Two-year-old trees of H. brasiliensis clones of Wenchang11 were grown in the Rubber Research Institute of Hainan Agricultural Reclamation, Wenchang, Hainan, China. Rubber tree shoots were treated by 0.07% methyl jasmonate (JA), 0.5% Ethrel (ET), 200 μm abscisic acid (ABA), and 200 μm salicylic acid (SA) as described previously (Hao and Wu, 2000). Then, seven groups of 10 trees were used in each treatment, in which the plant growth regulator was applied at 1, 3, 6, 9, 12, 24, and 48 h before tapping. One group served as an untreated control. After the treatments at all time points, latex samples from all the tested trees were collected and mixed together thoroughly. The resulting solution was then divided into five equal volumes for RNA extraction (Tang et al., 2007). The other tissues (leaves, barks, roots, and flowers) of the rubber tree were then collected and stored in liquid nitrogen for RNA extraction.

DNA was extracted from young leaves of H. brasiliensis through the cetyl trimethylammonium bromide method (Allen et al., 2006). The total RNA from latex was isolated in accordance with Tang’s method (Tang et al., 2007), whereas that from the other tissues was extracted as described previously (Li et al., 2011). Three biological replicates were used for RNA extraction.

The analytical software NCBI-Blast-2.2.28+-win32 and the genomic data of H. brasiliensis (Rahman et al., 2013; Tang et al., 2016) were downloaded from the National Center for Biotechnology Information (NCBI¹). They were used to establish a local H. brasiliensis genome database. The MYB unigenes were obtained from the transcriptome database of the rubber tree latex in our previous study (Li et al., 2016a). The MYB unigenes were employed as query sequences for a BLAST search in the local rubber tree genome database. All candidate MYB genes were further analyzed for confirmation by using the NCBI Conserved Domain Search database². The physical and chemical properties of HblMYB were analyzed with ExPASy³), and the CDS of HblMYBs were analyzed with GSDS⁴. The predicted amino-acid sequences of HblMYBs were aligned with ClustalX. The highly conservative domains of HblMYB proteins were illustrated with espript⁵, and the 3D structure was constructed with SWISS-MODEL⁶.

Arabidopsis MYB protein sequences were downloaded from the phytozome⁷. Arabidopsis MYB protein sequences and the deduced amino-acid sequences of HblMYBs were aligned using ClustalX. Then, using the neighbor-joining method and the MEGA6.0 program, we constructed the phylogenetic tree between HlMYBs and the known MYB from Arabidopsis, and bootstrap analysis was conducted with 1,000 replicates (Yang et al., 2015).

The cDNA synthesis for qRT-PCR was performed with a RevertAid^TM First-Strand cDNA Synthesis Kit (Fermentas, Lithuania). qRT-PCR was performed using a SYBR Premix EX Taq Kit (TaKaRa, Japan). The primers for the HblMYBs were designed using the Primer Premier 5 software (Supplementary Table S1). HbACT7 was amplified through the following primers as standard control: 5′-TGTCAGCAACT GGGACGATATGG-3’ as primer 1 and 5′-GAGTCATCTTCTCTCTGTTGGC-3′ as primer 2 (Li et al., 2016a). qRT-PCR was performed as follows: 3 min at 95°C for denaturation, 40 cycles for 10 s at 95°C, 20 s at 58°C, and 25 s at 72°C. The quantitative value obtained from qRT-PCR is considered the cycle threshold (Ct). The normalized expression values for each gene were calculated through the following formula: 2^{-(Ct[gene]-Ct[HbACT7])}, in which HbACT7 was used as a housekeeping gene for normalization. Three individual reactions were replicated. Data were analyzed by ANOVA to analyze the significant differences on the basis of Fischer’s LSD test (P < 0.05 and P < 0.01; Quirk et al., 2016).

The open reading frames (ORFs) of HblMYBs were amplified by PCR using primers (Supplementary Table S1). PCR products were ligated into the pCAMBIA1302 vector, generating pHblMYBs-GFP. pCAMBIA1302 and pHblMYBs-GFP were introduced into the onion epidermis by Agrobacterium-mediated transformation. The transformed onion epidermis was then cultured on an MS solid medium in the dark at 26°C for 5 h and then observed with a confocal microscope (Zeiss LSM510, Germany).

The HbFDPS1 promoter (1,066 bp) and HbSRPP promoter (1,735 bp) were amplified by PCR with the primers as described previously (Guo et al., 2010, 2014). The 1,136 bp HRT1 promoter was amplified using the primers (Supplementary Table S1) based on the HRT1 sequence from the rubber tree genome database (GenBank accession: LVXX01000000; Tang et al., 2016). These promoters were cloned into the SpeI/MluI sites of the pHiS2.1 vectors (Clontech) to form the bait vectors pHiS-pHbSRPP, pHiS-HbFDPS1, and pHiS-HRT1. The ORFs of HbMYB19 and HbMYB44 were fused into the GAL4 domains of the pGAD7 vectors to generate the prey constructs pGAD-HbMYBs. The bait and prey vectors were then transformed into the yeast strain Y187 (Clontech). Afterward, the introduced yeast was cultivated on an SD/-Trp/-His/-Leu medium supplemented with 70 mM 3-amino-1,24-triazole (3-AT) at 30°C for 3 days.

The Dual-LUC assay was performed as described previously (Hellens et al., 2005). In brief, the promoters of HbSRPP, HbFDPS1, and HRT1 were cloned into pGreenII 0800 vectors, in which the expression of Renilla luciferase (REN-Luc) provided an internal control (Hellens et al., 2005). The ORFs of HbMYB19 and HbMYB44 were amplified with the primers (Supplementary Table S1) and then with the inserted pGreenII 62Sk vectors. All constructs were introduced into the Agrobacterium tumefaciens strain GV3103. The introduced GV3103 harboring pGreen-pHbSRPP, pGreen-pHbFDPS1, or pGreen-pHRT1 were mixed with the introduced GV3103 harboring pGreenII 62Sk-HblMYBs in a volume ratio of 1:5. The mixtures were injected into tobacco leaves. After culturing for 3 days, the infected areas of the leaves were obtained by puncher, and the protein was extracted. The activities of the luciferase and REN-Luc were measured through the Dual-LUC Reporter Assay System in accordance with the manufacturer’s manual (Promega). The binding ability of the HblMYBs to the promoters of HbSRPP, HbFDPS1, and HRT1 were represented by LUC/REN (ratio of LUC to REN-Luc). Three biological repeats were measured. The data were analyzed by ANOVA to determine the significant differences on the basis of the Fischer’s LSD test (P < 0.05 and P < 0.01) (Quirk et al., 2016).

The transcriptome database of the rubber tree latex was obtained in our previous study (Li et al., 2016a). A total of 76 MYB unigenes were obtained using our previously established latex transcriptome database. A total of 44 MYB genes were confirmed in the present study after the 76 MYB unigenes were searched in BLAST against those from the public rubber tree genome database. These MYB genes were named as HblMYB1 to HblMYB44. The sequences and properties of the identified 44 HblMYBs from the laticifer cells of the rubber tree are listed in Supplementary Table S2. The genomic structure of the 44 HblMYB genes was analyzed by GSDS. The 44 HblMYBs vary with respect to exon–intron gene structure. In particular, six HblMYBs contain only one exon, six HblMYBs contain two exons and one intron, 17 HblMYBs contain three exons and two introns, and the other HblMYBs containing more than five exons (Figure 1).

The number of amino acids in the HblMYB proteins ranges from 246 (HblMYB26, HblMYB33) to 1,043 (HblMYB24); protein molecular weights, from 27.24 (HblMYB10) to 115.12 (HblMYB24); and pI, from 5 (HblMYB37) to 9.44 (HblMYB3; Supplementary Table S2). Further analysis showed that the MYB domain (R unit) was highly conserved in the N-terminus. Among all the 44 HblMYBs, 5 HblMYBs contained one R unit, 33 HblMYBs had two R units, and 6 HblMYB contained three R units (Figure 1). The R2 repeats of R2R3-HblMYBs harbored three highly conserved tryptophan residues (W) at positions 6, 26, and 46 (Figure 2A). By contrast, the tryptophan residues were highly conserved at positions 25 and 44 of the R3 repeats (Figure 2B). Other highly conserved residues, except the tryptophan residues, were confirmed in the R2 and R3 domains. These residues included Gly (G4), Glu (E10), Asp (D11), Gly (G22), Arg (R37), Gly (G39), Lys (K40), Cys (C42), Arg (R43), Arg (R45), Asn (N48), Leu (L50), and Pro (P52) in the R2 repeat (Figure 2A), as well as Glu (E10), His (H18), Gly (G22), Asn (N23), Gly (G34), Arg (R35), Thr (T36), Asp (D37), Asn (N38), Lys (K41), and Asn (N42) in the R3 repeat (Figure 2B). Additionally, the 3D protein structure prediction showed that the R2 and R3 domains of the HblMYB proteins formed three α-helices (Figures 2C,D), which participate in transcriptional regulation (Dubos et al., 2010).

To infer the evolutionary relationships between HlMYBs and the known MYB from Arabidopsis, we constructed a phylogenetic tree between the obtained 44 HblMYBs and 126 Arabidopsis MYB TFs (Figure 3). All the 170 MYB were classified into 29 subgroups. Meanwhile, 44 HblMYB proteins were divided into 17 subgroups (S2, 5, 8, 9, 13, 14, 15, 17, 18, 19, 20, 22, and 23 and G1, 2, 3, and 4). Their orthologous MYBs were from Arabidopsis. This result suggests the existence of few closely related orthologous MYBs between rubber trees and Arabidopsis. Of the 29 subgroups, 10 subgroups (S1, 4, 6, 7, 10, 11 12, 16, 21, and 25) did not exhibit any rubber tree ortholog, and two subgroups (G1 and G4) did not present any Arabidopsis ortholog. The phylogenetic tree indicated the existence of an ancestral set of MYB genes prior to the divergence of rubber tree and Arabidopsis.

The expression patterns of 44 HblMYBs were detected in roots, barks, leaves, flowers, and latex by qRT-PCR (Figure 4). The results showed that the expression profiles of the 44 HblMYBs differed across different tissues. Five HblMYBs (HblMYB19, 20, 25, 40, and 44) showed higher transcription levels in latex, whereas 36 HblMYBs in the leaves, and 22 HblMYBs in the flowers. By contrast, all HblMYBs clearly presented with lower expression in the barks and roots.

Given the expression of HblMYBs in different tissues, five HblMYBs (HblMYB19, 20, 25, 40, and 44, which showed high expression levels in latex), were selected for further analysis on their response to exogenous phytohormone. Overall, the five HblMYBs showed different expression patterns under MeJA, ET, ABA, and SA treatments (Figure 5). Results showed that JA induced the expression of HblMYB19, 20, 40, and 44 but down-regulated that of HblMYB25 at the 6 h time point. ET treatment upregulated the transcript abundance of HblMYB20, 25, and 40 but downregulated those of HblMYB19 and 44 at the 12 or 9 h time point. ABA stress induced the expression of HblMYB44 at 24 h time point. However, we repressed the HblMYB19, 20, 25, and 40 expression at 12 or 6 h time point. Lastly, SA treatment repressed the HblMYB20, 25, and 44 expression at 24, 9, or 3 h time point, but did not significantly affect the HblMYB19 and 44 expression.

Given the expression of HblMYBs and response to JA in laticifers, HblMYB19 and HblMYB44 were selected for further analysis. First, subcellular localization analysis was performed on HblMYB19 and HblMYB44. We found that the GFP signals expressed the fusion proteins of HblMYB19, and 44 were present only in the nucleus of onion epidermal cells. By contrast, the GFP signals expressing the GFP protein were present obviously both in the nuclei and cytosol (Figure 6).

Farnesyl diphosphate synthase (FDPS), small rubber particle protein (SRPP), and Hevea cis-prenyltransferases or rubber transferase (HRT) from H. brasiliensis are responsible for the cis-1,4-polymerization of isoprene units from isopentenyl diphosphate (IPP) and implicated in NR yield (Light et al., 1989; Oh et al., 1999; Asawatreratanakul et al., 2003). To determine whether HblMybs bind the promoters of HbFDPS1, HbSRPP, and HRT1, we performed a yeast one-hybrid analysis. The yeast clones harboring pHblMyb19+pHIS2-pHbSRPP, pHblMyb19+pHIS2-pHRT1, pHblMyb19+pHIS2-pHbFDPS1, pHblMyb44+pHIS2-HbSRPP, pHblMyb44+pHIS2-pHRT1, and pHblMyb44+pHIS2-pHbFDPS1 can grow on SD/-Trp/-His/-Leu selective medium added with 70 mM 3-AT (Figure 7). This result indicated that HblMyb19 and HblMyb44 bound the promoters of HbSRPP, HRT1, and HbFDPS1 in yeast.

Given the interaction between yeast HblMyb19 and HblMyb44 and the promoters of HRT1, HbSRPP, and HbFDPS1, we investigated whether HblMyb19 and HblMyb44 participate in the regulation of the promoters of HbHRT, HbSRPP, and HbFDPS1 in plants. For this purpose, HblMyb19 and HblMyb44 were transiently expressed in tobacco by Agrobacterium-mediated transformation (Figure 8A). The luciferase activity controlled by the HblMyb19 or HblMyb44 binding of the promoters of HRT1, HbSRPP, and HbFDPS1 was elevated (Figure 8B). Moreover, the expression of HblMyb19 or HblMyb44 resulted in a significant increase in luciferase activity. The data showed that the transient expression of HblMyb19 and HblMyb44 activated the promoters of HRT1, HbSRPP, and HbFDPS1.