Silencing of the Ortholog of DEFECTIVE IN ANTHER DEHISCENCE 1 Gene in the Woody Perennial Jatropha curcas Alters Flower and Fruit Development

DEFECTIVE IN ANTHER DEHISCENCE 1 (DAD1), a phospholipase A1, utilizes galactolipids (18:3) to generate α-linolenic acid (ALA) in the initial step of jasmonic acid (JA) biosynthesis in Arabidopsis thaliana. In this study, we isolated the JcDAD1 gene, an ortholog of Arabidopsis DAD1 in Jatropha curcas, and found that it is mainly expressed in the stems, roots, and male flowers of Jatropha. JcDAD1-RNAi transgenic plants with low endogenous jasmonate levels in inflorescences exhibited more and larger flowers, as well as a few abortive female flowers, although anther and pollen development were normal. In addition, fruit number was increased and the seed size, weight, and oil contents were reduced in the transgenic Jatropha plants. These results indicate that JcDAD1 regulates the development of flowers and fruits through the JA biosynthesis pathway, but does not alter androecium development in Jatropha. These findings strengthen our understanding of the roles of JA and DAD1 in the regulation of floral development in woody perennial plants.


Characterization of the JcDAD1 Gene
We obtained a JcDAD1 cDNA library (GenBank accession no. 105643375) from the Jatropha transcriptome [44]. The lengths of the JcDAD1 genomic sequence and coding sequence (CDS) are 1537 bp and 1323 bp, respectively (http://jcdb.xtbg.ac.cn/). The JcDAD1 gene, which encodes 440 amino acids, is located on the fifth chromosome of Jatropha [45]. Similar to Arabidopsis and rice, JcDAD1 possesses only one exon accompanied by no introns (Figure 1A), indicating that DAD1 is evolutionarily conserved among these plants. Phylogenetic analysis showed that JcDAD1 has a close relationship with HbDAD1 and RcDAD1 ( Figure 1B). In addition, the JcDAD1 protein containing a phospholipase A1 domain belongs to a member of the alpha/beta-hydrolase superfamily of proteins, which is similar to that of Arabidopsis [46].
The tissue-specific expression analysis showed that JcDAD1 is primarily expressed in the stems, roots, and male flowers but has low expression levels in other tissues of Jatropha ( Figure 2). In Arabidopsis, expression of DAD1 is highly restricted to the filaments of stamens, which is consistent with the function of JA in promoting water transport by synchronizing anther dehiscence, pollen maturation, and flower opening [22]. In rice, expression of the EG1 gene (an ortholog of DAD1) is high in inflorescence primordia, but weak in developing floral primordia, which is in accordance with its role in early flower development [28,47]. However, the transcript of tomato LeLID1 (a homolog of Arabidopsis DAD1) is hardly detected in reproductive organs (buds, flowers, or fruits) but strongly expressed in germinating seedlings, where the encoded protein functions as a TAG lipase [48]. These results show that DAD1 genes have various expression patterns in different species, indicating that they may have different functions. The tissue-specific expression analysis showed that JcDAD1 is primarily expressed in the stems, roots, and male flowers but has low expression levels in other tissues of Jatropha ( Figure 2). In Arabidopsis, expression of DAD1 is highly restricted to the filaments of stamens, which is consistent with the function of JA in promoting water transport by synchronizing anther dehiscence, pollen maturation, and flower opening [22]. In rice, expression of the EG1 gene (an ortholog of DAD1) is high in inflorescence primordia, but weak in developing floral primordia, which is in accordance with its role in early flower development [28,47]. However, the transcript of tomato LeLID1 (a homolog of Arabidopsis DAD1) is hardly detected in reproductive organs (buds, flowers, or fruits) but strongly expressed in germinating seedlings, where the encoded protein functions as a TAG lipase [48]. These results show that DAD1 genes have various expression patterns in different species, indicating that they may have different functions.    Supplementary Table S1. The tissue-specific expression analysis showed that JcDAD1 is primarily expressed in the stems, roots, and male flowers but has low expression levels in other tissues of Jatropha ( Figure 2). In Arabidopsis, expression of DAD1 is highly restricted to the filaments of stamens, which is consistent with the function of JA in promoting water transport by synchronizing anther dehiscence, pollen maturation, and flower opening [22]. In rice, expression of the EG1 gene (an ortholog of DAD1) is high in inflorescence primordia, but weak in developing floral primordia, which is in accordance with its role in early flower development [28,47]. However, the transcript of tomato LeLID1 (a homolog of Arabidopsis DAD1) is hardly detected in reproductive organs (buds, flowers, or fruits) but strongly expressed in germinating seedlings, where the encoded protein functions as a TAG lipase [48]. These results show that DAD1 genes have various expression patterns in different species, indicating that they may have different functions.

JcDAD1 Gene Silencing Increased Inflorescence Branching, Flower Number, and Flower Size
To investigate the functionality of the JcDAD1 gene, we transformed the JcDAD1-RNAi construct into Jatropha plants and identified 16 independent JcDAD1-RNAi transgenic lines (T1 generation), in which the transcript of JcDAD1 is repressed in inflorescence buds ( Figure 3). The order of inflorescence branches was increased to the fifth-order branches in the transgenic inflorescence, whereas the wild type (WT) inflorescence had only fourth-order branches ( Figure 4). Compared with the WT plants, notable increases in the female flower number (approximately three-fold) and male flower number (three-to five-fold) per inflorescence were observed in the transgenic Jatropha plants ( Figure 5A−F), which is similar to the phenotype of the Arabidopsis dad1 mutant [22]. The flower size of the transgenic plants was conspicuously larger than that of the WT plants ( Figure 6A−L), and the average diameters of the transgenic female and male flowers increased by 2−4 mm ( Figure 6M). These results show that JcDAD1 participates in flower development and promotes flower production and floral organ growth in Jatropha. To investigate the functionality of the JcDAD1 gene, we transformed the JcDAD1-RNAi construct into Jatropha plants and identified 16 independent JcDAD1-RNAi transgenic lines (T1 generation), in which the transcript of JcDAD1 is repressed in inflorescence buds ( Figure 3). The order of inflorescence branches was increased to the fifth-order branches in the transgenic inflorescence, whereas the wild type (WT) inflorescence had only fourth-order branches ( Figure 4). Compared with the WT plants, notable increases in the female flower number (approximately three-fold) and male flower number (three-to five-fold) per inflorescence were observed in the transgenic Jatropha plants ( Figure 5A−F), which is similar to the phenotype of the Arabidopsis dad1 mutant [22]. The flower size of the transgenic plants was conspicuously larger than that of the WT plants ( Figure 6A−L), and the average diameters of the transgenic female and male flowers increased by 2−4 mm ( Figure 6M). These results show that JcDAD1 participates in flower development and promotes flower production and floral organ growth in Jatropha.

Figure 3. Relative expression levels of JcDAD1 in inflorescence buds of the wild type (WT) plants and
JcDAD1-RNAi transgenic Jatropha lines (L2, L3, L9, L10, and L12). Two biological replicates and three technical replicates were prepared for qRT-PCR assays. JcActin was used as the internal reference. Error bars represent standard errors (n = 2). Student's t-test was performed to analyze significant differences. ** Extremely significant difference (p < 0.01). * Significant difference (p < 0.05).  . Two biological replicates and three technical replicates were prepared for qRT-PCR assays. JcActin was used as the internal reference. Error bars represent standard errors (n = 2). Student's t-test was performed to analyze significant differences. ** Extremely significant difference (p < 0.01). * Significant difference (p < 0.05).

JcDAD1 Gene Silencing Increased Inflorescence Branching, Flower Number, and Flower Size
To investigate the functionality of the JcDAD1 gene, we transformed the JcDAD1-RNAi construct into Jatropha plants and identified 16 independent JcDAD1-RNAi transgenic lines (T1 generation), in which the transcript of JcDAD1 is repressed in inflorescence buds ( Figure 3). The order of inflorescence branches was increased to the fifth-order branches in the transgenic inflorescence, whereas the wild type (WT) inflorescence had only fourth-order branches ( Figure 4). Compared with the WT plants, notable increases in the female flower number (approximately three-fold) and male flower number (three-to five-fold) per inflorescence were observed in the transgenic Jatropha plants ( Figure 5A−F), which is similar to the phenotype of the Arabidopsis dad1 mutant [22]. The flower size of the transgenic plants was conspicuously larger than that of the WT plants ( Figure 6A−L), and the average diameters of the transgenic female and male flowers increased by 2−4 mm ( Figure 6M). These results show that JcDAD1 participates in flower development and promotes flower production and floral organ growth in Jatropha.

Figure 3. Relative expression levels of JcDAD1 in inflorescence buds of the wild type (WT) plants and
JcDAD1-RNAi transgenic Jatropha lines (L2, L3, L9, L10, and L12). Two biological replicates and three technical replicates were prepared for qRT-PCR assays. JcActin was used as the internal reference. Error bars represent standard errors (n = 2). Student's t-test was performed to analyze significant differences. ** Extremely significant difference (p < 0.01). * Significant difference (p < 0.05).

JcDAD1 Gene Silencing Caused the Abortion of Some Female Flowers but Did Not Affect Anther Dehiscence
At the late stage of flower development (in the inflorescences of 21-30 days after emergence), a portion of flowers were abortive and most were female flowers in the transgenic inflorescences, while flower development was normal in the WT inflorescences ( Figure 7A−L). Abortion might be caused by a deficiency of the nutrient supply. In Arabidopsis, the dad1 mutant displays a defect in anther dehiscence [22], although obvious phenotypic changes in anthers and pollens were not observed in the transgenic Jatropha (Figure 8). These results suggest that DAD1 may play different roles in the regulation of anther dehiscence and pollen development in Arabidopsis and Jatropha.

JcDAD1 Gene Silencing Caused the Abortion of Some Female Flowers but Did Not Affect Anther Dehiscence
At the late stage of flower development (in the inflorescences of 21-30 days after emergence), a portion of flowers were abortive and most were female flowers in the transgenic inflorescences, while flower development was normal in the WT inflorescences ( Figure 7A−L). Abortion might be caused by a deficiency of the nutrient supply. In Arabidopsis, the dad1 mutant displays a defect in anther dehiscence [22], although obvious phenotypic changes in anthers and pollens were not observed in the transgenic Jatropha (Figure 8). These results suggest that DAD1 may play different roles in the regulation of anther dehiscence and pollen development in Arabidopsis and Jatropha.

JcDAD1 Gene Silencing Affected Jatropha Yield Traits
Compared with that of WT Jatropha, the fruit number per infructescence in the transgenic plants was increased by 2-3-fold ( Figure 9A−E). The transgenic Jatropha plants produced smaller fruit and seeds in length and width ( Figure 10A−G), and had a lighter ten-seed weight ( Figure 10H), and lower seed oil content ( Figure 10I). The results indicate that JcDAD1 regulates fruit development and therefore affects the traits of seeds in Jatropha.

JcDAD1 Gene Silencing Affected Jatropha Yield Traits
Compared with that of WT Jatropha, the fruit number per infructescence in the transgenic plants was increased by 2-3-fold ( Figure 9A−E). The transgenic Jatropha plants produced smaller fruit and seeds in length and width ( Figure 10A−G), and had a lighter ten-seed weight ( Figure 10H), and lower seed oil content ( Figure 10I). The results indicate that JcDAD1 regulates fruit development and therefore affects the traits of seeds in Jatropha.     Scale bars = 2.0 cm. L2, L3 and L10 represent different transgenic lines. Error bars represent standard deviations. Student's t-test was performed to analyze significant differences. ** Extremely significant difference (p < 0.01).

JcDAD1 Gene Silencing Reduced Endogenous JA and JA-Ile Contents in Jatropha Inflorescences
To determine whether the endogenous jasmonate contents in the transgenic Jatropha plants were affected by JcDAD1, two types of jasmonate (JA and the bioactive form JA-Ile) were measured in developing inflorescences (15−20 days after emergence) from the WT and transgenic Jatropha plants. Based on the results, the concentrations of both JA and JA-Ile were significantly decreased in the transgenic Jatropha inflorescences (Figure 11), indicating that JcDAD1 is a key positive regulator of JA biosynthesis.
length and width in the WT and transgenic Jatropha plants, n ≥ 34. (F,G) Seed length and width in the WT and transgenic Jatropha plants, n ≥ 120. (H) Ten-seed weight in the WT and transgenic Jatropha plants, n ≥ 14. (I) Seed oil content in the WT and transgenic Jatropha plants, n ≥ 14. Scale bars = 2.0 cm. L2, L3 and L10 represent different transgenic lines. Error bars represent standard deviations. Student's t-test was performed to analyze significant differences. ** Extremely significant difference (p < 0.01).

JcDAD1 Gene Silencing Reduced Endogenous JA and JA-Ile Contents in Jatropha Inflorescences
To determine whether the endogenous jasmonate contents in the transgenic Jatropha plants were affected by JcDAD1, two types of jasmonate (JA and the bioactive form JA-Ile) were measured in developing inflorescences (15−20 days after emergence) from the WT and transgenic Jatropha plants. Based on the results, the concentrations of both JA and JA-Ile were significantly decreased in the transgenic Jatropha inflorescences (Figure 11), indicating that JcDAD1 is a key positive regulator of JA biosynthesis. Figure 11. Contents of JA (A) and JA-Ile (B) in early developing inflorescences from the wild type (WT) and the transgenic Jatropha plants. FW, fresh weight; JA, jasmonic acid; JA-Ile, jasmonic acidisoleucine. Error bars represent the standard deviations (n ≥ 3). Student's t-test was performed to analyze significant differences. ** Extremely significant difference (p < 0.01).

Discussion
JA, a pivotal phytohormone, plays diversified functions in inflorescence and flower development [49]. The initial step of JA biosynthesis is catalyzed by DAD1 [19][20][21]50]. The Arabidopsis dad1 mutant generally displayed male sterility and abnormal anther dehiscence that can be rescued by exogenous methyl jasmonate (MeJA) treatment [22]. Similarly, antisense inhibition of BrDAD1 also resulted in male sterility in Brassica rapa [27]. Compared with the male sterile phenotypes in Arabidopsis and Brassica, EG1, an ortholog of Arabidopsis DAD1, controlled both empty glume fate and spikelet development in rice [22,27,28,47,49]. In this study, JcDAD1 silencing caused a decrease of endogenous JA levels and an increase in flower number and flower size in the JcDAD1-RNAi transgenic inflorescences (Figures 3, 5, 6 and 11), suggesting that JcDAD1 regulates flower development by controlling JA levels in Jatropha inflorescences. However, the JcDAD1-RNAi transgenic plants did not exhibit obviously abnormal androecium (Figure 8), which is not consistent with that in Arabidopsis, thus implying that JA could play different roles in regulating androecium development in Jatropha and Arabidopsis. Furthermore, the phenotype of male flowers showed increased size but normal fertility in the transgenic plants, which is inconsistent with the speculation that JcDAD1 might contribute to male abortion in gynoecious Jatropha plants [44]. It is possible that JcDAD1 may function in adjusting the balance between reproductive development and stress response via the JA synthetic pathway because DAD1 can simultaneously act in flower development and wound defense and inhibit pathogen infection [7,22]. In general, cell size or cell number, which are mainly controlled by the genes involved in the biosynthesis or signal transduction of auxin, Figure 11. Contents of JA (A) and JA-Ile (B) in early developing inflorescences from the wild type (WT) and the transgenic Jatropha plants. FW, fresh weight; JA, jasmonic acid; JA-Ile, jasmonic acid-isoleucine. Error bars represent the standard deviations (n ≥ 3). Student's t-test was performed to analyze significant differences. ** Extremely significant difference (p < 0.01).

Discussion
JA, a pivotal phytohormone, plays diversified functions in inflorescence and flower development [49].
The initial step of JA biosynthesis is catalyzed by DAD1 [19][20][21]50]. The Arabidopsis dad1 mutant generally displayed male sterility and abnormal anther dehiscence that can be rescued by exogenous methyl jasmonate (MeJA) treatment [22]. Similarly, antisense inhibition of BrDAD1 also resulted in male sterility in Brassica rapa [27]. Compared with the male sterile phenotypes in Arabidopsis and Brassica, EG1, an ortholog of Arabidopsis DAD1, controlled both empty glume fate and spikelet development in rice [22,27,28,47,49]. In this study, JcDAD1 silencing caused a decrease of endogenous JA levels and an increase in flower number and flower size in the JcDAD1-RNAi transgenic inflorescences (Figures 3, 5, 6 and 11), suggesting that JcDAD1 regulates flower development by controlling JA levels in Jatropha inflorescences. However, the JcDAD1-RNAi transgenic plants did not exhibit obviously abnormal androecium (Figure 8), which is not consistent with that in Arabidopsis, thus implying that JA could play different roles in regulating androecium development in Jatropha and Arabidopsis. Furthermore, the phenotype of male flowers showed increased size but normal fertility in the transgenic plants, which is inconsistent with the speculation that JcDAD1 might contribute to male abortion in gynoecious Jatropha plants [44]. It is possible that JcDAD1 may function in adjusting the balance between reproductive development and stress response via the JA synthetic pathway because DAD1 can simultaneously act in flower development and wound defense and inhibit pathogen infection [7,22]. In general, cell size or cell number, which are mainly controlled by the genes involved in the biosynthesis or signal transduction of auxin, ethylene, cytokinin, and brassinosteroid, determine flower size in the plant kingdom [51,52]. An increase in flower size in JcDAD1-RNAi transgenic Jatropha plants ( Figure 6) suggests that JA might participate in the regulation of flower size in woody perennial plants. In the transgenic Jatropha plants, some female flowers and a few male flowers were abortive at the early stage of inflorescence development (Figure 7), which is likely caused by insufficient nutrient supply due to the generation of more flowers in a single inflorescence.
In summary, this study suggests that DAD1 genes are involved in the JA biosynthesis pathway and play diverse roles in regulating flower development among different species.
To investigate whether abnormal flower phenotypes of the transgenic plants can be rescued by exogenous jasmonate treatment, JA (2.5 or 5 mM) and MeJA (200 or 400 µM) solutions were sprayed onto the emerging transgenic inflorescence buds in the field. Unfortunately, whether JA/MeJA treatment could recover flowers in the transgenic inflorescences to normal size or normal flower number could not be clarified. Given the lack of control of environmental cues in the field trials, an optimized scheme will be designed for conducting experiments in future studies. Treatment with exogenous gibberellic acid (GA) was found to increase the number of female flowers in Jatropha [53][54][55], similar to the phenotypes of the JcDAD1-RNAi transgenic plants ( Figure 5E). Moreover, both exogenous benzyladenine (BA) treatment and flower-specific overproduction of endogenous cytokinins (CKs) promoted the total flower number and female flower number in Jatropha [38,56], which again resembled the phenotypes of the JcDAD1-RNAi transgenic plants ( Figure 5E,F). These results indicate that JA may act together with GA and/or CK to regulate flower development in Jatropha. However, the underlying molecular mechanism remains unclear and needs further investigation in the future.

Plant Materials
Two-year-old WT and JcDAD1-RNAi transgenic Jatropha plants were cultivated on the hillside land of an experimental field located in the Xishuangbanna Tropical Botanical Garden (XTBG; 21 • 54 N, 101 • 46 E; 580 m in altitude) of the Chinese Academy of Sciences, Menglun County, Yunnan Province, Southwest China [38]. Inflorescences from the WT and transgenic T1 generation plants were collected for morphologic observation. Roots, stems, young leaves, mature leaves, developing inflorescences (15−20 days after emergence), inflorescence buds, female flowers, male flowers, pericarps, and seeds were harvested from WT or transgenic plants, immediately frozen in liquid nitrogen and then stored at −80 • C for quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) analysis.

Sequence Alignment and Phylogenetic Analysis
The gene structures of DAD1 in different species were analyzed by the Gene Structure Display Server (GSDS2.0, Center for Bioinformatics, Peking University, China) with default settings [57]. The deduced DAD1 amino acid sequences for phylogenetic analysis from NCBI databases were used. A phylogenetic tree was constructed with the neighbor-joining statistical method, the Poisson model, and the bootstrap method applied in 1000 replications via MEGA (version 7.0) software (http://www.megasoftware.net/).

Isolation of JcDAD1 cDNA
Total RNA was extracted from Jatropha leaves using the pBIOZOL Plant Total RNA Extraction Reagent (BioFlux, Hangzhou, China) following the manufacturer's instructions. Then, first-strand cDNA was synthesized with 1.0 µg of the extracted RNA using the TAKARA PrimeScript TM RT Reagent Kit (TAKARA, Dalian, China). A 153 bp fragment (TCCGTCAATCAGATGGAGATACGTGCTTAGCTCGTGACATGTGGCCACGTCTCTTTTGTTTAGAT ACGGTGACTGTTTGCTACTGAGCCTAAGCTCTTGCCCCACCTCAGCATAAACCCACTGCGTGC TCTCTACTCGCTTCTGTAACCAACT) was PCR-amplified from the synthesized cDNA using primers carrying appropriate restriction enzyme cutting sites (all primers used in this study are listed in Supplementary Table S2). The PCR products were subsequently cloned into a pEASY-Blunt Zero cloning vector (TransGen Biotech, Beijing, China) with the appropriate restriction enzymes and sequenced.

RNAi Silencing Vector Construction and Transformation
To construct the JcDAD1-RNAi expression vector, the sense and antisense fragments of JcDAD1 were cloned into the pJL10 binary vector [58] in opposing orientations on either side of a pdk intron to produce an invert repeat driven by the 35S promoter. The expression vector was transformed into Jatropha with Agrobacterium strain EHA105 as described previously [59]. The positive transgenic plants were confirmed by PCR and qRT-PCR.

Quantitative RT-PCR (qRT-PCR)
qRT-PCR was performed using SYBR ® Premix Ex Taq TM II (TAKARA, Dalian, China) on a Roche 480 Real-Time PCR Detection System (Roche Diagnostics, Mannheim, Germany). At least two biological replicates and three technical replicates for all samples were applied in the qRT-PCR analysis. We used the 2 −∆∆CT method described by Livak and Schmittgen [60] to analyze the data. The JcActin gene was used to normalize the transcript levels of specific genes of Jatropha.

Phenotypic Analysis of Flowers
The phenotypes of heterozygous (T1 generation) transgenic Jatropha plants were analyzed using a 3D Super Depth digital microscope (Smart Zoom 5, Carl Zeiss, Germany,).

Quantitation of the JA and JA-Ile Contents
The 15-to 20-day-old inflorescences of the JcDAD1-RNAi transgenic and WT plants were harvested, frozen rapidly in liquid nitrogen and stored at −80 • C for measuring the JA and JA-Ile contents. The measurement method was described previously [61].

Conclusions
Silencing of JcDAD1 reduced endogenous jasmonate contents in inflorescences, increased the size and number of flowers, and caused the abortion of a few female flowers in Jatropha. Furthermore, the JcDAD1-RNAi transgenic Jatropha plants displayed increases in fruit number and decreases in seed size, weight and oil contents. However, compared with Arabidopsis, anther and pollen development in the JcDAD1-RNAi transgenic Jatropha plants was normal. These results indicate that JcDAD1 participates in JA biosynthesis and acts in regulating flower and fruit development in Jatropha. JcDAD1 is also highly expressed in the stems and roots, implying that JcDAD1 not only plays important roles in reproductive growth, but may also have undiscovered roles in vegetative growth. Since synergy may occur among JA, GA, and/or CK in regulating flower development of Jatropha, interactions of JA and other phytohormones to control flower development in Jatropha will be investigated in further work.