Genome-Wide Identification and Characterization of the JAZ Gene Family in Rubber Tree (Hevea brasiliensis)

Jasmonate signaling plays a vital role in the regulation of secondary laticifer differentiation and natural rubber biosynthesis in Hevea brasiliensis. Jasmonate ZIM-domain (JAZ) proteins are the master regulators of jasmonate signaling. Although several JAZs have been reported in the laticifer cells of H. brasiliensis, the genome-wide screening of HbJAZ members has not yet been explored. In the present study, 18 HbJAZs were identified based on the recent H. brasiliensis genome. Phylogenetic construction revealed that the HbJAZs were clustered into five subgroups and that members within the same subgroup shared highly conserved gene structures and protein motifs. Cis-element analysis of HbJAZ promoters suggested the presence of hormone, stress and development-related cis-elements. HbJAZ1.0, HbJAZ2.0, and HbJAZ5.0 interacted with CORONATINE INSENSITIVE1 (COI1) in the presence of coronatine (COR, a JA mimic). HbJAZ1.0, HbJAZ2.0, HbJAZ5.0, and HbJAZ12.0 could also interact with each other. Of the 18 HbJAZs, transcripts of 15 HbJAZs were present in the vascular cambium region except for that of HbJAZ7.0, HbJAZ8.0d, and HbJAZ13.0. Fourteen of the 15 HbJAZs were significantly up-regulated upon COR treatment. The transcripts of three genes that were absent from vascular cambium region were also absent from the latex. Among the 15 HbJAZs in the latex, the expression patterns of 13 HbJAZs were different between the tapping and ethrel treatments. Eight of the 14 COR-up-regulated HbJAZs in the vascular cambium region were also activated by tapping in latex. Of the eight tapping-activated HbJAZs, 5 HbJAZs were repressed by ethrel application. Based on the computational analyses and gene expression patterns described in this study, the HbJAZ5.0 and HbJAZ10.0b may be associated with laticifer differentiation while the HbJAZ8.0b is a negative regulator for natural rubber biosynthesis in H. brasiliensis.


INTRODUCTION
Jasmonates regulate various biological processes by the COI1-JAZ-MYC signaling pathway (Wasternack and Hause, 2013;Song et al., 2014;Yuan and Zhang, 2015). As repressors of jasmonate signaling, Jasmonate ZIM-domain (JAZ) proteins interact with MYC in the absence of active jasmonate (Singh et al., 2015). They are degraded by COI1-mediated ubiquitination in the presence of active jasmonate. This degradation releases MYC, which activates downstream gene transcription (Fonseca et al., 2009;Wager and Browse, 2012). The released MYC in turn activates the transcription of JAZ . The JAZ proteins are a subgroup of the TIFY family, characterized by the N-terminal located ZIM domain and the C-terminal located Jas domain (Thines et al., 2007;Melotto et al., 2008;Li et al., 2017). The ZIM domain, which consists of 30 amino acids with a highly conserved TIFY motif (TIF[F/Y]XG), is required for the formation of homo-and hetero-dimers within the JAZ proteins (Vanholme et al., 2007). The Jas domain, containing 29 amino acids with a conserved motif (SLX 2 FX 2 KRX 2 RX 5 PY), is specific to JAZ members and interacts with a wide range of other proteins, such as COI1, and members of the MYC, MYB, and bHLH families (Yan et al., 2007;Sheard et al., 2010;Pauwels and Goossens, 2011). As a master regulator of jasmonate signaling, JAZ has been shown to directly regulate plant morphology, flower initiation, cotton fiber initiation, salvianolic acids, and tanshinone biosynthesis (Boter et al., 2015;Hu et al., 2016;Pei et al., 2018;Yu et al., 2018).
As an important industrial raw material, natural rubber is mainly produced by the commercialized tropical plant, the rubber tree (Hevea brasiliensis) . The secondary laticifers located in the bark of the trunk of H. brasiliensis are the site closely related to natural rubber production (Hao and Wu, 2000). Latex, a milky cytoplasm of laticifers, contains 20-40% of the rubber and is used as the raw material for refining natural rubber (Chao et al., 2017;Zhang et al., 2017). Although application of ethrel (an ethylene releaser) causes a significant increase in rubber yield per tapping via prolonging the duration of latex flow (Zhu and Zhang, 2009), it inhibits rubber biosynthesis (Deng et al., 2018). Activation of jasmonate signaling in laticifer cells is closely associated with the tapping-enhanced natural rubber biosynthesis (Zhao, 2011;Cao et al., 2017;Deng et al., 2018). Moreover, jasmonates are the key signal molecules for the secondary laticifer differentiation from vascular cambia in H. brasiliensis (Hao and Wu, 2000;Tian et al., 2015;Wu et al., 2016).
Until now, the core JA signaling components, such as HbCOI1 and HblMYC1-2, have been identified in the latex of H. brasiliensis, and a total of 10 HbJAZs have been reported in different studies (Peng et al., 2009;Tian et al., 2010;Zhao et al., 2011;Pirrello et al., 2014;Hong et al., 2015). The publication of several versions of the genome of H. brasiliensis (Rahman et al., 2013;Lau et al., 2016;Tang et al., 2016;Pootakham et al., 2017;Makita et al., 2018) facilitates the genome-wide identification of JAZ family members. In the present study, a total of 18 HbJAZs were identified on the basis of the latest version of the H. brasiliensis genome (Tang et al., 2016), analysis of the gene structures, phylogenetics, interactions, and expression patterns of HbJAZ family members were performed, and candidate members that may be related to laticifer differentiation and rubber biosynthesis were suggested.

Plant Materials and Treatments
Plantlets, 10-year-old virgin trees and regularly tapped trees of the H. brasiliensis clone CATAS7-33-97 were grown in the field of the Experimental Farm of the Chinese Academy of Tropical Agricultural Sciences (CATAS) on Hainan Island. The plantlets were developed from the latent buds on pruned trunks and thus called epicormic shoots. The COR treatment and collection of vascular cambia-containing samples of epicormic shoots were previously described in Wu et al. (2016). At each time interval, the bark tissues were collected from 10 epicormic shoots and combined as one sample. The regularly tapped trees were tapped by using a half spiral pattern every 3 days (S/2, d/3). The virgin trees had not been tapped until sampling. Application of 1.5% ethrel on the tapping panel of the regularly tapped rubber trees were performed. The latex samples were collected 24 h after ethrel treatment. The latex samples were also collected from regularly tapped rubber trees without ethrel treatment and virgin rubber trees. The latex samples from 10 tapped trees with or without ethrel treatment and 10 virgin trees were combined as one sample, respectively. Three biological replications were conducted.

RNA Isolation, DNA Isolation, and cDNA Synthesis
Total RNA was extracted from samples using RNAplant Plus reagent and evaluated by the NanoDrop 2000 (Thermo Scientific, Inc., United States). Approximately, 1 µg of RNA was used for reverse transcription and cDNA was synthesized using RevertAid TM First Strand cDNA Synthesis Kit (Thermo Scientific, Inc., United States). Genomic DNA was extracted from the leaves of epicormic shoots using DNAplant Plus reagent.

Computational Analysis
The whole genome sequence of H. brasiliensis was downloaded from NCBI 1 (Tang et al., 2016). The Arabidopsis JAZ (AtJAZ) proteins were acquired from the TAIR database 2 . The HMM profiles of the TIFY domain (PF06200) and Jas domain (PF09425) were used as queries to search for predicted JAZ proteins in the H. brasiliensis genome using HMMER software (HMMER 3 ). The BLAST algorithm was also used to identify the predicted H. brasiliensis JAZ proteins using AtJAZ as the queries. All putative HbJAZ proteins were further confirmed by the CDD database 4 . The AtJAZs and HbJAZs were aligned using the online software Multiple Sequence Alignment 5 , and then the phylogenetic tree was constructed using iTOL online software 6 . The molecular weight and isoelectric points of the HbJAZ proteins were predicted from the ExPASy database 7 . The genes structures and protein motif analysis were performed by GSDS2.0 software 8 and MEME software 9 , respectively. For MEME analysis, the number of motifs was set as 12. The identified motifs were annotated by the InterProScan database 10 . Cis-acting regulatory elements within the HbJAZ promoters were analyzed by Plantcare 11 . The 1,500 bp genomic sequences upstream of the ATG of the HbJAZ genes were downloaded from the published H. brasiliensis genomes and used for the cis-regulatory element analysis.

Yeast Two-Hybrid Assay
The full-length CDSs of five HbJAZ proteins (HbJAZ1.0/2.0/5.0/10.0b/12.0) and HbCOI1 were amplified and inserted into the pGBKT7 vector to generate a bait plasmid. The CDS of HbJAZ genes were then fused into the pGADT7 vector to generate a prey plasmid. The bait vector and prey vector were transformed into Y187 and Y2H gold strains, respectively. The yeast Y187 and Y2H gold strains were mated, and protein interaction was assessed by the expression of different reporters under control of GAL4-responsive promoters. For the interactions between HbJAZs and HbCOI1, 20 or 60 µM COR and 10% ethanol were respectively added as treatment or MOCK, according to the reference (Melotto et al., 2008).

qRT-PCR Analysis
The primers for the HbJAZs were designed using the Primer Premier 5 software (Supplementary Table S1). The experiment was performed with the CFX384 real-time PCR system (Bio-Rad, United States) using the SYBR Prime Script RT-PCR Kit (TaKaRa, Dalian). HbUBC2b was used as the standard control for gene normalization . Three biological replicates were performed.

Statistical Analysis
The relative expression of HbJAZs were assessed by the 2 − Ct method. Significant differences in the expression level were tested by Student's t-test using GraphPad Prism 5 software.

Identification and Isolation of JAZ Family Proteins in H. brasiliensis
The HMM profiles of the TIFY domain (PF06200) and Jas domain (PF09425), as well as BLAST searches using Arabidopsis JAZ sequences as queries, were used to perform genome-wide identification of JAZ proteins in H. brasiliensis based on the most recent genome database. After validating the TIFY domain and Jas domain by NCBI Conserved Domain Search Service, a total of 18 HbJAZ proteins were identified. The identified HbJAZs were designated based on the names of their presumptive Arabidopsis orthologs (Figure 1). The 18 HbJAZs ranged from 127 (HbJAZ7.0) to 382 (HbJAZ9.0a) amino acid residues in length, with molecular weights of 14.39 kDa (HbJAZ7.0) to 40.82 kDa (HbJAZ9.0a) ( Table 1 and Supplementary Table S2). The predicted isoelectric points of these proteins ranged from 8.44 (HbJAZ7.0) to 9.91 (HbJAZ10.0b) ( Table 1).
Gene family expansion is a source of new gene generation (Jia et al., 2015). Tandem duplication and segmental duplication are the two strategies for expanding the number of genes in a family (Cannon et al., 2004;Jin et al., 2017). Compared to genes with tandem duplication, genes that come from segmental duplication at different chromosomes or scaffolds have been suggested to perform different functions (Dennis and Eichler, 2016). In rice, the clusters of OsJAZ9/10/11 and OsJAZ12/13/14 are from tandem duplication, while the clusters of OsJAZ3/4, OsJAZ2/15, and OsJAZ6/7 are from segmental duplication (Ye et al., 2009). In apple, Li et al. (2015) determine that the clusters of MdJAZ3/4/5/6/7/8, MdJAZ11/12, and MdJAZ13/14 are from tandem duplication, while the clusters of MdJAZ4/7, MdJAZ7/3, MdJAZ1/15, MdJAZ12/18, and MdJAZ18/11 are from segmental duplication. In grape, Zhang et al. (2012) report that VvJAZ5/6/7/8 are from tandem duplication, while the pairs of VvJAZ1/11 and VvJAZ4/9 are from segmental duplication. In rubber tree, the lack of high-density genetic map precludes the identification of paralogs derived from Whole Genome Duplication (WGD) or segmental duplication. However, Zou et al. (2016Zou et al. ( , 2017Zou et al. ( , 2019 identify several segmental duplication events in rubber tree genome, and firstly reveal that ρ WGD occurred in rubber tree is shared by cassava. In the present study, we identified at least three clusters (HbJAZ3.0/4.0, HbJAZ9.0a/9.0b, HbJAZ10.0a/10.0b) that were segmental duplications while one pair (HbJAZ8.0c/8.0d) was tandem duplicated, based on the scaffold location, gene structure, and protein conserved motifs (Figure 1 and Table 1). It is clear that the two segmentally duplicated clusters (HbJAZ3.0/4.0, HbJAZ9.0a/9.0b) have differential expression patterns in response to COR, ethrel or tapping treatments, suggesting the putative functional diversity of these genes. Phylogenetic tree analysis of JAZs from multiple species shows that the JAZ members that come from segmental duplication separate into different clusters while those from the tandem duplication are clustered together. The exception to this is JAZ members in rice (Supplementary Figure S3 and Supplementary Table S4), which may be attributed to the fact that rice is a monocot while the other species are dicots.
It has been suggested that JA signaling is involved in the regulation of plant development (Singh et al., 2015;Huang et al., 2017). In Arabidopsis, AtJAZ7 and AtJAZ10 are the key regulators in JA-mediated cambium differentiation (Sehr et al., 2010). Li et al. (2015 show knockdown of AsJAZ1 decreases the number of nodules in Astragalus sinicus. In cotton, over-expression of GhJAZ2 inhibits both lint and fuzz fiber initiation and reduces the fiber length (Hu et al., 2016). Over-expression of SlJAZ2 in tomato directly activates meristem-related genes and results in production of more leaves and flowers (Yu et al., 2018). To further explore the key members involved in secondary laticifer differentiation of rubber trees, we constructed a phylogenetic tree using the 18 HbJAZs together with AtJAZ7, AtJAZ10, AsJAZ1, GhJAZ2, SlJAZ2 (Supplementary Figure S4 and Supplementary Table S4). It is intriguing that 11 of the HbJAZs (HbJAZ1.0/2.0/5.0/6.0/7.0/8.0a/8.0b/8.0c/8.0d/10.0a/10.0b) are homologous to the plant development-related homologs, and five of them (HbJAZ5.0/8.0b/8.0c/10.0a/10.0b) are significantly up-regulated greater than 10 times upon COR treatment. Moreover, a meristem specific element (CATbox) is present in the promoter regions of HbJAZ5.0 and HbJAZ10.0b. The CAT-box is a cis-regulatory element related to meristem expression. In Brassica juncea, mutation of the CAT-box in the promoter of BjuA07.CLV1 disturbs the CLAVATA (CLV)/WUSCHEL (WUS) signaling pathway, leading to the enlargement of the shoot and floral meristem, and ultimately multilocular siliques (Xiao et al., 2018). Thus, we speculate that HbJAZ5.0 and HbJAZ10.0b may play a key role in laticifer differentiation in H. brasiliensis.
Jasmonate is widely involved in the regulation of secondary metabolism in plants. Boter et al. (2015) show that degradation of JAZ3 activates a subset of JA-regulated genes in leaves, leading to anthocyanin accumulation in Arabidopsis. Mutation of the Jas motif of NtJAZ1 or NtJAZ3 reduces nicotine content in tobacco BY-2 cells (Shoji et al., 2008). In Salvia miltiorrhiza, over-expression of SmJAZ3 or SmJAZ8 notably reduces tanshinone production Pei et al., 2018). Natural rubber is a typical secondary metabolite and its biosynthesis is significantly activated by tapping (Deng et al., 2018). The tapping-enhanced rubber biosynthesis is closely related to the activation of jasmonate signaling in laticifer cells (Deng et al., 2018). Exogenous methyl jasmonate enhances rubber biosynthesis while ethrel, inhibits rubber biosynthesis (Deng et al., 2018). In the present study, the expression patterns of HbJAZs in response to tapping were similar to those in response to COR, but generally opposite to the patterns in response to ethrel treatment (Figures 5, 6). The differences in the expression levels of 13 of the HbJAZs between the tapping and ethrel treatments may be associated with the different effects of methyl jasmonate and ethrel on rubber biosynthesis. Phylogenetic tree analysis reveals that 11 HbJAZs (HbJAZ1.0/2.0/3.0/4.0/7.0/8.0a/8.0b/8.0c/8.0d/9.0a/9.0b) are homologous to AtJAZ3, NtJAZ1, NtJAZ3, SmJAZ3, and SmJAZ8 (Supplementary Figure S4 and Supplementary  Table S4), and tapping causes the expression level of HbJAZ8.0a and HbJAZ8.0b more than five times of control ( Figure 6A). We also identified a secondary MBSI in the promoter region of HbJAZ8.0b. MBSI is present in the promoter of flavonoid biosynthetic genes, such as chsJ.
In Petunia hybrida, the petal epidermis-specific MYB.Ph3 binds the MBSI element that regulates the biosynthesis of flavonoids (Solano et al., 1995). Based on the evidence described above, HbJAZ8.0b is likely a key regulator of natural rubber biosynthesis.

CONCLUSION
Jasmonate signaling plays a vital role in the regulation of secondary laticifer differentiation and natural rubber biosynthesis. In the present study, JAZs, the repressors of jasmonate signaling, are genome-wide identified. Based on the computational analyses and gene expression patterns, HbJAZ5.0 and HbJAZ10.0b might be key regulators of laticifer differentiation whereas HbJAZ8.0b was crucial for regulation of natural rubber biosynthesis in H. brasiliensis. The genome-wide identification of HbJAZs will facilitate the jasmonate signaling-mediated laticifer differentiation and natural rubber biosynthesis in rubber tree.

AUTHOR CONTRIBUTIONS
JC designed and carried out the experiment of this study and wrote the manuscript. YZ, JJ, SW, XM, and YC participated and analyzed data in the experiment. W-MT planned the study and participated in the design of the experiment. All authors have read and approved the manuscript in its final form.

SUPPLEMENTARY MATERIAL
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fgene. 2019.00372/full#supplementary-material FIGURE S1 | Conserved amino acid sequences of 12 motifs.  The trees were constructed by neighboring-joining phylogeny test, and 1,000 bootstrap replicates. The accession numbers for the genes are provided in Supplementary Table S4. Red color represented tandem duplication, while blue color represented segmental duplication. FIGURE S4 | Phylogenetic relationships of HbJAZs with other species homolog related to plant development or secondary metabolites biosynthesis. The trees were constructed by neighboring-joining phylogeny test, and 1,000 bootstrap replicates. The accession numbers for the genes are provided in Supplementary  Table S4. Blue circle represented pigments biosynthesis related TFs in other species. Red circle represented genes identified in this study.
TABLE S1 | Primers used in this paper.