogenous GA 3 3 ous GA also regulated cytokinin metabolism 2 and 4 DAT DELLA protein has been characterized as a negative regu- (Fig. 7). Zeatin concentration increased dramatically in the lator in GA signaling [48]. One DELLA gene exhibited up- receptacle 2 DAT, then decreased to a markedly lower level regulation after GA application, most markedly in female than controls, suggesting that cytokinin-biosynthesis genes flowers 2 DAT, in accordance with previously reported re- respond more rapidly (in under 2 days) in the receptacle sults [7, 13, 43]. This indicates GA-perception feedback following GA application. regulation in female flowers and receptacle. Aside from growth-promoting hormones, GA treat- Studies of the crosstalk between GA and auxin have ment also modulated the biosynthesis of ABA and ethyl- revealed that GA-induced parthenocarpy arises from the ene, which mainly play GA-antagonistic roles in the interaction of GA with auxin signaling [42], and that control of many plant developmental processes [47]. auxin-induced fruit set is mediated in part by GA [9]. In Enhancement of ABA and ethylene biosynthesis 2 DAT the present study, AUX/IAA gene expression was signifi- might reflect ‘feedback’ regulation of the highly signifi- cantly upregulated in the receptacle on all sampling days cantly increased GA content in fig fruit following and in female flowers 10 DAT, suggesting feedback regu- exogenous GA application. This ‘feedback’ response lation of the auxin response following GA treatment, in 3 3 then declines, with almost no ABA- or ethylene-related agreement with results in grape [15]. In addition to the DEGs being identified 4 DAT. Along with fruit develop- AUX/IAAs, differential expression of SAUR transcript ment, both female flowers and receptacle revealed sup- levels was seen in both female flowers and receptacles pression of ABA and ethylene biosynthesis 10 DAT, with 10 DAT (Fig. 3). Further study is needed to elucidate downregulation of NCED and ACO and upregulation of their roles in fig fruit development. DEGs related to ABA 8′-h, in accordance with the lower ABA level in cytokinin signaling were only identified 10 DAT in the GA -treated fruits. Previous reports have shown that female flowers and receptacle (Fig. 7). 3 Chai et al. BMC Plant Biology (2018) 18:100 Page 13 of 15 GA treatment also changed the signaling of ABA and control and GA -treated female flowers and receptacles. DAT, days after ethylene. Enhancement of ethylene signaling was identi- treatment, Figure S4. Heat maps of hormone-related genes with low expression. FPKM of all samples was < 10. (PDF 1241 kb) fied 2 DAT. Crosstalk internodes between GA and the ethylene-signaling pathway were elucidated by the role Abbreviations of group VII ERFs as DELLA partners [49], and ERF11 ABA 8′-h: ABA 8′-hydroxylase; ABA: Abscisic acid; ABF: ABRE binding factor; activates GA biosynthesis and signaling [50]. Genes en- ACC: 1-aminocyclopropane-1-carboxylate; ACS: ACC synthase; ACO: ACC coding ABA receptor PYR/PYL and PP2C, a negative oxidase; ARRs: Arabidopsis response regulators; AUX1: auxin-influx carrier; CIS- ZOG: cis-zeatin O-glucosyltransferase; CKX: Cytokinin oxidase/dehydrogenase; regulator in the ABA-response pathway, displayed up- CRE1: Cytokinin response1; DAT: Days after treatment; DEGs: Differentially regulation and downregulation, respectively, in both tis- expressed genes; FDR: False discovery rate; FPKM: Fragments per kilobase of sues 10 DAT, leading to enhancement of ABA-signal exon model per million mapped reads; GA20ox: GA 20-oxidase; GA2ox: GA 2-oxidase; GA3ox: GA 3-oxidase; GAs: Gibberellins; GH3: Gretchen Hagen 3; transduction in both female flowers and receptacle. A GO: Gene Ontology; IAA: Indole-3-acetic acid; IAH: IAA-amino acid hydrolase; recent study on rice revealed the central role of PYR/ IAld: Indole-3-acetaldehyde dehydrogenase; IPT: Isopentenyl transferase; PYL in the antagonistic action of GA and ABA [51]. KEGG: Kyoto Encyclopedia of Genes and Genomes; NECD: 9- cisepoxycarotenoid dioxygenase; PP2C: Protein phosphatase 2C; qRT- PCR: Quantitative real-time PCR; RNA-Seq: RNA sequencing; SAM: S- Conclusions adenosylmethionine synthetase; SAUR: Small auxin-up RNA GA treatment, performed to induce parthenocarpy in San Funding Pedro-type fig main crop, resulted in a highly significant This work was financially supported by the grants from National Natural GA content increment in both female flowers and recep- Science Foundation of China project NSFC [31372007]. tacle tissues of fig fruit. Changes in the expression of key Availability of data and materials genes in plant hormone-synthesis and metabolism Raw data of all sample-sequencing results have been submitted to the NCBI pathways—as reflected by modulated endogenous plant Sequence Read Archive (SRA, http://www.ncbi.nlm.nih.gov/Traces/sra) with accession number SRP113799. hormone levels—and in the hormone-signaling pathway, converted the pre-dropping fruit pattern to an artificial par- Authors’ contributions thenocarpic developmental pattern. Spatiotemporal charac- HM and SC designed the experiments. LC and PC conducted the experiments and analyzed the results. LC and HM prepared the manuscript. teristics of the changes in plant hormone levels and SC and MF assisted in drafting the manuscript. All authors read and gene-expression patterns in GA -induced fig parthenocarpy approved the final manuscript. were reflected by differentially reacting female flowers and receptacle tissue, and temporal waves in major plant hor- Ethics approval and consent to participate The experiments did not involve endangered or protected species. No mone and relevant genes’ expression were illustrated. specific permits were required for these activities because the figs used in Further study on fig fruit development with the application this study were obtained from an orchard in Beijing, which was a of other exogenous plant hormones, e.g., auxin and cytoki- demonstration base of China Agricultural University. nin, will serve to further understand the mechanisms of Competing interests GA-induced parthenocarpy and the key plant hormones in The authors declare that they have no competing interests. fig fruit set and development. Author details College of Horticulture, China Agricultural University, Beijing, People’s Additional file Republic of China. College of Beijing Laboratory for Food Quality and Safety, College of Food Science and Nutritional Engineering, China Agricultural University, Beijing, People’s Republic of China. Department of Fruit Tree Additional file 1: Table S1. Primer sequences of genes used for Sciences, Agricultural Research Organization, Volcani Center, Bet-Dagan, validation of RNA-Seq results by quantitative real-time PCR, Table S2. Israel. Summary of sequencing results for control and GA -treated female flowers and receptacles, Table S3. Significant GO terms (corrected Received: 9 September 2017 Accepted: 24 May 2018 P-value ≤0.05) identified between control and GA -treated female flowers and receptacles, Table S4. Significant KEGG pathways (corrected P-value ≤0.05) identified between control and GA -treated female flowers and References receptacles, Table S5. All gibberellin-, auxin- and cytokinin-synthesis, 1. Blazquez MR, Nilsson O, Sussman MR, Weigel D. Gibberellins promote catabolism and response transcripts identified in this study which were flowering of Arabidopsis by activating the LEAFY promoter. Plant Cell. differentially expressed (FDR < 0.01 and the absolute value of log 1998;10(5):791. (FPKM /FPKM ) ≥ 1) in at least one pairwise comparison group, treatment control 2. Serrani JC, Sanjuan R, Ruizrivero O, Fos M, Garciamartinez JL. Gibberellin Table S6. All abscisic acid- and ethylene-synthesis, catabolism and re- regulation of fruit set and growth in tomato. Plant Physiol. 2007;145(1): sponse transcripts identified in this study which were differentially 246–57. expressed (FDR < 0.01 and the absolute value of log (FPKM / 2 treatment 3. Ayele BT, Ozga JA, Reinecke DM. Regulation of GA biosynthesis genes FPKM ) ≥ 1) in at least one pairwise comparison group, Figure S1. control during germination and young seedling growth of pea (Pisum sativum Verification of RNA-Seq results by qRT-PCR. Bars represent standard L.). J Plant Growth Regul. 2006;25(3):219–32. deviation. F, female flowers; R, receptacle, Figure S2. Correlation of fold 4. Tyler L, Thomas SG, Hu J, Dill A, Alonso JM, Ecker JR, Sun T. DELLA proteins changes in gene expression between RNA-Seq and qRT-PCR. Equation of 2 and gibberellin-regulated seed germination and floral development in linear regression and correlation coefficient (R ) are shown. DAT, days Arabidopsis. Plant Physiol. 2004;135(2):1008–19. after treatment, Figure S3. Number of differentially expressed genes 5. Hedden P, Phillips AL. Gibberellin metabolism: new insights revealed by the (FDR ≤ 0.001 and log (FPKM /FPKM ) ≥ 1or≤− 1) between 2 treatment control genes. Trends Plant Sci. 2000;5(12):523–30. Chai et al. BMC Plant Biology (2018) 18:100 Page 14 of 15 6. Yarushnykov VV, Blanke MM. Alleviation of frost damage to pear flowers by 28. Langmead B, Salzberg SL. Fast gapped-read alignment with bowtie 2. Nat application of gibberellin. Plant Growth Regul. 2005;45(1):21–7. Methods. 2012;9(4):357. 7. Serrani JC, Ruizrivero O, Fos M, Garciamartinez JL. Auxin-induced fruit-set in 29. Robinson MD, Mccarthy DJ, Smyth GK. edgeR: a Bioconductor package for tomato is mediated in part by gibberellins. Plant J. 2008;56(6):922–34. differential expression analysis of digital gene expression data. Bioinformatics. 2010;26(1):139. 8. Mesejo C, Yuste R, Reig C, Martínez-Fuentes A, Iglesias DJ, Muñoz-Fambuena N, Bermejo A, Germanà MA, Primo-Millo E, Agustí M. Gibberellin reactivates 30. Xie C, Mao X, Huang J, Ding Y, Wu J, Dong S, Kong L, Gao G, Li C, Wei L. and maintains ovary-wall cell division causing fruit set in parthenocarpic KOBAS 2.0: a web server for annotation and identification of enriched Citrus species. Plant Sci. 2016;247:13–24. pathways and diseases. Nucleic Acids Res. 2011;39 9. Lu L, Liang J, Zhu X, Xiao K, Li T, Hu J. Auxin- and Cytokinin-induced berries 31. Upton G. Fisher's exact test. J R Stat Soc A Sta. 1992;155:395–402. set in grapevine partly rely on enhanced gibberellin biosynthesis. Tree 32. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a Genet Genomes. 2016;12(3):41. practical and powerful approach to multiple testing. J R Stat Soc B. 10. Shinozaki Y, Hao S, Kojima M, Sakakibara H, Ozekiiida Y, Zheng Y, Fei Z, 1995;57(1):289–300. Zhong S, Giovannoni JJ, Rose JKC. Ethylene suppresses tomato (Solanum 33. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using lycopersicum) fruit set through modification of gibberellin metabolism. real-time quantitative PCR and the 2− ΔΔCT method. Methods. 2001; Plant J. 2015;83(2):237–51. 25(4):402–8. 11. Nitsch L, Kohlen W, Oplaat C, Charnikhova T, Cristescu S, Michieli P, Wolters- 34. Freiman ZE, Rosianskey Y, Dasmohapatra R, Kamara I, Flaishman MA. The Arts M, Bouwmeester H, Mariani C, Vriezen WH. ABA-deficiency results in ambiguous ripening nature of the fig (Ficus carica L.) fruit: a gene- reduced plant and fruit size in tomato. J Plant Physiol. 2012;169(9):878–83. expression study of potential ripening regulators and ethylene-related 12. Mariotti L, Picciarelli P, Lombardi L, Ceccarelli N. Fruit-set and early fruit genes. J Exp Bot. 2015;66(11):3309. growth in tomato are associated with increases in Indoleacetic acid, 35. Marei N, Crane JC. Growth and respiratory response of fig (Ficus carica L. cv. Cytokinin, and bioactive gibberellin contents. J Plant Growth Regul. Mission) fruits to ethylene. Plant Physiol. 1971;48(3):249–54. 2011;30(4):405. 36. Kudo T, Makita N, Kojima M, Tokunaga H, Sakakibara H. Cytokinin activity of 13. Tang N, Deng W, Hu G, Hu N, Li Z. Transcriptome profiling reveals the cis-zeatin and phenotypic alterations induced by overexpression of putative regulatory mechanism underlying pollination dependent and cis-Zeatin-O-glucosyltransferase in rice. Plant Physiol. 2012;160(1):319. parthenocarpic fruit set mainly mediated by auxin and gibberellin. PLoS 37. Nambara E, Marionpoll A. Abscisic acid biosynthesis and catabolism. Annu One. 2015;10(4) Rev Plant Biol. 2005;56(1):165–85. 14. Ji L, Zhe W, Li C, Tinglin Z, Qinwei G, Jian X, Li J, Qunfeng L, Sanwen H, 38. RosianskiY,Doronfaigenboim A,Freiman ZE,LamaK,Milocochavi S, Zhengguo L. Transcriptome comparison of global distinctive features Dahan Y, Kerem Z, Flaishman MA. Tissue-specific transcriptome and between pollination and parthenocarpic fruit set reveals transcriptional hormonal regulation of pollinated and Parthenocarpic fig (Ficus carica phytohormone cross-talk in cucumber (Cucumis sativus L.). Plant Cell L.) fruit suggest that fruit ripening is coordinated by the reproductive Physiol. 2014;55(7):1325. part of the Syconium. Front Plant Sci. 2016;7 15. Chai L, Li Y, Chen S, Perl A, Zhao F, Ma H. RNA sequencing reveals high 39. Rieu I, Ruizrivero O, Fernandezgarcia N, Griffiths J, Powers SJ, Gong F, resolution expression change of major plant hormone pathway genes Linhartova T, Eriksson S, Nilsson O, Thomas SG. The gibberellin biosynthetic after young seedless grape berries treated with gibberellin. Plant Sci. genes AtGA20ox1 and AtGA20ox2 act, partially redundantly, to promote 2014;229:215–24. growth and development throughout the Arabidopsis life cycle. Plant J. 16. Ding J, Chen B, Xia X, Mao W, Shi K, Zhou Y, Yu J. Cytokinin-induced 2008;53(3):488. parthenocarpic fruit development in tomato is partly dependent on 40. Middleton AM, Ubedatomas S, Griffiths J, Holman T, Hedden P, Thomas SG, enhanced gibberellin and auxin biosynthesis. PLoS One. 2013;8(7):e70080. Phillips AL, Holdsworth MJ, Bennett MJ, King JR. Mathematical modeling 17. Mignolli F, Vidoz ML, Mariotti L, Lombardi L, Picciarelli P. Induction of elucidates the role of transcriptional feedback in gibberellin signaling. Proc gibberellin 20-oxidases and repression of gibberellin 2β-oxidases in Natl Acad Sci U S A. 2012;109(19):7571–6. unfertilized ovaries of entire tomato mutant, leads to accumulation of 41. Yamaguchi S, Kamiya Y, Sun T. Distinct cell-specific expression patterns of active gibberellins and parthenocarpic fruit formation. Plant Growth early and late gibberellin biosynthetic genes during Arabidopsis seed Regul. 2015;75(2):415–25. germination. Plant J. 2001;28(4):443–53. 18. Wang H, Jones B, Li Z, Frasse P, Delalande C, Regad F, Chaabouni S, 42. Jung CJ, Hur YY, Jung S-M, Noh J-H, Do G-R, Park S-J, Nam J-C, Park K-S, Latche A, Pech J, Bouzayen M. The tomato aux/IAA transcription factor Hwang H-S, Choi D, et al. Transcriptional changes of gibberellin oxidase IAA9 is involved in fruit development and leaf morphogenesis. Plant genes in grapevines with or without gibberellin application during Cell. 2005;17(10):2676–92. inflorescence development. J Plant Res. 2014;127(2):359–71. 19. Tiwari A, Offringa R, Heuvelink E. Auxin-induced fruit set in Capsicum 43. Cheng C, Jiao C, Singer SD, Gao M, Xu X, Zhou Y, Li Z, Fei Z, Wang Y, Wang annuum L. requires downstream gibberellin biosynthesis. J Plant Growth X. Gibberellin-induced changes in the transcriptome of grapevine ( Vitis Regul. 2012;31(4):570–8. labrusca × V. Vinifera ) cv. Kyoho flowers. BMC Genomics. 2015;16(1):128. 20. Flaishman MA, Rodov V, Stover E. The fig: botany, horticulture, and 44. Gallegogiraldo L, Ubedatomás S, Gisbert C, Garcíamartínez JL, Moritz T, breeding. Horticultural Rev-Westport Then N Y. 2008;34:113. Lópezdíaz I. Gibberellin homeostasis in tobacco is regulated by gibberellin 21. Beck NG, Lord EM. Breeding system in Ficus carica, the common fig. II. metabolism genes with different gibberellin sensitivity. Plant Cell Physiol. Pollination events. Am J Bot. 1988;75(12):1913–22. 2008;49(5):679. 22. Crane J, Campbell R. Breaking rest and inducing parthenocarpy in the 45. Wang ZR, Zhao F, Zhao XB, Ge H, Chai L, Chen S, Perl A, Ma H. Proteomic Calimyrna fig with gibberellin: Proc 15th Int Cong Hort, Nice; 1958. analysis of berry-sizing effect of GA3 on seedless Vitis vinifera L. Proteomics. 23. Crane JC, Van Overbeek J. Kinin-induced parthenocarpy in the fig, Ficus 2012;12(1):86–94. carica L. Science. 1965;147(3664):1468–9. 46. Rampey RA, LeClere S, Kowalczyk M, Ljung K, Sandberg G, Bartel B. A family 24. Lodhi F, Bradley MV, Crane JC. Auxins and gibberellin-like substances in of auxin-conjugate hydrolases that contributes to free indole-3-acetic acid parthenocarpic and non-parthenocarpic syconia of Ficus carica L., cv. King. levels during Arabidopsis germination. Plant Physiol. 2004;135(2):978–88. Plant Physiol. 1969;44(4):555–61. 47. Weiss D, Ori N. Mechanisms of cross talk between gibberellin and other 25. Chai L, Wang Z, Chai P, Chen S, Ma H. Transcriptome analysis of San Pedro- hormones. Plant Physiol. 2007;144(3):1240–6. type fig (Ficus carica L.) parthenocarpic breba and non-parthenocarpic main 48. Sun T. Gibberellin-GID1-DELLA: a pivotal regulatory module for plant crop reveals divergent phytohormone-related gene expression. Tree Genet growth and development. Plant Physiol. 2010;154(2):567–70. Genomes. 2017;13(4):83. 49. Marínde lRN, Sotillo B, Miskolczi P, Gibbs DJ, Vicente J, Carbonero P, 26. Pan X, Welti R, Wang X. Quantitative analysis of major plant hormones in Oñatesánchez L, Holdsworth MJ, Bhalerao R, Alabadí D. Large-scale crude plant extracts by high-performance liquid chromatography-mass identification of gibberellin-related transcription factors defines group VII spectrometry. Nat Protoc. 2010;5(6):986. ETHYLENE RESPONSE FACTORS as functional DELLA partners. Plant Physiol. 27. Reid KE, Olsson N, Schlosser J, Peng FY, Lund ST. An optimized grapevine 2014;166(2):1022–32. RNA isolation procedure and statistical determination of reference genes for 50. Zhou X, Zhang Z, Park J, Tyler L, Yusuke J, Qiu KQ, Nam E, Lumba S, real-time RT-PCR during berry development. BMC Plant Biol. 2006;6(1):27. Desveaux D, Mccourt P. The ERF11 transcription factor promotes internode Chai et al. BMC Plant Biology (2018) 18:100 Page 15 of 15 elongation by activating gibberellin biosynthesis and signaling. Plant Physiol. 2016;171(4):2760–70. 51. Lin Q, Wu F, Sheng P, Zhang Z, Zhang X, Guo X, Wang J, Cheng Z, Wang J, Wang H. The SnRK2-APC/CTE regulatory module mediates the antagonistic action of gibberellic acid and abscisic acid pathways. Nat Commun. 2015;6:7981. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Plant Biology Springer Journals

Transcriptome analysis unravels spatiotemporal modulation of phytohormone-pathway expression underlying gibberellin-induced parthenocarpic fruit set in San Pedro-type fig (Ficus carica L.)

Free
15 pages

Loading next page...
 
/lp/springer_journal/transcriptome-analysis-unravels-spatiotemporal-modulation-of-3Sp3PU9EoA
Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s).
Subject
Life Sciences; Plant Sciences; Agriculture; Tree Biology
eISSN
1471-2229
D.O.I.
10.1186/s12870-018-1318-1
Publisher site
See Article on Publisher Site

Abstract

Background: Gibberellin (GA) treatments can induce parthenocarpy in the main crop of San Pedro-type figs, the native non-parthenocarpic fruit, however, the underlying mechanism is still largely unclear. Results: In our study, GA was applied to San Pedro-type fig main crop at anthesis. Sharply increased GA content 3 3 was detected in both female flowers and receptacle, along with significantly decreased indole-3-acetic acid (IAA), zeatin and abscisic acid (ABA) levels in female flowers, and increased zeatin peak intensity and earlier ABA peak in receptacles. Transcriptome comparison between control and treatment groups identified more differentially expressed genes (DEGs) in receptacles than in female flowers 2 and 4 days after treatment (DAT); 10 DAT, the number of DEGs became similar in the two tissues. Synchronized changing trends of phytohormone-associated DEGs were observed in female flowers and receptacles with fruit development. Modulation of ethylene and GA signaling and auxin metabolism by exogenous GA occurred mainly 2 DAT, whereas changes in auxin, cytokinin and ABA signaling occurred mainly 10 DAT. Auxin-, ethylene- and ABA-metabolism and response pathways were largely regulated in the two tissues, mostly 2 and 10 DAT. The major components altering fig phytohormone metabolic and response patterns included downregulated GA2ox, BAS1, NCED and ACO, and upregulated ABA 8′-h and AUX/IAA. Conclusions: Thus GA-induced parthenocarpy in fig is co-modulated by the female flowers and receptacle, and repression of ABA and ethylene biosynthesis and GA catabolism might be the main forces deflecting abscission and producing fig parthenocarpy. Keywords: Ficus carica, Gibberellin treatment, Parthenocarpy, Transcriptome analysis, Plant hormone * Correspondence: hqma@cau.edu.cn Lijuan Chai and Peng Chai contributed equally to this work. College of Horticulture, China Agricultural University, Beijing, People’s Republic of China Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. 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. Chai et al. BMC Plant Biology (2018) 18:100 Page 2 of 15 Background of GA metabolism [17]. SlIAA9, a member of the Aux/ Gibberellins (GAs) are involved in almost every aspect IAA family, is an auxin-signaling repressor that prevents of plant growth and development. In plant reproductive tomato ovary development [18]. Moreover, fruit set in to- processes, GAs regulate floral initiation [1], fruit set and mato induced by auxin analogue 2,4-D was significantly growth [2], and seed maturation and germination [3, 4]. suppressed by GA-biosynthesis inhibitors [7]. Similar to Among the abundant GAs identified to date, only GA , the situation in Arabidopsis and tomato, auxin is the GA ,GA and GA have been recognized as bioactive major inducer of fruit set in Capsicum annuum, depend- 3 4 7 [5], and only GA ,GA and/or GA have been success- ing in part on enhanced GA biosynthesis [19]. Ethylene is 3 4 7 fully applied for parthenocarpic fruit induction [6]. also considered to be a key regulator in coordinating fruit Parthenocarpy, a highly valuable agronomic character- set, and ethylene reduces fruit set by suppressing GA me- istic, especially in edible fruit crops, means that fruit set tabolism [10]. and development can be uncoupled from pollination Ficus carica L. (Moraceae) bears urn-shaped fruit with and fertilization and thus the fruit is seedless, such as in an enclosed inflorescence structure termed syconium. fig, banana, persimmon, pineapple and pear [6]. There are two major sex types in fig trees: caprifig (male Parthenocarpic fruit can also be triggered by application fig) and female fig; the latter is further classified into of exogenous plant hormones, termed artificial par- three types, common, Smyrna and San Pedro, based on thenocarpy. Auxin and GAs are the most frequently cropping characteristics [20]. Parthenocarpic or used hormones in parthenocarpic fruit induction [7, 8]. pollinated non-parthenocarpic fig fruit show a typical Other plant hormones, including cytokinin [9], ethylene double-sigmoid growth curve, with two rapid [10] and abscisic acid (ABA) [11], also regulate or par- fruit-size-increment phases, i.e., stages I and III, sepa- ticipate in parthenocarpic fruit set and growth. rated by a lag phase (stage II) [21]. Every year, San Pedro Parthenocarpy requires the unique and coordinated action fig produces a parthenocarpic first crop called breba, of different phytohormones, whereas non-parthenocarpic and a non-parthenocarpic main crop which constitutes fruit will senescence and abscise if fertilization is not the main yield [20]. Studies on non-parthenocarpic achieved. Auxin, GA and cytokinin are generally recognized Smyrna fig cv. Calimyrna revealed that parthenocarpic as the major regulators of fruit set. Elevated levels of these fruit set can be induced through auxin, GA and cytoki- hormones in tomato ovaries have been associated with fruit nin treatments [22, 23]. Analysis of auxin and GA setand early fruitgrowth [12]. Auxin and GA, and the cross- content in the two crops of San Pedro fig cv. King dem- talk between them, largely modulate pollination-dependent onstrated higher auxin content in the parthenocarpic and parthenocarpic fruit set in tomato [13]. Transcriptome breba than the pollinated non-parthenocarpic main crop, comparisons between pollinated and parthenocarpic cucum- whereas GA was higher in the latter than in the ber fruit have shown crosstalk of auxin, GA and cytokinin at former [24]. Our previous studies revealed that compared the transcriptional level during parthenocarpic fruit set [14]. with San Pedro parthenocarpic breba, expression of GA- Exogenous GA application significantly modulated the and auxin-biosynthesis genes is repressed in its expression patterns of plant hormone metabolism and non-parthenocarpic main crop female flowers, whereas signaling genes in seedless grape, as determined by transcrip- ABA- and ethylene-biosynthesis genes are enhanced [25]. tome analysis [15]. Previous studies in several plants have Though exogenous hormone treatment can induce fruit demonstrated that parthenocarpy induced by auxin and set and development of non-parthenocarpic fig, our un- cytokinin requires downstream GA biosynthesis. Application derstanding of the underlying mechanisms is limited. of exogenous auxin and cytokinin in grape increased the ex- In this study, the effect of exogenous GA treatment on pression of GA-biosynthesis genes, such as VvGA3ox1,and the main crop of San Pedro fig was studied, and hormone repressed GA-catabolism genes, such as VvGA2ox3 and contents of the female flowers and receptacle were VvGA2ox4, suggesting that auxin- and cytokinin-induced assayed. A corresponding global transcriptome compari- parthenocarpic fruit set requires enhanced GA biosynthesis son demonstrated a significant number of gene ontology [9]. Parthenocarpy induction in tomato by cytokinin depends (GO) terms and Kyoto Encyclopedia of Genes and in part on modulation of GA and indole-3-acetic acid (IAA) Genomes (KEGG) pathways with spatiotemporal regula- metabolism, as reflected by the upregulated expression tion characteristics. The expression patterns of annotated levels of GA-biosynthesis genes such as GPS, GA20ox and plant hormone metabolism and signal-transduction genes GA3ox, and the IAA-biosynthesis gene ToFZY, and down- were analyzed in depth. GA treatment led mainly to the regulation of the GA-inactivation gene GA2ox [16]. regulation of auxin, ethylene and ABA biosynthesis and Auxin-induced parthenocarpic tomato fruit set is partially signaling in fig. Differential expression of plant hormone mediated by GAs, as indicated by the dependence of par- genes revealed synchronized change trends with fruit thenocarpic fruit formation in the unfertilized entire to- development between female flowers and receptacles mato mutant with a defective SlIAA9 gene on regulation following GA treatment. Our work provides new Chai et al. BMC Plant Biology (2018) 18:100 Page 3 of 15 information on GA-induced fruit set in San Pedro-type fig by were assessed by Nanodrop 2000 (Thermo Scientific, unraveling key pathways and genes in the expression-regulation USA) and electrophoresis in a 1% agarose gel. RNA in- network, which could help elucidate the molecular mechanisms tegrity number, analyzed by Agilent 2100 Bioanalyzer, underlying the diversified parthenocarpic nature of fig species. was > 8.0. Methods Library preparation and paired-end transcriptome Plant materials and exogenous GA treatment sequencing A San Pedro-type fig cultivar (Ficus carica L. ‘Asteran’) The mRNA of each sample’s three biological replicates was used in this study. The trees were planted in a com- were respectively enriched using cellulose oligo (dT) mercial orchard in Beijing with regular management. magnetic beads (Invitrogen, USA), then fragmented into Main crop fruits at pre-anthesis (15–20 mm), anthesis ca. 200-bp fragments. The fragments were transcribed (20–25 mm) and post-anthesis (> 25 mm) stages were and double-stranded cDNA was synthesized, then end selected to evaluate the time window of GA treatment; repair, 3′-end single-nucleotide A (adenine) addition and 200 μl of 25 mg/l GA solution was injected through the ligation of adaptors were performed according to the ostiole into the fruit cavity with a 1-ml syringe. Solution manufacturer’s instructions. The resultant fragments without GA was injected into the control group. For were enriched by PCR and purified using 2% Certified each treatment and control group, 100 fruits were Low Range Ultra Agarose (Bio-Rad) to create the final injected and fruit-set rates were calculated. cDNA library. The quantitative assay was conducted Growth curves of control and GA -treated fruit were using Picogreen fluorescent dye (Invitrogen, USA) with a established with 30 tagged fruits using a digital slide cali- TBS-380 fluorimeter (Invitrogen, USA). After bridge per. Three biological replicates were collected from pools PCR amplification on Illumina cBot using Truseq PE of at least 30 fruits 2, 4 and 10 days after treatment Cluster Kit v3-cBot-HS, paired-end (2 × 151 bp) sequen- (DAT). For transcriptome sequencing and plant hormone cing of the cDNA library products was carried out using analysis, main crop fruits at anthesis, 22 ± 1 mm in trans- the Illumina HiSeq4000 platform. verse diameter, were used. Female flowers and receptacle were carefully separated, frozen in liquid nitrogen on site, Illumina read processing and functional annotation pulverized in the lab, and stored at − 80 °C for further use. Clean reads were generated by removing reads with adaptors and more than 10% unknown nucleotides, and Phytohormone content assays low-quality reads (the rate of reads with quality value The levels of IAA, GA ,GA , zeatin and ABA were moni- ≤10 was more than 20%) from the raw data. Then clean 3 4 tored by reverse-phase HPLC–MS/MS [26]. In brief, about reads were mapped to our reported reference sequences 50 mg of the frozen and pulverized plant material was deposited at DDBJ/EMBL/GenBank under the accession placed in 500 μl extraction solvent (isopropanol:H O:con- number GDKC00000000 [25] using Bowtiealigner centrated HCl, 2:1:0.002, v/v), then shaken (100 rpm) at (http://bowtie-bio.sourceforge.net/index.shtml)[28] with 4 °C for 30 min. Dichloromethane (1 ml) was added, no more than two mismatches allowed. The number of followed by shaking (100 rpm) at 4 °C for 30 min. After mapped clean reads for each transcript was normalized centrifugation (10,000 × g) at 4 °C for 5 min, the solution to FPKM (fragments per kilobase of exon model per mil- was divided into two phases, and the lower phase (ap- lion mapped reads) by RSEM (http://deweylab.biostat.- proximately 1 ml) was collected. The solvent mixture was wisc.edu/rsem/). Genes that were differentially expressed concentrated to dryness using a Termovap sample con- between GA -treated samples and their corresponding centrator (N-EVAP, Organomation, USA), then redis- controls were analyzed by edgeR (http://www.biocon- solved in 100 μl methanol. The extracted hormone ductor.org/packages/2.12/bioc/html/edgeR.html)[29], solutions were run in an AB SCIEX (USA) Triple Quad™ and “false discovery rate (FDR) < 0.01 and absolute value 5500 LC–MS/MS system. For each sample, three of log (FPKM /FPKM ) ≥ 1” were set as the 2 treatment control independent and parallel extractions were carried out. thresholds to determine significant differences in gene Presented values are means of all replicates ± SD. Signifi- expression. For GO annotation of all transcripts (http:// cance analysis was performed by SPSS 17.0. www.geneontology.org/), Goatools (https://github.com/ tanghaibao/Goatools) was used. Pathway-enrichment RNA extraction analysis was carried out based on the KEGG pathway data- Total RNA from female flowers and receptacles taken 2, base by KOBAS (http://kobas.cbi.pku.edu.cn/)[30]. Both 4 and 10 DAT and the corresponding controls was iso- GO and KEGG enrichment analyses were based on Fisher’s lated using the CTAB method [27] and treated with Exact Test [31] with multiple testing correction of FDR RNase-free DNase I (TaKaRa, China) according to the [32], and corrected P-value ≤0.05 was selected as the manufacturer’s instructions. RNA quality and quantity threshold for significance. Chai et al. BMC Plant Biology (2018) 18:100 Page 4 of 15 Quantitative real-time PCR (qRT-PCR) verification its level 2 DAT (Fig. 1c); thereafter, IAA decreased to Thirteen phytohormone-related genes were selected for very low levels in both control and treated receptacles. validation by qRT-PCR and the specific primers are Exogenous GA treatment dramatically decreased zeatin shown in Additional file 1: Table S1. Isolation of total peak intensity in the female flowers 4 DAT, whereas it RNA was as described previously and RNase-free DNase specifically enhanced the zeatin peak to 2.4-fold that in I (D2270A, TaKaRa) treatment was performed to digest control receptacles 2 DAT (Fig. 1c); this action was op- DNA in the sample. For the first-strand cDNA synthesis, posite that in female flowers. ABA content was higher in 1000 ng of total RNA was reverse-transcribed with a GA -treated female flowers and receptacles than in their synthesis kit (D6210A, TaKaRa). Ten-fold-diluted cDNA control counterparts 2 and 4 DAT, whereas its level was template was used for the qRT-PCR assay performed by repressed in both tissues 10 DAT, to a level that was less the SYBR Premix Ex Taq Kit (DRR420A, TaKaRa) on a than half of the control. Thus, GA caused a shift in the 7500 Real-Time PCR System (Applied Biosystems, USA) ABA peak to 6 days earlier than its natural peak in fe- according to the manufacturer’s protocol. Relative quan- male flowers and receptacle (Fig. 1c). −ΔΔCt titative method (2 ) was used to calculate the level of expression of the tested genes [33]. All of the reac- Transcriptome sequencing analysis and qRT-PCR tions were performed in three replicates using actin as validation the internal gene [34]. The transcriptomes of female flowers and receptacles from control and GA -treated fruit were sequenced 2, 4 Results and 10 DAT. After removing adaptors and low-quality Developmental characteristics and endogenous hormone sequences, each library generated 29.7 to 40.5 million changes upon GA treatment clean reads (Additional file 1: Table S2). Sequencing To optimize the GA treatment time window at stage I, quality compliance was indicated by an error ratio of fruits at pre-anthesis (15–20 mm in diameter), anthesis single bases of less than 0.0097%; Q20 and Q30 (the (20–25 mm) and post-anthesis (> 25 mm) stages were percentage of bases with base quality greater than 20 or 30 grouped and treated with GA .GA treatment at anthe- 3 3 among the total bases) were around 98.6 and 96.0%, re- sis gave a fruit-set rate of about 72% 42 DAT, much spectively. The number of G and C bases was about 47.0% higher than the pre- and post-anthesis stage treatments for each gene-expression library. Alignment of the clean (40 and 10%, respectively) (Fig. 1a). Anthesis-stage fruit reads to the reference sequence showed over 24.7 million with transverse diameter of 22 ± 1 mm were therefore total mapped reads, with the ratio varying from 82.11 to selected for further plant hormone and transcriptome 83.89% of the total clean reads (Additional file 1:Table S2). experiments. The growth curve of GA -induced par- Thirteen genes related to phytohormone metabolism thenocarpic fruit is shown in Fig. 1b; the typical or signal transduction were selected for validation. double-sigmoid curve pattern was in agreement with Positive correlations were found between the RNA se- previous reports [35]. However, fruit abscission was hap- quencing (RNA-Seq) and qRT-PCR data in all pairwise pened in control group at the beginning of stage II, comparison groups (Additional file 1: Figure S1 and S2), which was called “premature drop”. confirming the consistency, validity and representative- Exogenous GA treatment directly increased GA con- ness of our RNA-Seq data. centration in the female flowers, yielding about 26.8-, 38.8- and 28.7-fold its amount in controls 2, 4 and 10 DAT, respectively (Fig. 1c). Exogenous GA was also im- Differences in gene expression pattern between control mediately transported into the receptacle, with 15-fold and GA -treated fruit control GA levels in the treatment group 2 DAT. De- Comparative transcriptome analysis was performed to creasing trends of GA content were seen from 2 to 10 identify the differentially expressed genes (DEGs) in DAT in both control and GA -treated female flowers GA -treated samples compared to controls. Filtered by a 3 3 and receptacles, indicating the fig tissues’ strong capacity minimum 2-fold FPKM difference with FDR < 0.01, for to deplete this hormone (Fig. 1c). GA was present in fig female flower samples, pairwise comparisons revealed tissues at a low and stable level, and was insensitive to 685, 294 and 1811 DEGs at 2, 4 and 10 DAT, respect- GA treatment. The declining phase of native IAA in ively, with 400 (58.4%), 103 (35.0%) and 701 (38.7%) up- stage I female flowers was greatly accelerated by GA regulated and 255 (41.6%), 191 (65.0%) and 1110 (61.3%) treatment 2 and 4 DAT, but 10 DAT, IAA degradation downregulated in the GA -treated samples relative to ceased in the treated flowers and its concentration accu- controls; for receptacle samples, 908, 911 and 1672 mulated to almost 2-fold that in the controls. At this genes were differentially expressed at 2, 4 and 10 DAT, stage, the receptacle contained much less IAA than the respectively, with 436 (48.0%), 509 (55.9%) and 755 female flowers, and GA treatment only slightly changed (45.2%) upregulated and 472 (52.0%), 402 (44.1%) and 3 Chai et al. BMC Plant Biology (2018) 18:100 Page 5 of 15 Fig. 1 Physiological features of San Pedro-type fig main crop fruit after gibberellin (GA ) application. a Effect of application time window on fruit- set ratio. b Growth curves of control and GA -treated fruits (GA was applied at anthesis). c GA ,GA , indole acetic acid (IAA), zeatin and ABA 3 3 3 4 concentration in control and GA -treated female flowers and receptacle. Error bars indicate standard deviation. Significance analysis was performed by SPSS 17.0. *P ≤ 0.05, **P ≤ 0.01; FW, fresh weight 917 (54.8%) downregulated in the treated samples rela- the shared enriched terms in the two different tissues 2, 4 tive to controls (Additional file 1: Figure S3). and 10 DAT were: binding (GO:0005488) and catalytic ac- Significantly enriched terms in the GO and KEGG da- tivity (GO:0003824) terms in the molecular function cat- tabases were analyzed to identify the principle biological egory, cell part (GO:0044464) and cell (GO:0005623) functions of the DEGs. In female flowers, 38, 37 and 45 terms in the cellular component group, and metabolic significant GO terms of biological process, cellular com- process (GO:0008152), cellular process (GO:0009987) and ponent and molecular function were enriched 2, 4 and single-organism process (GO:0044699) terms in the bio- 10 DAT, respectively, and in the receptacle, 40, 39 and logical process group. 41 GO terms were enriched (Additional file 1: Table S3). DEGs identified 2, 4 and 10 DAT were assigned to 4, 1 In general, the enrichment patterns of up- and downregu- and 13 significantly different KEGG pathways in female lated genes were similar in female flowers and receptacles; flowers and 9, 4 and 12 pathways in the receptacle, Chai et al. BMC Plant Biology (2018) 18:100 Page 6 of 15 respectively (Additional file 1: Table S4). The largest With respect to GA-signal transduction, only one DEG number of significantly changed pathways was found 10 encoding DELLA protein was identified, which was more DAT in both tissues. Among the hormone-related path- highly expressed in GA -treated fruit than in controls; its ways, tryptophan metabolism (ko00380) was significantly upregulation was significant in GA -treated female flowers enriched in female flowers 10 DAT, and plant 2 DAT (Fig. 2d). These results suggest that exogenous hormone-signal transduction (ko04075) was significantly GA treatment mainly influences GA-biosynthesis and ca- enriched in receptacles 2 and 10 DAT. tabolism pathways in both female flowers and receptacle, but not the signal-transduction process. Expression pattern of genes involved in plant hormone Auxin metabolism and signal-transduction pathways In GA -treated fig, auxin-pathway genes constituted Plant hormone-related genes differentially expressed in the largest group of DEGs among growth- and female flowers and/or receptacles are summarized in development-promoting phytohormones (Additional Additional file 1: Tables S5 and S6. Briefly, most of the file 1: Table S5). The expression level of indole-3-ace- genes belonged to the ethylene-metabolism and response taldehyde dehydrogenase (IAld) remained stable in pathways, followed by auxin, ABA and GA; relatively female flowers and receptacle 2 and 4 DAT, whereas fewer DEGs were related to cytokinin. A FPKM value > 10 10 DAT, it was upregulated in the two control tissues, for at least one sample was set as the threshold to select to significantly higher levels than in their GA -treated genes for further analysis. The expression profiles of the counterparts (Fig. 3). Two auxin-responsive Gretchen rest of the hormone-related DEGs, with FPKM values of Hagen 3 (GH3) family genes revealed different expression all samples < 10, are shown in Additional file 1:Figure S4. patterns in the female flowers and similar patterns in the receptacles. In female flowers, comp25108_c0 showed Gibberellin continuously increasing expression during development in We identified 10 GA biosynthesis and catabolism genes control tissues, more than 2-fold higher than in the encoding GA 20-oxidase (GA20ox), GA 3-oxidase treated tissues 10 DAT; in contrast, the other GH3 (GA3ox) and GA 2-oxidase (GA2ox) (Additional file 1: (comp29694_c0) decreased from 2 to 10 DAT in female Table S5). The expression level of the only GA20ox identi- flowers, and showed no significant difference between fied in this study was dramatically higher in female flowers control and treated flowers. In receptacles, both GH3 than receptacles, and it was significantly upregulated in genes exhibited over 2-fold upregulation 2 DAT (Fig. 3). GA -treated female flowers 2 DAT compared to controls, Three IAA-amino acid hydrolase (IAH) genes demon- then downregulated until 10 DAT in both control and strated overall moderate changes in expression during de- treated samples (Fig. 2a). Two GA3ox genes showed dif- velopment—increasing from 2 to 4 DAT and then ferent expression patterns: comp22862_c0 was weakly decreasing in control female flowers and receptacle; how- expressed in all studied samples 2 and 4 DAT and showed ever, GA treatment significantly elevated IAH expression sharp upregulation in control female flowers and recep- 2 DAT, followed by a continuous decrease, resulting in tacle 10 DAT, to significantly higher levels compared to similar expression levels between control and treated tis- their GA -treated counterparts; comp12680_c0 showed sues 4 and 10 DAT. This implies that GA only transiently the highest expression level 2 DAT and then a decrease modulates IAH expression in female flowers and recepta- during development, and was generally downregulated in cles (Fig. 3). the two different tissues after GA treatment (Fig. 2b). With respect to the auxin-response pathway (Additional Among all GA2ox genes, comp16885_c0 was most file 1: Table S5), auxin-influx carrier (AUX1)demon- highly expressed, but its expression was repressed 3.3- strated higher expression in GA -treated fruit female and 41.1-fold in GA -treated female flowers and recep- flowers than in controls on all sampling days, and in the tacle compared to controls at 10 DAT, respectively. treated receptacle, upregulation was observed 2 DAT, Another GA2ox gene (comp26664_c0) also exhibited whereas downregulation was identified 10 DAT (Fig. 3). significantly decreased expression in GA -treated vs. Expression patterns of four small auxin-up RNA (SAUR) control samples at 10 DAT. However, comp32778_c0 ex- genes could be separated into two groups: DEGs pression was upregulated after treatment at all sampling comp31670_c0 and comp33019_c0 were significantly stages in both female flowers and receptacle, and re- repressed in GA -treated fruit female flowers and recep- markably higher expression was found in the receptacle tacle 10 DAT, whereas comp16559_c0 and comp36657_c0 vs. female flowers. In addition, comp18866_c0 only were generally upregulated in the treatment group on all showed a significant difference in expression between three sampling days relative to their control counterparts control and treated receptacles 2 DAT, with the latter (Fig. 3). The negative auxin-response regulator AUX/IAA being 7.5-fold lower than the former (Fig. 2c). revealed divergent expression patterns between female Chai et al. BMC Plant Biology (2018) 18:100 Page 7 of 15 Fig. 2 Expression patterns of GA-metabolism and signaling genes. a GA20ox. b GA3ox. c GA2ox. d DELLA. F, female flowers; R, receptacle. *Fold change of FPKM ≥2 and FDR ≤ 0.001 flowers and receptacle and between control and treatment expression in GA -treated fruit, whereas 10 DAT, it was groups. In general, AUX/IAA presented higher expression 14.7- and 8.6-fold higher in control vs. GA -treated female in GA -treated samples than in controls (Fig. 3). flowers and receptacle, respectively (Fig. 4). Note that in rice, the activity of cis-zeatin is comparable to that of Cytokinin trans-zeatin and can upregulate cytokinin-inducible genes, Differential expression of cytokinin metabolism and whereas overexpression of cis-ZOG delays rice leaf senes- signal-transduction genes was identified (Fig. 4,Additional cence [36]. file 1:Table S5). Isopentenyl transferase (IPT)expression In cytokinin-signal transduction, the cytokinin receptor was higher in GA -treated female flowers than controls 2 cytokinin response1(CRE1) was downregulated in both fe- DAT, but lower 4 and 10 DAT. On the other hand, in the male flowers and receptacle after GA treatment, espe- receptacle, IPT was downregulated 2 and 4 DAT, and cially 10 DAT (Fig. 4). The negative A-type Arabidopsis moderately upregulated 10 DAT. Compared to controls, response regulators (ARRs) remained stable with a low ex- cytokinin oxidase/dehydrogenase (CKX) was upregulated pression level in GA -treated fruit at all tested stages, in GA -treated samples, and its expression was remark- whereas a very significant increase was identified in con- ably higher in female flowers vs. receptacle. Cis-zeatin trols 10 DAT, presenting 16- and 31-fold upregulation in O-glucosyltransferase (CIS-ZOG) showed very low expres- female flowers and receptacle, respectively, compared to sion in all samples 2 and 4 DAT, and maintained this low the corresponding GA -treated tissues (Fig. 4). 3 Chai et al. BMC Plant Biology (2018) 18:100 Page 8 of 15 Fig. 3 Expression patterns of auxin-metabolism and signaling genes. Grids with 11 different colors represent FPKM values of 0–10, 10–20, 20–40, 40–60, 60–80, 80–100, 100–200, 200–300, 300–400, 400–500, and over 500, respectively. For each gene, the upper and lower rows refer to the results of control and GA treatment, respectively. 2, 4 and 10 = days after treatment. *Fold change of FPKM ≥2 and FDR ≤ 0.001 Fig. 4 Expression patterns of cytokinin-metabolism and signaling genes. F, female flowers; R, receptacle. *Fold change of FPKM ≥2 and FDR ≤ 0.001 Chai et al. BMC Plant Biology (2018) 18:100 Page 9 of 15 ABA they demonstrated higher expression after treatment than All identified ABA-synthesis, catabolism and response in controls. The other two SAM genes—comp30786_c0 and DEGs are shown in Additional file 1: Table S6. NECD comp7483_c0—increased during the control samples’ devel- (9-cisepoxycarotenoid dioxygenase) is the first commit- opment, but showed decreasing trends after GA treat- ted and rate-limiting enzyme in ABA biosynthesis [37]. ment, leading to 6.8- and 7.7-fold downregulation 10 DAT, Two NCED genes—comp17110_c0 and comp22655_c0— respectively, in the GA -treated receptacle (Fig. 6). One were enhanced in the both tissues 2 and 4 DAT with gene annotated as 1-aminocyclopropane-1-carboxylate GA (Fig. 5a). In controls, their expression was relatively (ACC) synthase (ACS) presented higher expression in stable during development in female flowers, and rose GA -treated samples than in controls. ACC oxidase (ACO) first and then fell at the 10 DAT in the receptacle; how- genes, except comp23635_c1, had low expression in all ever, decreasing trends were found in both female tested control and GA -treated samples 2 and 4 DAT, flowers and receptacle of GA -treated fruit. The other whereas dramatic upregulation was detected in control fruit two NCED—comp19377_c0 and comp26438_c0—gener- female flowers and receptacle 10 DAT, resulting in signifi- ally increased in the two tissues of control fruit during cantly higher expression in the two control vs. treated tis- development, especially from 4 to 10 DAT, whereas their sues (2.5- to 42-fold) (Fig. 6). levels changed moderately in GA -treated tissues; ultim- In the ethylene-response pathway, ETR transcripts in- ately, their expression was significantly lower in treat- creased during development, showing the highest ex- ment vs. control tissues 10 DAT (Fig. 5a). ABA2, pression levels 10 DAT in controls. For GA -treated encoding another ABA-biosynthesis enzyme, was down- samples, their expression levels fell first and then rose, regulated in GA -treated vs. control tissues, but with no with a peak at 2 or 10 DAT (Fig. 6). The expression significant difference in the female flowers (Fig. 5b). The levels of ETR and EBF1/2 were significantly upregulated process catalyzed by ABA 8′-hydroxylase (ABA 8′-h) is in GA -treated female flowers and receptacle 2 DAT. In considered to be the main pathway for ABA catabolism general, ERF transcripts revealed an increasing trend in [37]. ABA 8′-h expression remained relatively stable in control fruit female flowers and receptacle, whereas they the controls, whereas remarkable upregulation was presented divergent expression patterns after GA treat- found in the GA -treated samples 10 DAT—4.9- and ment: comp16781_c0 and comp16998_c0 had the highest 3.3-fold higher in female flowers and receptacle, respect- expression 2 DAT in female flowers and receptacle; ively, than in controls (Fig. 5c). comp28870_c0 and comp17108_c0 displayed upward In the ABA-signaling pathway, PYR/PYL expression was trends during development similar to their expression in slightly lower in GA -treated samples than in controls 2 the controls (Fig. 6). ERFs were generally upregulated in and 4 DAT, but significantly higher 10 DAT (Fig. 5d). the two treatment tissues 2 DAT, whereas 10 DAT, their Protein phosphatase 2C (PP2C) usually represses ABA re- expression levels were slightly lower in GA -treated vs. sponses. PP2C expression was not significantly different control receptacles. between the control and treated fruit 2 and 4 DAT, whereas it was dramatically downregulated in GA -treated Discussion female flowers and receptacle 10 DAT (Fig. 5e). ABRE Our aim was to investigate the role of GA in partheno- binding factor (ABF) functions as a positive regulator of carpy induction in fig by comparing control and ABA signaling. The ABF genes’ expression was consistent GA -treated fruit’s female flowers and receptacles. We with an increasing tendency from 2 to 10 DAT in both focused our study on 2, 4 and 10 days after exogenous control and GA -treated samples, and their expression GA application in a San Pedro-type fig cultivar’s main 3 3 was more or less downregulated in all studied samples crop, which is genetically non-parthenocarpic. By Illu- after GA treatment (Fig. 5f). mina RNA-Seq transcriptome analysis, 1.3- and 3.1-fold more DEGs were identified in the receptacle vs. female Ethylene flowers at 2 and 4 DAT, respectively, whereas the num- A large number of ethylene biosynthesis- and signaling-related ber of DEGs was a little higher in female flowers than in DEGs were found in female flowers and/or recepta- the receptacle 10 DAT. This implied that aside from the cles after GA treatment (Additional file 1:Table S6). reproductive female flowers, which mainly influence S-adenosylmethionine synthetase (SAM) displayed the fruit development [38], the vegetative receptacle tissue highest expression 2 DAT in control and GA -treated might also have some regulatory effect on partheno- female flowers, followed by a gradual decrease, with no sig- carpic fruit set and growth. GA, auxin, cytokinin, ABA nificant differences between control and treatment (Fig. 6). and ethylene metabolism- and response-related genes Three SAM genes—comp28816_c0, comp11492_c0 and showed differential expression between control and comp28987_c0—peaked 4 DAT and showed their lowest GA -treated fruit, and phytohormone content assay re- expression 10 DAT in control and GA -treated receptacles; vealed a new balance of endogenous phytohormone 3 Chai et al. BMC Plant Biology (2018) 18:100 Page 10 of 15 Fig. 5 Expression patterns of ABA-metabolism and signaling genes. a NCED. b ABA2. c ABA 8′-h. d PYR/PYL. e PP2C. f ABF. F, female flowers; R, receptacle. *Fold change of FPKM ≥2 and FDR ≤ 0.001 levels in both female flowers and receptacle following Exogenous GA application reset the transcriptional exogenous GA treatment (Fig. 1c). This indicated that expression of plant hormone-related genes. In female parthenocarpy induction by GA in fig is the result of the flowers, 4, 1 and 15 phytohormone-metabolism genes coordinated action of different phytohormones. and 8, 2 and 12 response-related genes were significantly Chai et al. BMC Plant Biology (2018) 18:100 Page 11 of 15 Fig. 6 Expression patterns of ethylene-biosynthesis and signaling genes. Grids with 11 different colors represent FPKM values 0–10, 10–20, 20–40, 40–60, 60–80, 80–100, 100–200, 200–300, 300–400, 400–500 and over 500, respectively. For each gene, upper and lower rows refer to the results of control and GA treatment, respectively. 2, 4 and 10 = days after treatment. *Fold change of FPKM ≥2 and FDR ≤ 0.001 differently expressed 2, 4 and 10 DAT, respectively. In the tissue and developmental stage [39, 41]. Exogenous receptacle, 19, 4 and 18 phytohormone-metabolism and GA application before or after bloom in grapevine 10, 2 and 12 signal-transduction DEGs were identified 2, 4 leads to downregulation of the GA-biosynthetic genes and 10 DAT, respectively. The higher representation of GA20ox and GA3ox and upregulation of the GA-catabolic plant hormone transcripts’ differential expression in re- gene GA2ox [15, 42, 43]; this has also been observed in ceptacles compared to female flowers may further suggest Arabidopsis [39] and tobacco [44]. Conversely, in the present an indelible role for receptacle tissue in controlling fruit study, GA20ox was upregulated in female flowers 2 DAT and set and development. Concrete spatiotemporal transcrip- GA2ox was repressed in both female flowers and receptacle tional changes in phytohormone-related genes induced by 10 DAT. This difference could be explained by the change in GA are summarized in Fig. 7. There were more common endogenous GA content: in both fig and grapevine, GA DEGs with the same changing trends in female flowers content increases significantly and immediately after treat- and receptacles 10 DAT than 2 and 4 DAT, indicating that ment and then decreases; however, for grapevine, GA con- GA treatment induces synchronized patterns of changing tent drops to levels equal to or significantly less than the gene expression between the two tissues as fruit develop- control 3 DAT [43, 45], whereas in fig, it is still significantly ment progresses. In addition, in both female flowers and higher than the control 10 DAT (Fig. 1c). receptacles 10 DAT, the number of DEGs related to ethyl- Auxin is one of the main regulators of fruit set and devel- ene metabolism was highest, followed by those related to opment. Similar to [42], auxin-biosynthesis genes, e.g. GA and ABA metabolism; for phytohormone response, YUCCA, did not show significant regulation after GA the highest number of DEGs were related to auxin. treatment in fig. The main differentially expressed auxin-metabolism genes identified in the present study Phytohormone-metabolism response to exogenous GA were IAH and GH3. The former encodes enzymes that pro- Previous studies have shown that transcription of duce free IAA from IAA–amino acid conjugates, whereas GA-metabolism genes is regulated by the content of the latter catalyzes auxin conjugation [46]. IAH and GH3 bioactive GA [2, 39]. GA oxidases modulate the GA-biosynthetic were mainly significantly upregulated 2 DAT; upregulation pathway mainly through feedback-loop mechanisms of GH3 has also been reported previously [42], indicating [40] and expression regulation that depends on the that bioactive auxin levels are co-mediated by IAH and Chai et al. BMC Plant Biology (2018) 18:100 Page 12 of 15 Fig. 7 Summary of main differentially expressed phytohormone-related genes identified in female flowers and receptacle after GA treatment. Different colored circles represent different phytohormones and circle size indicates relative number of DEGs—the higher the number, the bigger the circle. Genes in red (green) color are upregulated (downregulated). Gray highlight represents genes with same changing trend in both female flowers and receptacle following GA application GH3 in fig. Cytokinin induces parthenocarpy by altering expression of the ethylene-biosynthesis genes ACOs de- the expression of GA-biosynthesis genes [9, 16]. Onlyafew creases during parthenocarpic fruit set and development metabolic genes’ transcriptional levels were regulated by or after GA treatment [13]. GA treatment in our dataset. For female flowers, cytokinin-metabolism transcripts were only mediated by GA treatment 10 DAT, whereas for the receptacle, exogen- Phytohormone signaling in response to exogenous GA 3 3 ous GA also regulated cytokinin metabolism 2 and 4 DAT DELLA protein has been characterized as a negative regu- (Fig. 7). Zeatin concentration increased dramatically in the lator in GA signaling [48]. One DELLA gene exhibited up- receptacle 2 DAT, then decreased to a markedly lower level regulation after GA application, most markedly in female than controls, suggesting that cytokinin-biosynthesis genes flowers 2 DAT, in accordance with previously reported re- respond more rapidly (in under 2 days) in the receptacle sults [7, 13, 43]. This indicates GA-perception feedback following GA application. regulation in female flowers and receptacle. Aside from growth-promoting hormones, GA treat- Studies of the crosstalk between GA and auxin have ment also modulated the biosynthesis of ABA and ethyl- revealed that GA-induced parthenocarpy arises from the ene, which mainly play GA-antagonistic roles in the interaction of GA with auxin signaling [42], and that control of many plant developmental processes [47]. auxin-induced fruit set is mediated in part by GA [9]. In Enhancement of ABA and ethylene biosynthesis 2 DAT the present study, AUX/IAA gene expression was signifi- might reflect ‘feedback’ regulation of the highly signifi- cantly upregulated in the receptacle on all sampling days cantly increased GA content in fig fruit following and in female flowers 10 DAT, suggesting feedback regu- exogenous GA application. This ‘feedback’ response lation of the auxin response following GA treatment, in 3 3 then declines, with almost no ABA- or ethylene-related agreement with results in grape [15]. In addition to the DEGs being identified 4 DAT. Along with fruit develop- AUX/IAAs, differential expression of SAUR transcript ment, both female flowers and receptacle revealed sup- levels was seen in both female flowers and receptacles pression of ABA and ethylene biosynthesis 10 DAT, with 10 DAT (Fig. 3). Further study is needed to elucidate downregulation of NCED and ACO and upregulation of their roles in fig fruit development. DEGs related to ABA 8′-h, in accordance with the lower ABA level in cytokinin signaling were only identified 10 DAT in the GA -treated fruits. Previous reports have shown that female flowers and receptacle (Fig. 7). 3 Chai et al. BMC Plant Biology (2018) 18:100 Page 13 of 15 GA treatment also changed the signaling of ABA and control and GA -treated female flowers and receptacles. DAT, days after ethylene. Enhancement of ethylene signaling was identi- treatment, Figure S4. Heat maps of hormone-related genes with low expression. FPKM of all samples was < 10. (PDF 1241 kb) fied 2 DAT. Crosstalk internodes between GA and the ethylene-signaling pathway were elucidated by the role Abbreviations of group VII ERFs as DELLA partners [49], and ERF11 ABA 8′-h: ABA 8′-hydroxylase; ABA: Abscisic acid; ABF: ABRE binding factor; activates GA biosynthesis and signaling [50]. Genes en- ACC: 1-aminocyclopropane-1-carboxylate; ACS: ACC synthase; ACO: ACC coding ABA receptor PYR/PYL and PP2C, a negative oxidase; ARRs: Arabidopsis response regulators; AUX1: auxin-influx carrier; CIS- ZOG: cis-zeatin O-glucosyltransferase; CKX: Cytokinin oxidase/dehydrogenase; regulator in the ABA-response pathway, displayed up- CRE1: Cytokinin response1; DAT: Days after treatment; DEGs: Differentially regulation and downregulation, respectively, in both tis- expressed genes; FDR: False discovery rate; FPKM: Fragments per kilobase of sues 10 DAT, leading to enhancement of ABA-signal exon model per million mapped reads; GA20ox: GA 20-oxidase; GA2ox: GA 2-oxidase; GA3ox: GA 3-oxidase; GAs: Gibberellins; GH3: Gretchen Hagen 3; transduction in both female flowers and receptacle. A GO: Gene Ontology; IAA: Indole-3-acetic acid; IAH: IAA-amino acid hydrolase; recent study on rice revealed the central role of PYR/ IAld: Indole-3-acetaldehyde dehydrogenase; IPT: Isopentenyl transferase; PYL in the antagonistic action of GA and ABA [51]. KEGG: Kyoto Encyclopedia of Genes and Genomes; NECD: 9- cisepoxycarotenoid dioxygenase; PP2C: Protein phosphatase 2C; qRT- PCR: Quantitative real-time PCR; RNA-Seq: RNA sequencing; SAM: S- Conclusions adenosylmethionine synthetase; SAUR: Small auxin-up RNA GA treatment, performed to induce parthenocarpy in San Funding Pedro-type fig main crop, resulted in a highly significant This work was financially supported by the grants from National Natural GA content increment in both female flowers and recep- Science Foundation of China project NSFC [31372007]. tacle tissues of fig fruit. Changes in the expression of key Availability of data and materials genes in plant hormone-synthesis and metabolism Raw data of all sample-sequencing results have been submitted to the NCBI pathways—as reflected by modulated endogenous plant Sequence Read Archive (SRA, http://www.ncbi.nlm.nih.gov/Traces/sra) with accession number SRP113799. hormone levels—and in the hormone-signaling pathway, converted the pre-dropping fruit pattern to an artificial par- Authors’ contributions thenocarpic developmental pattern. Spatiotemporal charac- HM and SC designed the experiments. LC and PC conducted the experiments and analyzed the results. LC and HM prepared the manuscript. teristics of the changes in plant hormone levels and SC and MF assisted in drafting the manuscript. All authors read and gene-expression patterns in GA -induced fig parthenocarpy approved the final manuscript. were reflected by differentially reacting female flowers and receptacle tissue, and temporal waves in major plant hor- Ethics approval and consent to participate The experiments did not involve endangered or protected species. No mone and relevant genes’ expression were illustrated. specific permits were required for these activities because the figs used in Further study on fig fruit development with the application this study were obtained from an orchard in Beijing, which was a of other exogenous plant hormones, e.g., auxin and cytoki- demonstration base of China Agricultural University. nin, will serve to further understand the mechanisms of Competing interests GA-induced parthenocarpy and the key plant hormones in The authors declare that they have no competing interests. fig fruit set and development. Author details College of Horticulture, China Agricultural University, Beijing, People’s Additional file Republic of China. College of Beijing Laboratory for Food Quality and Safety, College of Food Science and Nutritional Engineering, China Agricultural University, Beijing, People’s Republic of China. Department of Fruit Tree Additional file 1: Table S1. Primer sequences of genes used for Sciences, Agricultural Research Organization, Volcani Center, Bet-Dagan, validation of RNA-Seq results by quantitative real-time PCR, Table S2. Israel. Summary of sequencing results for control and GA -treated female flowers and receptacles, Table S3. Significant GO terms (corrected Received: 9 September 2017 Accepted: 24 May 2018 P-value ≤0.05) identified between control and GA -treated female flowers and receptacles, Table S4. Significant KEGG pathways (corrected P-value ≤0.05) identified between control and GA -treated female flowers and References receptacles, Table S5. All gibberellin-, auxin- and cytokinin-synthesis, 1. Blazquez MR, Nilsson O, Sussman MR, Weigel D. Gibberellins promote catabolism and response transcripts identified in this study which were flowering of Arabidopsis by activating the LEAFY promoter. Plant Cell. differentially expressed (FDR < 0.01 and the absolute value of log 1998;10(5):791. (FPKM /FPKM ) ≥ 1) in at least one pairwise comparison group, treatment control 2. Serrani JC, Sanjuan R, Ruizrivero O, Fos M, Garciamartinez JL. Gibberellin Table S6. All abscisic acid- and ethylene-synthesis, catabolism and re- regulation of fruit set and growth in tomato. Plant Physiol. 2007;145(1): sponse transcripts identified in this study which were differentially 246–57. expressed (FDR < 0.01 and the absolute value of log (FPKM / 2 treatment 3. Ayele BT, Ozga JA, Reinecke DM. Regulation of GA biosynthesis genes FPKM ) ≥ 1) in at least one pairwise comparison group, Figure S1. control during germination and young seedling growth of pea (Pisum sativum Verification of RNA-Seq results by qRT-PCR. Bars represent standard L.). J Plant Growth Regul. 2006;25(3):219–32. deviation. F, female flowers; R, receptacle, Figure S2. Correlation of fold 4. Tyler L, Thomas SG, Hu J, Dill A, Alonso JM, Ecker JR, Sun T. DELLA proteins changes in gene expression between RNA-Seq and qRT-PCR. Equation of 2 and gibberellin-regulated seed germination and floral development in linear regression and correlation coefficient (R ) are shown. DAT, days Arabidopsis. Plant Physiol. 2004;135(2):1008–19. after treatment, Figure S3. Number of differentially expressed genes 5. Hedden P, Phillips AL. Gibberellin metabolism: new insights revealed by the (FDR ≤ 0.001 and log (FPKM /FPKM ) ≥ 1or≤− 1) between 2 treatment control genes. Trends Plant Sci. 2000;5(12):523–30. Chai et al. BMC Plant Biology (2018) 18:100 Page 14 of 15 6. Yarushnykov VV, Blanke MM. Alleviation of frost damage to pear flowers by 28. Langmead B, Salzberg SL. Fast gapped-read alignment with bowtie 2. Nat application of gibberellin. Plant Growth Regul. 2005;45(1):21–7. Methods. 2012;9(4):357. 7. Serrani JC, Ruizrivero O, Fos M, Garciamartinez JL. Auxin-induced fruit-set in 29. Robinson MD, Mccarthy DJ, Smyth GK. edgeR: a Bioconductor package for tomato is mediated in part by gibberellins. Plant J. 2008;56(6):922–34. differential expression analysis of digital gene expression data. Bioinformatics. 2010;26(1):139. 8. Mesejo C, Yuste R, Reig C, Martínez-Fuentes A, Iglesias DJ, Muñoz-Fambuena N, Bermejo A, Germanà MA, Primo-Millo E, Agustí M. Gibberellin reactivates 30. Xie C, Mao X, Huang J, Ding Y, Wu J, Dong S, Kong L, Gao G, Li C, Wei L. and maintains ovary-wall cell division causing fruit set in parthenocarpic KOBAS 2.0: a web server for annotation and identification of enriched Citrus species. Plant Sci. 2016;247:13–24. pathways and diseases. Nucleic Acids Res. 2011;39 9. Lu L, Liang J, Zhu X, Xiao K, Li T, Hu J. Auxin- and Cytokinin-induced berries 31. Upton G. Fisher's exact test. J R Stat Soc A Sta. 1992;155:395–402. set in grapevine partly rely on enhanced gibberellin biosynthesis. Tree 32. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a Genet Genomes. 2016;12(3):41. practical and powerful approach to multiple testing. J R Stat Soc B. 10. Shinozaki Y, Hao S, Kojima M, Sakakibara H, Ozekiiida Y, Zheng Y, Fei Z, 1995;57(1):289–300. Zhong S, Giovannoni JJ, Rose JKC. Ethylene suppresses tomato (Solanum 33. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using lycopersicum) fruit set through modification of gibberellin metabolism. real-time quantitative PCR and the 2− ΔΔCT method. Methods. 2001; Plant J. 2015;83(2):237–51. 25(4):402–8. 11. Nitsch L, Kohlen W, Oplaat C, Charnikhova T, Cristescu S, Michieli P, Wolters- 34. Freiman ZE, Rosianskey Y, Dasmohapatra R, Kamara I, Flaishman MA. The Arts M, Bouwmeester H, Mariani C, Vriezen WH. ABA-deficiency results in ambiguous ripening nature of the fig (Ficus carica L.) fruit: a gene- reduced plant and fruit size in tomato. J Plant Physiol. 2012;169(9):878–83. expression study of potential ripening regulators and ethylene-related 12. Mariotti L, Picciarelli P, Lombardi L, Ceccarelli N. Fruit-set and early fruit genes. J Exp Bot. 2015;66(11):3309. growth in tomato are associated with increases in Indoleacetic acid, 35. Marei N, Crane JC. Growth and respiratory response of fig (Ficus carica L. cv. Cytokinin, and bioactive gibberellin contents. J Plant Growth Regul. Mission) fruits to ethylene. Plant Physiol. 1971;48(3):249–54. 2011;30(4):405. 36. Kudo T, Makita N, Kojima M, Tokunaga H, Sakakibara H. Cytokinin activity of 13. Tang N, Deng W, Hu G, Hu N, Li Z. Transcriptome profiling reveals the cis-zeatin and phenotypic alterations induced by overexpression of putative regulatory mechanism underlying pollination dependent and cis-Zeatin-O-glucosyltransferase in rice. Plant Physiol. 2012;160(1):319. parthenocarpic fruit set mainly mediated by auxin and gibberellin. PLoS 37. Nambara E, Marionpoll A. Abscisic acid biosynthesis and catabolism. Annu One. 2015;10(4) Rev Plant Biol. 2005;56(1):165–85. 14. Ji L, Zhe W, Li C, Tinglin Z, Qinwei G, Jian X, Li J, Qunfeng L, Sanwen H, 38. RosianskiY,Doronfaigenboim A,Freiman ZE,LamaK,Milocochavi S, Zhengguo L. Transcriptome comparison of global distinctive features Dahan Y, Kerem Z, Flaishman MA. Tissue-specific transcriptome and between pollination and parthenocarpic fruit set reveals transcriptional hormonal regulation of pollinated and Parthenocarpic fig (Ficus carica phytohormone cross-talk in cucumber (Cucumis sativus L.). Plant Cell L.) fruit suggest that fruit ripening is coordinated by the reproductive Physiol. 2014;55(7):1325. part of the Syconium. Front Plant Sci. 2016;7 15. Chai L, Li Y, Chen S, Perl A, Zhao F, Ma H. RNA sequencing reveals high 39. Rieu I, Ruizrivero O, Fernandezgarcia N, Griffiths J, Powers SJ, Gong F, resolution expression change of major plant hormone pathway genes Linhartova T, Eriksson S, Nilsson O, Thomas SG. The gibberellin biosynthetic after young seedless grape berries treated with gibberellin. Plant Sci. genes AtGA20ox1 and AtGA20ox2 act, partially redundantly, to promote 2014;229:215–24. growth and development throughout the Arabidopsis life cycle. Plant J. 16. Ding J, Chen B, Xia X, Mao W, Shi K, Zhou Y, Yu J. Cytokinin-induced 2008;53(3):488. parthenocarpic fruit development in tomato is partly dependent on 40. Middleton AM, Ubedatomas S, Griffiths J, Holman T, Hedden P, Thomas SG, enhanced gibberellin and auxin biosynthesis. PLoS One. 2013;8(7):e70080. Phillips AL, Holdsworth MJ, Bennett MJ, King JR. Mathematical modeling 17. Mignolli F, Vidoz ML, Mariotti L, Lombardi L, Picciarelli P. Induction of elucidates the role of transcriptional feedback in gibberellin signaling. Proc gibberellin 20-oxidases and repression of gibberellin 2β-oxidases in Natl Acad Sci U S A. 2012;109(19):7571–6. unfertilized ovaries of entire tomato mutant, leads to accumulation of 41. Yamaguchi S, Kamiya Y, Sun T. Distinct cell-specific expression patterns of active gibberellins and parthenocarpic fruit formation. Plant Growth early and late gibberellin biosynthetic genes during Arabidopsis seed Regul. 2015;75(2):415–25. germination. Plant J. 2001;28(4):443–53. 18. Wang H, Jones B, Li Z, Frasse P, Delalande C, Regad F, Chaabouni S, 42. Jung CJ, Hur YY, Jung S-M, Noh J-H, Do G-R, Park S-J, Nam J-C, Park K-S, Latche A, Pech J, Bouzayen M. The tomato aux/IAA transcription factor Hwang H-S, Choi D, et al. Transcriptional changes of gibberellin oxidase IAA9 is involved in fruit development and leaf morphogenesis. Plant genes in grapevines with or without gibberellin application during Cell. 2005;17(10):2676–92. inflorescence development. J Plant Res. 2014;127(2):359–71. 19. Tiwari A, Offringa R, Heuvelink E. Auxin-induced fruit set in Capsicum 43. Cheng C, Jiao C, Singer SD, Gao M, Xu X, Zhou Y, Li Z, Fei Z, Wang Y, Wang annuum L. requires downstream gibberellin biosynthesis. J Plant Growth X. Gibberellin-induced changes in the transcriptome of grapevine ( Vitis Regul. 2012;31(4):570–8. labrusca × V. Vinifera ) cv. Kyoho flowers. BMC Genomics. 2015;16(1):128. 20. Flaishman MA, Rodov V, Stover E. The fig: botany, horticulture, and 44. Gallegogiraldo L, Ubedatomás S, Gisbert C, Garcíamartínez JL, Moritz T, breeding. Horticultural Rev-Westport Then N Y. 2008;34:113. Lópezdíaz I. Gibberellin homeostasis in tobacco is regulated by gibberellin 21. Beck NG, Lord EM. Breeding system in Ficus carica, the common fig. II. metabolism genes with different gibberellin sensitivity. Plant Cell Physiol. Pollination events. Am J Bot. 1988;75(12):1913–22. 2008;49(5):679. 22. Crane J, Campbell R. Breaking rest and inducing parthenocarpy in the 45. Wang ZR, Zhao F, Zhao XB, Ge H, Chai L, Chen S, Perl A, Ma H. Proteomic Calimyrna fig with gibberellin: Proc 15th Int Cong Hort, Nice; 1958. analysis of berry-sizing effect of GA3 on seedless Vitis vinifera L. Proteomics. 23. Crane JC, Van Overbeek J. Kinin-induced parthenocarpy in the fig, Ficus 2012;12(1):86–94. carica L. Science. 1965;147(3664):1468–9. 46. Rampey RA, LeClere S, Kowalczyk M, Ljung K, Sandberg G, Bartel B. A family 24. Lodhi F, Bradley MV, Crane JC. Auxins and gibberellin-like substances in of auxin-conjugate hydrolases that contributes to free indole-3-acetic acid parthenocarpic and non-parthenocarpic syconia of Ficus carica L., cv. King. levels during Arabidopsis germination. Plant Physiol. 2004;135(2):978–88. Plant Physiol. 1969;44(4):555–61. 47. Weiss D, Ori N. Mechanisms of cross talk between gibberellin and other 25. Chai L, Wang Z, Chai P, Chen S, Ma H. Transcriptome analysis of San Pedro- hormones. Plant Physiol. 2007;144(3):1240–6. type fig (Ficus carica L.) parthenocarpic breba and non-parthenocarpic main 48. Sun T. Gibberellin-GID1-DELLA: a pivotal regulatory module for plant crop reveals divergent phytohormone-related gene expression. Tree Genet growth and development. Plant Physiol. 2010;154(2):567–70. Genomes. 2017;13(4):83. 49. Marínde lRN, Sotillo B, Miskolczi P, Gibbs DJ, Vicente J, Carbonero P, 26. Pan X, Welti R, Wang X. Quantitative analysis of major plant hormones in Oñatesánchez L, Holdsworth MJ, Bhalerao R, Alabadí D. Large-scale crude plant extracts by high-performance liquid chromatography-mass identification of gibberellin-related transcription factors defines group VII spectrometry. Nat Protoc. 2010;5(6):986. ETHYLENE RESPONSE FACTORS as functional DELLA partners. Plant Physiol. 27. Reid KE, Olsson N, Schlosser J, Peng FY, Lund ST. An optimized grapevine 2014;166(2):1022–32. RNA isolation procedure and statistical determination of reference genes for 50. Zhou X, Zhang Z, Park J, Tyler L, Yusuke J, Qiu KQ, Nam E, Lumba S, real-time RT-PCR during berry development. BMC Plant Biol. 2006;6(1):27. Desveaux D, Mccourt P. The ERF11 transcription factor promotes internode Chai et al. BMC Plant Biology (2018) 18:100 Page 15 of 15 elongation by activating gibberellin biosynthesis and signaling. Plant Physiol. 2016;171(4):2760–70. 51. Lin Q, Wu F, Sheng P, Zhang Z, Zhang X, Guo X, Wang J, Cheng Z, Wang J, Wang H. The SnRK2-APC/CTE regulatory module mediates the antagonistic action of gibberellic acid and abscisic acid pathways. Nat Commun. 2015;6:7981.

Journal

BMC Plant BiologySpringer Journals

Published: Jun 1, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off