Genomics analysis of genes encoding respiratory burst oxidase homologs (RBOHs) in jatropha and the comparison with castor bean

Identification and manual curation of Rboh family genes

Arabidopsis and rice Rboh genes described before were retrieved from TAIR10 (Lamesch et al., 2012) or RGAP7 (Sakai et al., 2013), respectively, and their accession numbers are shown Table S1. At/OsRBOH proteins were used as queries to search for homologs from jatropha and castor bean genomes that were downloaded from NCBI (NCBI Resource Coordinators, 2018) or Phytozome v12 (Goodstein et al., 2012), respectively. Sequences with an E-value of less than 1E-5 in the tBLASTn search (Altschul et al., 1997) were collected and computationally predicted gene models were manually revised with cDNAs, ESTs (expressed sequence tags) and RNA-seq (RNA sequencing) reads that are available in NCBI (last accessed Jan 2018). Presence of the conserved NADPH_Ox domain in candidate RBOHs was confirmed using MOTIF Search (http://www.genome.jp/tools/motif/) and gene structures were displayed using GSDS (Hu et al., 2015). Homology search for nucleotides or ESTs and expression annotation using RNA-seq data were performed as previously described (Zou et al., 2015a; Zou et al., 2015b).

Synteny analysis, sequence alignment and phylogenetic analysis

Synteny analysis and multiple sequence alignment of deduced RBOH proteins were carried out as previously described (Zou, Xie & Yang, 2017; Zou, Zhu & Zhang, 2019b). Phylogenetic trees were constructed using MEGA (version 6.0) (Tamura et al., 2013), implementing the maximum likelihood method with a bootstrap of 1,000 replicates. Orthologs were determined using the BRH (Best Reciprocal Hit) method as described before (Zou et al., 2018) and further confirmed using the result from synteny analysis of jatropha and castor bean. Sequence alignment of Jc/RcRBOHs was displayed using Boxshade (http://sourceforge.net/projects/boxshade/).

Protein properties and conserved motif analysis

Various physical and chemical parameters of RBOH proteins were calculated using ProtParam (Gasteiger et al., 2005) and subcellular localization was predicted using Plant-mPLoc (http://www.csbio.sjtu.edu.cn/bioinf/plant/). Analysis of conserved motifs was performed using MEME (http://meme-suite.org/tools/meme): any number of repetitions distributed in sequences, the maximum number of 15 motifs, and the width of each motif ranging from six to 180 residues.

Gene expression analysis

Various transcriptome data were retrieved from NCBI SRA and detailed information is shown in Table S2. Raw reads were first filtered using fastQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/), and resulted clean reads were mapped to the coding sequences (CDS) of revised JcRbohs or RcRbohs as well as other protein-coding genes using Bowtie 2 (Langmead & Salzberg, 2012), and the relative transcript level was represented by FPKM (fragments per kilobase of exon per million fragments mapped, for pair-ended samples) or RPKM (Reads per kilobase per million mapped reads, for single-ended samples) (Mortazavi et al., 2008; Trapnell et al., 2010). The significance of gene expression difference was determined with parameters “FDR < 0.001” and “log2Ratio ≥ 1” with our in-house scripts. Unless specific statements, the tools used in this study were performed with default parameters.

Characterization of seven Rboh family genes in jatropha

A survey of the jatropha genome resulted in seven loci that were proven to encode Rboh genes. Considering the extensively functional analysis performed in arabidopsis, JcRbohB, JcRbohC, JcRbohD, JcRbohE, JcRbohF, and JcRbohH were named after their best orthologs in arabidopsis, whereas JcRbohN represents a novel group member with no ortholog in arabidopsis as well as rice (see below). Based on the result of mining RNA-seq data, all these genes were shown to be expressed. Moreover, JcRbohC, JcRbohD, and JcRbohF have corresponding ESTs in NCBI GenBank (Table 1). Compared with the original genome annotation, one gene model was manually optimized on the basis of read alignment: the locus JCGZ_22480 (JcRbohN) was predicted to contain 14 introns putatively encoding 755 residues (KDP26234) and actually it only contains 13 introns encoding 736 residues (see File S1). Based on available genetic markers (Wu et al., 2015), six Rboh-encoding scaffolds were further anchored to five out of the 11 chromosomes (Chrs) (Fig. 1).

Characterization of seven Rboh family genes in castor bean

To facilitate evolutionary analysis, Rboh family genes present in castor bean were also identified, resulting in seven genes that are also distributed across six scaffolds. RcRbohB and RcRbohD are located on the same scaffold, which is similar to JcRbohB and JcRbohD as observed in jatropha (Table 1), corresponding to the close phylogenetic relationship between these two Euphorbiaceous plants. A significant level of syntenic relation and one-to-one orthologous relationship were observed between jatropha and castor bean Rboh genes, supporting their divergence before speciation of these two species. As shown in Fig. 1, synteny analysis also allows anchoring all RcRbohs to five jatropha chromosomes. Based on manual curation, three of the computationally predicted gene models were optimized. The locus 30147.t000648 (RcRbohC) was predicted to encode 710 residues (30147.m014377), and it actually represents only the 3′ sequence of the gene which encodes 913 residues (see File S2). The locus 30128.t000051 (RcRbohD) was predicted to harbor 10 introns putatively encoding 709 residues (30128.m008590), which is relatively shorter than its ortholog in jatropha. In fact, RcRbohD was also proven to harbor 11 introns, though it contains at least two alternative splicing isoforms. The most common isoform encodes 916 residues (see File S3), whereas the short isoform (i.e., 30128.m008590) is derived from the skipping of the tenth intron. The locus 29941.t000003 (RcRbohN) was predicted to harbor 12 introns putatively encoding 666 residues (29941.m000220), which is relatively shorter than its ortholog in jatropha. Actually, read alignment indicated that this gene contains 13 introns encoding 731 residues (see File S4).

Phylogenetic analysis, exon-intron structures, and conserved motifs

To reveal the evolutionary relationship of Rboh family genes and facilitate the transfer of functional information obtained in model plants, an unrooted phylogenetic tree was constructed from full-length Jc/Rc/At/OsRBOHs, where arabidopsis and rice represent the model eudicot or monocot, respectively. As shown in Fig. 2A, these RBOHs were clustered into seven groups named RBOHC, -D, -B, -E, -F, -N, and -H, which is highly consistent with the BRH-based homologous analysis (Table 1). As for both jatropha and castor bean, a single member was found in each group. By contrast, recent duplicates were found in most groups of arabidopsis and rice Rboh gene families, which were shown to result from WGD as well as local duplication (see Table S1). However, RBOHC is absent from rice, whereas RBOHN is absent from both arabidopsis and rice, suggesting lineage-specific gene loss. The classification is further supported by exon-intron structure patterns as well as conserved motifs. As shown in Fig. 2B, RBOHB, RBOHC, and RBOHD feature 11 introns, whereas RBOHF, RBOHE, RBOHH, and RBOHN usually contain 13 introns. However, gene-specific gain or loss of certain introns was also observed, i.e., JcRbohB, JcRbohD, RcRbohN as well as AtRbohC, AtRbohG, AtRbohD, AtRbohH, AtRbohJ, OsRbohH, OsRbohA, and OsRbohD. The variety of exon-intron structure in both jatropha and castor bean seems to be considerably less frequent than that in arabidopsis and rice. Compared with exon, the intron length of Rboh genes examined in this study is relatively more variable, where the third intron of OsRbohA harbors the maximum of 6,173 bp (Fig. 2B).

Several parameters of deduced RBOH proteins were also computed as shown in Table 1. JcRBOHs and RcRBOHs consist of 736–953 or 731–940 amino acids (AA), respectively, and the average sequence length of 887 AA in these two species is relatively smaller than 906 AA in arabidopsis or 903 AA in rice. The average theoretical MW (molecular weight) of 100.72 kDa in jatropha and 100.86 kDa in castor bean is similar to 101.55 kDa in rice or 102.93 kDa in arabidopsis. The average pI (isoelectric point) value of 9.21 in jatropha and 9.11 in castor bean is also similar to 9.20 in arabidopsis or 9.34 in rice. The GRAVY (grand average of hydropathicity) value varies from −0.081 to −0.362, supporting their hydrophilic feature (Table 1 and Table S1 ). Subcellular localization analysis showed that all examined RBOHs are located to the cell membrane. In addition to six transmembrane α-helices containing two hemes, several typical domains such as Ferric_reduct, FAD_binding_8, NAD_binding_6, NADPH_Ox, and EF-hand motif were found in all Jc/RcRBOHs, though Jc/RcRBOHN lack a large part of the N-terminal as observed in other proteins (Fig. 3). In fact, the N-terminal of RBOHs was shown to be relatively variable related to the C-terminal.

Conserved motifs were further analyzed by using MEME as shown in Fig. 2C and Fig. S1. Among 15 motifs identified, Motifs 1–3, 5–12 are broadly distributed; Motif 4 is only absent from AtRBOHI and OsRBOHF, and Motifs 13–15 are absent from JcRBOHN and RcRBOHN. Moreover, Motif 13 is also absent from AtRBOHB and OsRBOHA; Motif 14 is also absent from AtRBOHC, AtRBOHA, AtRBOHG, AtRBOHD, OsRBOHD, and OsRBOHA; and, Motif 15 is also absent from AtRBOHA, AtRBOHI, AtRBOHE, OsRBOHF, and OsRBOHG. Motifs 13, 14, and 11 are part of the NADPH_Ox domain, which was proven to catalyze superoxide production; Motif 4 includes two EF-hand motifs, which are related to Ca²⁺-binding and Ca²⁺-dependent phosphorylation; Motifs 3, 12, 8, 9, and 6 are part of the Ferric_reduct domain that is similar to ferric reductase; Motif 1 is characterized as the FAD_binding_8 domain that bears the FAD-binding site; and, Motifs 5, 7, 2, and 10 are part of the NAD_binding_6 domain that bears the NAD-binding site (Fig. 2C).

Expression profiles of JcRboh genes

Despite the expression of all JcRbohs, the transcript of JcRbohH was only detected in late stages of ovule and stamen development as well as a mixed sample of root, mature leaf, flower, developing seed, and embryo (Natarajan & Parani, 2011; Hui et al., 2017), and the transcript level was shown to be extremely low. A global view of JcRboh expression profiles was further investigated based on transcriptomes of several typical tissues, i.e., roots from 15-day-old seedlings, half expanded and mature leaves from 4-year-old plants, developing seeds from fruits harvested 19–28 days after pollination (DAP), undifferentiated inflorescence of 0.5 cm diameter (IND), female flowers with carpel primordia beginning to differentiate (PID1), female flowers with three distinct carpels formed (PID2), male flowers with stamen primordia beginning to differentiate (STD1), and male flowers with ten complete stamens formed (STD2). Results showed that the total family transcripts are most abundant in PID1 (defined as Class I); moderate in IND, seed, STD2, PID2, root, and STD1 (defined as Class II, accounting for 36–59% of Class I); and, relatively low in mature leaf and leafage (defined as Class III, accounting for 21–27% of Class I). JcRbohC and JcRbohD are constitutively expressed in most tissues examined: JcRbohC occupies 72%, 63%, 46%, 25%, 23% or 23% of the total transcripts in seed, mature leaf, root, leafage, and IND, respectively; and, JcRbohD occupies 62%, 50%, 49%, 42%, 38% or 24% in STD2, PID1, STD1, PID2, INDs, and root, respectively. Additionally, JcRbohF occupies 68%, 27% or 21% of the total transcripts in leafage, mature leaf, and root, respectively, though its transcript level is extremely low in IND, PID1, PID2, STD1, and STD2. Compared with leafage, JcRbohC is upregulated by about 3.0 folds in mature leaf, whereas JcRbohF is downregulated by about 2.3 folds. Compared with IND, JcRbohD is upregulated by about 2.2 folds in PID1; JcRbohC is downregulated by about 2.3 folds in PID2; JcRbohB and JcRbohC are downregulated by about 2.9 or 2.7 folds in STD1, respectively; and, JcRbohC is downregulated by about 2.2 folds in STD2. Compared with PID1, there are 0.4, 0.5 or 0.4 fold-changes in PID2 for JcRbohB, JcRbohC, and JcRbohD, respectively. According to their expression patterns, seven JcRbohs could be classed into four main clusters: Cluster I is predominantly expressed in PID1, including JcRbohB, JcRbohD, and JcRbohN; Cluster II which includes JcRbohC is preferentially expressed in seed; Cluster III which includes JcRbohF is typically expressed in leafage; and Cluster IV which includes JcRbohE is mostly expressed in root (Fig. 4).

The response to drought and salt stresses was investigated in leaves and roots of 8-week old seedlings as described before (Zhang et al., 2014; Zhang et al., 2015). After withholding irrigation for 1, 4 or 7 d, all JcRbohs were shown to be significantly regulated at least one time point in at least one of the examined tissues. For 1 d, about 3.8-folds downregulation was observed for JcRbohD in leaf, whereas in root, decrease of about 2.6, 3.9, and 2.2 folds was observed for JcRbohC, JcRbohN or JcRbohE, respectively. For 4 d, about 4.8-folds decrease was observed for JcRbohN in root, while in leaf, increase of about 2.8 and 23.4 folds was observed for JcRbohB or JcRbohN, respectively. For 7 d, decrease of 2.0 and 2.2 folds was observed for JcRbohD or JcRbohF in root, respectively; by contrast, in leaf, increase of about 3.4, 10.7, 399.1 and 5.9 folds was observed for JcRbohB, JcRbohD, JcRbohN or JcRbohE, respectively. After the treatment with 100 mM NaCl for 2 h, 2 d or 7 d, two genes were significantly regulated (i.e., JcRbohD and JcRbohN): for 2 h, downregulation of JcRbohD (5.1 folds) and JcRbohN (7.8 folds) was observed in leaf or root, respectively; and upregulation of JcRbohN in both leaf and root was observed at later two time points, i.e., 5.3 and 12.3 folds for 2 d in leaf or root, respectively; and, 8.6 and 8.1 folds for 7 d in leaf or root, respectively (Fig. 5A).

The response to Colletotrichum gloeosporioides in leaf was analyzed on the basis of transcriptomes of two different genotypes, i.e., the susceptible RJ127 and the resistant 9-1. The result showed that JcRbohN was significantly upregulated in leaves of both the susceptible RJ127 and the resistant 9-1, though the fold-change is highly distinct (i.e., 10.4 vs 2.9 folds) (Fig. 5B). Additionally, in the case of RJ127, JcRbohD was significantly induced, whereas JcRbohF was inhibited (Fig. 5B), implying different regulation mechanism of these two cultivars.

Regulation of JcRbohs by hormones such as gibberellin acid (GA) and 6-benzylaminopurine (BA) (Ni et al., 2017) was also investigated. After application of 10 µM BA to young axillary buds for 12 h, the expression of JcRbohN increased by about 9.5 folds, by contrast, no such effect was observed for the same concentration of GA (Fig. 5B).