Skip to main content
Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Survey of the rubber tree genome reveals a high number of cysteine protease-encoding genes homologous to Arabidopsis SAG12

  • Zhi Zou ,

    Affiliation Danzhou Investigation & Experiment Station of Tropical Crops, Ministry of Agriculture, Rubber Research Institute, Chinese Academy of Tropical Agricultural Sciences, Danzhou, Hainan, P. R. China

  • Jianting Liu,

    Affiliations Danzhou Investigation & Experiment Station of Tropical Crops, Ministry of Agriculture, Rubber Research Institute, Chinese Academy of Tropical Agricultural Sciences, Danzhou, Hainan, P. R. China, Crops Research Institute, Fujian Academy of Agricultural Sciences, Fuzhou, Fujian, P. R. China

  • Lifu Yang,

    Affiliation Danzhou Investigation & Experiment Station of Tropical Crops, Ministry of Agriculture, Rubber Research Institute, Chinese Academy of Tropical Agricultural Sciences, Danzhou, Hainan, P. R. China

  • Guishui Xie

    Affiliation Danzhou Investigation & Experiment Station of Tropical Crops, Ministry of Agriculture, Rubber Research Institute, Chinese Academy of Tropical Agricultural Sciences, Danzhou, Hainan, P. R. China


Arabidopsis thaliana SAG12, a senescence-specific gene encoding a cysteine protease, is widely used as a molecular marker for the study of leaf senescence. To date, its potential orthologues have been isolated from several plant species such as Brassica napus and Nicotiana tabacum. However, little information is available in rubber tree (Hevea brasiliensis), a rubber-producing plant of the Euphorbiaceae family. This study presents the identification of SAG12-like genes from the rubber tree genome. Results showed that an unexpected high number of 17 rubber orthologues with a single intron were found, contrasting the single copy with two introns in Arabidopsis. The gene expansion was also observed in another two Euphorbiaceae plants, castor bean (Ricinus communis) and physic nut (Jatropha curcas), both of which contain 8 orthologues. In accordance with no occurrence of recent whole-genome duplication (WGD) events, most duplicates in castor and physic nut were resulted from tandem duplications. In contrast, the duplicated HbSAG12H genes were derived from tandem duplications as well as the recent WGD. Expression analysis showed that most HbSAG12H genes were lowly expressed in examined tissues except for root and male flower. Furthermore, HbSAG12H1 exhibits a strictly senescence-associated expression pattern in rubber tree leaves, and thus can be used as a marker gene for the study of senescence mechanism in Hevea.


Leaf senescence, the last stage of leaf development, is a complex but highly regulated developmental process that is controlled by an array of internal and external factors [1,2]. Internal factors include age and levels of plant hormones, whereas external factors include shade/darkness, desiccation, drought, heat, nutrient stresses, pathogen infection and various hormones such as cytokinin (CK), auxin, ethylene (ET), jasmonic acid (JA), salicylic acid (SA) and abscisic acid (ABA) [3,4]. These factors trigger a series of coordinated events such as shifts in gene expression, loss of chlorophyll, reduction of photosynthesis, degradation of macromolecules, relocation of nutrients, the breakdown of organelles, and finally, cell death [5,6]. In the senescent leaves typical with a yellow color [7], protein degradation is one of the most important hydrolytic processes and many SAGs (senescence-associated genes) encoding proteases are synthesized de novo or induced [8]. SAG12, a well-known Arabidopsis gene encoding a papain-like cysteine protease, exhibits a strictly senescence-associated expression pattern in leaves and flowers [911]. AtSAG12 is widely used as a molecular marker for the study of developmental senescence, since its expression is developmentally controlled and cannot be induced by various stresses (e.g. desiccation, dark incubation and wounding) or hormones (e.g. ABA and ET) in young, detached Arabidopsis leaves [11]. However, in mature leaves that are ready for senescence, it can be induced by detachment, ABA and ET [12]. Although the exact physiological role of AtSAG12 is still not known, its colocalization with Rubisco and other stromal proteins in SAVs (senescence-associated vacuoles) in Arabidopsis was observed [13]. Moreover, studies indicated that the AtSAG12 promoter is the target of transcription factors AtWRKY53 (a senescence inducer) and AtWRKY57 (a senescence inhibitor) [14,15], and the promoter has been fused to IPT to form an autoregulatory senescence inhibition system widely applied to plant species [16,17]. To date, the potential orthologues of AtSAG12 were isolated from several species, e.g., BnSAG12-1 and BnSAG12-2 in Brassica napus [18], SPG31 in Ipomoea batatas [19], ccyp in Gossypium hirsutum [20], NtCP1 and NtSAG12 in Nicotiana tabacum [21,22].

Para rubber tree (Hevea brasiliensis Muell. Arg., 2n = 36), a perennial big tree of the Euphorbiaceae family, is economically important for the production of natural rubber (cis-1,4-polyisoprene), an essential industrial raw material [23,24]. Although native to the Amazon basin, the increasing demand of natural rubber has prompted the commercial cultivation of rubber trees in the tropical regions of Asia, Africa, and Latin America [25]. Natural rubber is specifically synthesized and stored in the highly differentiated vessels termed laticifers, which are located in the soft inner bark of the tree trunk and periodically differentiated from the vascular cambium [26]. Upon tapping, the laticifer cytoplasm is expelled in the form of latex due to the high turgor pressure inside [27,28]. Over the past decades, the latex yield has been significantly increased for the wide cultivation of high-yielding clones and the extensive utilization of ethephon (an ET generator) [29,30]. Meanwhile, steadily increased occurrence of a complex physiological syndrome called TPD (Tapping Panel Dryness), which is characterized as tapping incision blocked partly or entirely during latex exploiting, cause great losses of latex production [30,31]. Although a great deal of effort has been made on TPD, the molecular mechanism underlying remains poorly understood. Nuclear DNA fragmentation, upregulation of a high number of senescence-associated genes and downregulation of many anti-apoptosis-associated genes in TPD-affected trees [3236] suggest that TPD is a type of programmed cell death (PCD), probably due to overproduction of reactive oxygen species (ROS) resulted from high-strength tapping and ET over-stimulating [37]. Nevertheless, the functional characterization of TPD-associated genes is hindered for several reasons: rubber tree is a perennial plant species with long breeding time and long juvenile phase of 7–8 years before tapping, where the genetic transformation system is still not well established yet [25]; the induced TPD is destructive on rubber production; and, no laticifer can be found in model plants such as Arabidopsis and tobacco [33]. As a type of PCD, TPD is more likely to share some features of leaf senescence. Thereby, the information about rubber tree leaf senescence may improve our knowledge on TPD.

To explore molecular markers for the study of senescence mechanism in rubber tree, the present study took advantage of the recently available genome sequences [38] to identify the potential orthologues of AtSAG12. Furthermore, their evolutionary pattern and expression profiles over various tissues including natural and ET-induced senescent leaves were also investigated.

Materials and methods

Identification of SAG12-like genes in the rubber tree genome

All the deduced protein sequences of Arabidopsis were downloaded from TAIR10 [39]. To identify the orthologues of Arabidopsis SAG12 in rubber tree, a two-step approach was used. First, the amino acid sequence of AtSAG12 (TAIR10 accession number AT5G45890) was used as the query to search the Reyan7-33-97 genome [38], and sequences with an E-value of less than 1e-5 in the TBLASTN search [40] were collected; the positive genomic sequences were predicted using GeneMark.hmm [41], and the gene models were further validated with cDNAs, expressed sequence tags (ESTs) and RNA sequencing reads when available. Homology search for nucleotides or ESTs was performed using BLASTN, and sequences with the identity of more than 98% were taken into account. RNA sequencing data available in NCBI SRA ( were also used for the expression annotation: the reads were first filtered by removing adaptor sequences, adaptor-only reads and low quality reads containing more than 50% bases with Q-value≤5; read alignment was performed using Bowtie 2 [42], and mapped read number of more than one was counted as expressed. Unless specific statements, the tools used in this study were performed with default parameters.

Subsequently, the deduced amino acid sequences of putative homologues were used as the queries to search the local Arabidopsis proteome database; if the best hit in the BLASTP search was AtSAG12, the gene was defined as the true orthologue of AtSAG12. Tandem or proximal duplications were considered when two duplicated genes were consecutive in the genome and separated by 20 or fewer gene loci, respectively.

Identification of SAG12-like genes in another two Euphorbiaceae plants

Rubber tree is a diploid plant that was shown to have undergone a recent whole-genome duplication (WGD) event [38,43]. To investigate the recent evolutionary pattern of rubber tree SAG12 homologues, a similar approach as described above was adopted to identify orthologues from the genomes of another two Euphorbiaceae plants, castor bean (Ricinus communis L.) and physic nut (Jatropha curcas L.). Both castor and physic nut underwent no recent WGD [44,45], and their genome sequences were downloaded from Phytozome v11 [46] and NCBI GenBank, respectively.

Sequence alignments and phylogenetic analysis

Multiple alignments were performed using MUSCLE [47]. The alignment was displayed using BoxShade ( The unrooted phylogenetic tree was constructed using MEGA 6.0 [48] with the maximum likelihood method and with the bootstrap test replicated 1000 times.

Protein properties and conserved motif analysis

Protein properties such as the molecular weight (MW), isoelectric point (pI), and grand average of hydropathicity (GRAVY) were calculated using ProtParam ( The protein subcellular localization was predicted using iPSORT [49], and the location of signal peptide cleavage site was predicted using SignalP 4 [50]. Analysis for conserved motifs in proteins was performed using MEME [51]. The optimized parameters were: any number of repetitions; maximum number of motifs, 15; and the optimum width of each motif, between 6 and 50 residues. And the MAST program was used to search detected motifs in protein databases.

Plant materials, RNA isolation and cDNA synthesis

In vitro plantlets of clone Reyan7-33-97 used in this study were obtained via secondary somatic embryogenesis [52], and the bagged plantlets were grown in a greenhouse illuminated with natural sunlight. Various tissue samples such as root, bark, laticifer, xylem, shoot apex, leaf and petiole were collected and subjected to total RNA extraction as described before [53]. The leaf tissue included leaves of different developmental stages such as bronze, color-change, pale-green, mature, early-senescent and mid-senescent, where the early- and mid-senescent leaves were defined by the chlorophyll content of 75–85% and 45–55% relative to mature leaves, respectively. The mature leaves were also treated with 50 μM ethephon to induce senescence in vitro, and the induced senescent leaves were collected when the chlorophyll content was dropped to 45–55%.

The isolated RNA was subsequently treated with RNase-Free DNase I (Takara), and the first-strand cDNA was synthesized from 2 μg of total RNA with M-MLV reverse transcriptase (Promega) according to the manufacturer's instructions.

Gene expression analysis

Along with the genome sequencing, we have also sequenced several tissue transcriptomes of Reyan7-33-97 with the Illumina platform, i.e., root (NCBI SRA accession number SRX1554786), leaf (SRX1554799), bark (SRX1554797), laticifer (SRX1554800), female flower (SRX1554813), male flower (SRX1554814) and seed (SRX1554817) [38]. Thereby, the relative expression levels of HbSAG12H genes in these tissues were first examined: the filtered reads were mapped to the coding sequences (CDS) of identified HbSAG12H genes using Bowtie 2 [42], and the FPKM (fragments per kilobase of exon per million fragments mapped) method [54] was adopted for the quantification.

To identify HbSAG12H genes expressed in senescent leaves, 17 primer pairs (see S1 Table) were designed according to the genome sequences, and the RT-PCR (reverse transcriptase polymerase chain reaction) was performed to amplify the target CDS by using the synthesized cDNAs as the template. The PCR products were cloned into the pMD18-T vector (Takara) and sequenced with an ABI3730xl DNA Analyzer (Life Technologies). Semi-quantitative RT-PCR analysis was performed with gene-specific primers (SAG12H1F: 5’ AAC CCT TTG TCG TCC TCT GG 3'; SAG12H1R: 5’ TTT GCT TCT CGT CTG CGT CT 3'), and the Hb18S rRNA [53] was used as the internal control. The PCR conditions were listed in S1 File, and at least three replicates were performed for each sample of three biological replicates. A 10 μL sample of the PCR products was analyzed by electrophoresis on 1.5% agarose gel containing ethidium bromide.


Characterization of 17 SAG12-like genes in rubber tree

The initiative BLAST search resulted in 54 loci putatively encoding Arabidopsis SAG12 homologues from the rubber tree genome, and 17 out of them were proved to be true orthologues by the reciprocal BLASTP. These orthologues were denoted as HbSAG12H117, which were found to be distributed across 9 out of 7,453 scaffolds [38], i.e., scaffold0048 (5), scaffold0247 (2), scaffold0583 (2), scaffold0696 (2), scaffold1445 (1), scaffold2360 (2), scaffold0683 (1), scaffold0420 (1) and scaffold1086 (1) (Table 1). Interesting, a high number of HbSAG12H genes are organized in cluster, and the CDS of these genes exhibit relatively high identity, e.g., 97.8% between HbSAG12H2 and HbSAG12H3, 80.2% between HbSAG12H6 and HbSAG12H7, 96.0% between HbSAG12H8 and HbSAG12H9, 99.8% between HbSAG12H10 and HbSAG12H11, 78.2% between HbSAG12H13 and HbSAG12H14, 80.0% between HbSAG12H15 and HbSAG12H16.

Table 1. List of HbSAG12H genes identified in this study.

Although no corresponding ESTs can be found in GenBank (as of July 2016) for all HbSAG12H genes, read alignment supported their expression in at least one of the examined tissues, i.e., shoot apex, leaf, laticifer, bark, root, flower, seed and somatic embryogenesis [38,5559] (Table 1). The cDNAs of HbSAG12H1–4, HbSAG12H6, HbSAG12H13–17, can also been isolated from roots or senescent leaves with gene-specific primers via RT-PCR (data not shown). Without any exception, all HbSAG12H genes were shown to contain a single intron. Compared with the similar length of CDS (1020–1044 bp), the gene size (from start to stop codons) of HbSAG12H genes is highly distinct (1111–2215 bp). The average length of the intron is about 190 bp, with the minimum of 83 bp in HbSAG12H15 and the maximum of 1177 bp in HbSAG12H1 (Table 1).

Characterization of SAG12-like genes in castor and physic nut

The homology search resulted in 8 RcSAG12Hs and 8 JcSAG12Hs from the genomic sequences of castor and physic nut, respectively. The expression of these genes were all supported by available RNA sequencing reads, and a single intron was observed as that in rubber tree (Tables 2 and 3).

Table 2. List of JcSAG12H genes identified in this study.

Table 3. List of RcSAG12H genes identified in this study.

In physic nut, the identified JcSAG12H genes were shown to be distributed across 5 out of 6,023 scaffolds [44], i.e., scaffold26 (1), scaffold684 (3), scaffold84 (1), scaffold5 (1) and scaffold872 (2) (Table 2). Based on the linkage map containing 1208 markers [44], these scaffolds can be further anchored to 4 chromosomes, i.e., LG3 (scaffold26 and scaffold684), LG2 (scaffold84), LG4 (scaffold5) and LG9 (scaffold872). Compared with the automatic annotation, one more locus (denoted as JcSAG12H8) was predicted from scaffold684 (see S2 File). In addition, three pseudogenes, i.e., JCGZ_05109 and another two loci, were also identified from scaffold170. Three gene pairs exhibit high levels of homology, i.e., 97.8% between JcSAG12H1 (JCGZ_09604) and JcSAG12H2 (JCGZ_21557), 88.6% between JcSAG12H5 (JCGZ_25372) and JcSAG12H6 (JCGZ_25371) and 76.4% between JcSAG12H7 (JCGZ_21549) and JcSAG12H8. JcSAG12H5/6 and JcSAG12H7/8 can be defined as tandem duplications, whereas JcSAG12H1 and JcSAG12H2 were defined as proximal duplications for their distribution on two distinct scaffolds of chromosome 3. Interesting, 3 out of the 7 annotated gene models were proved to be inaccurate. The locus JCGZ_09604 (JcSAG12H1) was predicted to encode 311 residues, however, read alignment indicated that it represents only the 3’ sequence of the gene which is promised to encode 345 residues (see S3 File). The locus JCGZ_21557 (JcSAG12H2) was predicted to encode 155 residues, however, read alignment indicated that it represents only the 5’ sequence of the gene which is promised to encode 345 residues (see S4 File). The locus JCGZ_21549 (JcSAG12H7) was predicted to encode 324 residues, however, read alignment indicated that it represents only the 3’ sequence of the gene which is promised to encode 344 residues (see S5 File).

In castor, 8 RcSAG12H genes are also distributed across 5 out of 25,878 scaffolds [43], i.e., scaffold30131 (1), scaffold28962 (2), scaffold29646 (2), scaffold29900 (1) and scaffold29910 (2) (Table 3). These genes on the same scaffold were defined as tandem duplications for their close organization and high sequence identity, i.e., 97.4% between RcSAG12H7 (29910.t000015) and RcSAG12H8 (29910.t000014), 96.8% between RcSAG12H2 (28962.t000017) and RcSAG12H3 (28962.t000018), and 87.5% between RcSAG12H4 (29646.t000033) and RcSAG12H5 (29646.t000034). Compared with RcSAG12H4, the CDS length of RcSAG12H5 is 21-bp shorter which was resulted from a C/T mutation at the 3’ terminus, and this gives rise to relatively low identity between these two genes. In addition, a pseudogene (29900.t000066) derived from RcSAG12H6 (29900.t000065) via tandem duplication was also found. Except for RcSAG12H8, the transcription regions of other seven genes were successfully extended based on read alignment. It's worth noting that the CDS length of RcSAG12H1 is considerably longer than other orthologues in castor as well as that in physic nut and rubber tree, which was shown to be resulted from fragment loss of its 3’ terminus.

Phylogenetic analysis of HbSAG12Hs

To reveal the evolutionary relationships of rubber SAG12 orthologues, an unrooted phylogenetic tree was constructed using MEGA6 from 17 HbSAG12Hs together with AtSAG12, 8 RcSAG12Hs, 8 JcSAG12Hs and orthologues reported in other plant species. As shown in Fig 1, these SAG12 orthologues exhibit the identity of 42.1% (RcSAG12H1) to 84.4% (BnSAG12-1) with AtSAG12, and were split into four main groups containing members from at least two examined species. Group I includes HbSAG12H1, JcSAG12H1–2, AtSAG12 and other reported orthologues. HbSAG12H1 and two JcSAG12Hs were clustered together, showing closer relationship with two Nicotiana orthologues than AtSAG12. Group II includes HbSAG12H2–4, JcSAG12H3–6 and RcSAG12H1–3. Group III includes HbSAG12H5–12 and RcSAG12H4–5. Group IV includes HbSAG12H13–17, JcSAG12H7–8 and RcSAG12H6–8, which can be further divided into two subgroups (IVa and IVb) (Fig 1).

Fig 1. Phylogenetic analysis of SAG12 orthologues.

Sequence alignment was performed using MUSCLE and the phylogenetic tree was constructed using bootstrap maximum likelihood tree (1000 replicates) method of the MEGA6 software. The distance scale denotes the number of amino acid substitutions per site, and the sequence identity of each orthologue to AtSAG12 is indicated in brackets. Species and Genbank accession numbers of reported SAG12 orthologues are as follows: BnSAG12-1 (AAD53011, Brassica napus), BnSAG12-2 (AAD53012, Brassica napus), Ghccyp (AAT34987, Gossypium hirsutum), NtCP1 (AAW78661, Nicotiana tabacum), NtSAG12 (ADV41672, Nicotiana tabacum) and IbSPG31 (AAK48495, Ipomoea batatas).

Obviously, a high number of genes were grouped in pairs, which is consistent with the homologous analysis above. HbSAG12H2 and 3 were characterized by same-direction neighbors on scaffold0696 and can be defined as tandem duplications. In contrast, HbSAG12H4 from scaffold1445, exhibits the high identity of 89.2% or 89.6% with HbSAG12H2 and 3, respectively, and is more likely to be resulted from the WGD. HbSAG12H5–9 from scaffold0048, HbSAG12H10–11 from scaffold2360 can be defined as tandem duplications, whereas their parental loci and HbSAG12H12 from scaffold0420 are more likely to be resulted from the WGD. Interesting, HbSAG12H13 and 14 (with 78.2% identity) from scaffold0247, HbSAG12H15 and 16 (with 80.0% identity) from scaffold0583, were found to be tandem distributed on same scaffolds. However, according to the phylogenetic analysis, they were grouped into IVa or IVb, respectively, just like tandem duplications JcSAG12H7 and 8. The cluster result was further supported by the homologous analysis: HbSAG12H13 and 14 exhibit relatively higher identity with HbSAG12H16 (90.3%) or HbSAG12H15 (93.0%), respectively; HbSAG12H14 and 15 also show the high identity of 92.0% or 95.4% with HbSAG12H17 from scaffold1086, respectively. Thereby, scaffold0247 and scaffold0583 are more likely to be resulted from the WGD, and the tandem duplications HbSAG12H13/14 or HbSAG12H15/16 are promised to be appeared before the divergence of rubber tree and physic nut. However, whether HbSAG12H17 is a proximal duplication of HbSAG12H15 still needs to be confirmed, since the 7,453 assembled scaffolds have not been anchored to the chromosomes yet [38].

Sequence feathers of HbSAG12Hs

Sequence analysis showed that the 17 deduced HbSAG12H proteins consist of 339 to 347 amino acids with the theoretical MW ranging from 37352.7 to 38578.8 Da, which is consistent with AtSAG12, JcSAG12Hs, RcSAG12Hs and other reported SAG12 orthologues. The pI value of HbSAG12Hs ranges from 4.65 to 9.10. Although about 47.06% out of 17 HbSAG12Hs were predicted to be basic as AtSAG12, BnSAG12-2 and NtSAG12, the remainings are acid as BnSAG12-1, IbSPG31, Ghccyp and NtCP1. All HbSAG12Hs were predicted to harbor the GRAVY value (from -0.485 to -0.248) of less than 0, indicating their hydrophilic feather. As shown in Fig 2, all HbSAG12Hs contain the conserved catalytic triad (Cys-His-Asn), and a conserved ERFNIN motif with the exception of the R/H mutation in HbSAG12H6. All HbSAG12Hs were also predicted to harbor a hydrophobic signal peptide at the N-terminus (Fig 2 and Table 1).

Fig 2. Alignment of precursor proteins of HbSAG12Hs.

Sequence alignment was performed using MUSCLE and the alignment was displayed using Boxshade. Black shading shows identical amino acids, whereas light gray shading shows similar amino acids. The numbers indicate the positions of the amino acids within individual proteins. The consensus ERFNIN motif is marked with black dots. The conserved catalytic triad (Cys, His and Asn) is marked with asterisks. The predicted signal peptide is boxed. The putative cleavage site to generate the mature enzyme is marked with a down arrow, which is predicted based on sequence alignment with IbSPG31.

To learn more about the diversity of motif compositions among different HbSAG12Hs, a phylogenetic tree from 17 HbSAG12Hs was constructed and the motifs in protein sequences were predicted using MEME (Fig 3). Among the 15 identified motifs, motifs 1–5, 7 and 8 are broadly distributed. Motif 1 includes the ERFNIN consensus sequence and is characterized as the Inhibitor_I29 (Pfam accession number PF08246) which provides the core structure of the autoinhibitory prodomain. Motif 2 includes the Cys active site. Motif 3 includes the Asn active site. Motif 4 includes the putative cleavage site to generate the mature enzyme. Motif 7 includes the His active site. Motif 14 identified in HbSAG12H1 and 5 as well as motif 6 found in other HbSAG12Hs includes the GCE/Q/G/K/N/DGG consensus sequence, where the Cys residue is involved in the formation of a disulphide bridge. Motifs 9, 10, 12 and 13 belong to the predicted signal peptide. In contrast, little is known for other motifs: motifs 5 and 8 were found in all HbSAG12Hs; motif 11 was found in most HbSAG12Hs excluding HbSAG12H8–11; motif 15 was limited to HbSAG12H1 and 2 (Fig 3 and Table 4).

Fig 3. Structural and phylogenetic analysis of HbSAG12H proteins.

The unrooted phylogenetic tree resulting from all the HbSAG12H proteins is shown on the left. The distribution of conserved motifs among the HbSAG12H proteins is shown on the right. Different motifs are represented by different color blocks as indicated at the bottom of the figure. The same color in different proteins indicates the same group or motif.

Table 4. Motif sequences of 17 HbSAG12H proteins identified by MEME tools.

Expression profiles of HbSAG12H genes

As shown in Fig 4, transcriptome profiling showed that most HbSAG12H genes were not or lowly expressed in examined tissues except for root and male flower. Among 10 genes detected in male flower, the transcripts of HbSAG12H11, HbSAG12H10, HbSAG12H9 and HbSAG12H1 were relatively abundant. The high abundance of AtSAG12 and four RcSAG12Hs (i.e. RcSAG12H5, RcSAG12H4, RcSAG12H1 and RcSAG12H3) in flower was also observed. In root, the more abundant transcripts include HbSAG12H4, HbSAG12H3, HbSAG12H17, HbSAG12H2, HbSAG12H15, HbSAG12H14, HbSAG12H16 and HbSAG12H6, and their cDNAs can indeed be amplified using roots as the template (data not shown). In addition, the abundant HbSAG12H7 seems to be a seed-specific gene since its transcript was not detected in other examined tissues.

Fig 4. Tissue-specific expression profiles of 17 HbSAG12H genes.

Color scale represents FPKM normalized log10 transformed counts where green indicates low expression and red indicates high expression.

Moreover, 17 primer pairs were also used to amplify HbSAG12H genes expressed in senescent leaves. Interesting, the only amplified gene was confirmed to be HbSAG12H1. Thereby, subsequent semi-quantitative RT-PCR analysis was focused on HbSAG12H1. Except for flower and seed, all profiled tissues as well as shoot apex, petiole, xylem and leaf of six more developmental stages, were examined. As shown in Fig 5, the expression of HbSAG12H1 was only found in senescent leaves, i.e., lowly in early-senescent leaves, moderately in mid-senescent leaves, and highly in ET-induced senescent leaves. The result is consistent with its transcriptome profiles in laticifers, barks and mature leaves. In contrast, no detected PCR product of this gene in roots is more likely to be due to its low expression in this specific tissue, which usually results in no visible signal by electrophoresis. And RNA-seq (RNA sequencing) has been proved to be the most sensitive technology to determine gene expression [60].


Gene duplication is a major mechanism for acquiring new genes, which may be resulted from single gene duplications such as local (tandem or proximal), dispersed and transposed duplications, or large-scale duplications such as WGDs and segmental duplications [61]. WGDs are widespread in plants. For example, one so-called γ whole-genome triplication event was shown to occur in all core eudicot plants including Arabidopsis, rubber tree, castor and physic nut [62]. Moreover, it is well established that Arabidopsis underwent two recent doubling events [63,64]. Nevertheless, Arabidopsis SAG12 was found to exist as a single copy gene [65]. The present study performed the first genome-wide identification of SAG12-like genes in rubber tree, and an unexpected high number of 17 orthologues were found. Thereby, to reveal their expansion and evolutionary pattern is particularly interesting.

In Euphorbiaceae, the genome sequences of castor and physic nut are also available. Unlike rubber tree, both castor and physic nut didn’t undergo any recent WGD [38,4345]. Thereby, the identification of SAG12 orthologues in these two plants may provide insights into the evolutionary pattern of HbSAG12H genes. The genome-wide survey indicated that both castor and physic nut contain 8 SAG12 orthologues, and the number is two-folds smaller than that in rubber tree. Without any exception, all SAG12H genes in rubber tree, castor and physic nut contain a single intron as observed in sweet potato [19], but not in Arabidopsis and rapeseed that harbor two introns [11,18].

The phylogenetic analysis was further performed to investigate the evolutionary relationships of HbSAG12Hs, which divided the tested SAG12 orthologues into four groups. Interesting, rubber tree harbors at least one orthologue in each group. In contrast, no orthologue can be found in Group I and III for castor or physic nut, respectively. This means that the diversification of SAG12 orthologues into four groups can date back to the ancestral Euphorbiaceae plant, and group-specific gene loss has occurred in both castor and physic nut. In fact, all group members can be found in poplar (Populus trichocarpa), and members of Group I-III in papaya (Carica papaya) [66,67]. Since papaya underwent no recent WGD and stands very close to Arabidopsis, the unique AtSAG12 is more likely to be resulted from massive gene loss and chromosomal rearrangement after WGDs [64].

According to the phylogenetic analysis, a high number of gene pairs were identified, i.e., 3 in castor, 3 in physic nut and 5 in rubber tree. The duplication mechanism was investigated based on their locations. In castor and physic nut, except for JcSAG12H1/2 that can be defined as proximal duplications, other gene pairs were all derived from tandem duplications, supporting its major role in the expansion of SAG12 orthologues. In contrast, a role of WGD is also involved in rubber tree: HbSAG12H2/3, HbSAG12H5–9, HbSAG12H10/11, HbSAG12H13/14 and HbSAG12H15/16 can be defined as tandem duplications, while HbSAG12H4 or HbSAG12H2/3, HbSAG12H12 or HbSAG12H10/11, HbSAG12H13/14 or HbSAG12H15/16 are more likely to be resulted from the WGD.

Like AtSAG12 and other reported papain-like cysteine proteases, sequence analysis indicated that HbSAG12Hs are synthesized as inactive prepropeptides, consisting of an N-terminal signal sequence followed by an autoinhibitory prodomain, and the mature enzyme (i.e. the Peptidase_C1 domain under the Pfam accession number of PF00112) at the C-terminus. And the mature, active enzyme is promised to be resulted from proteolysis cleaving off the pre and pro domains [68]. The putative cysteine protease activity of HbSAG12Hs is supported by the presence of the conserved catalytic triad Cys-His-Asn, which was shown to be essential for the cysteine protease activity [69,70]. Moreover, the expression of HbSAG12H genes were all supported by available RNA sequencing reads, suggesting that they have function in rubber tree. Nevertheless, transcriptome profiling indicated that the expression level of most HbSAG12H genes was considerably low in most examined tissues, and the result was further confirmed by the RT-PCR analysis. When using roots as the PCR template, 8 abundant genes can be successfully amplified, whereas no amplification can be observed for the less abundant HbSAG12H1 and other HbSAG12H genes (data not shown). In fact, HbSAG12H1 presents the only gene that can be amplified from senescent leaves (Fig 5). The senescence-specific expression of HbSAG12H1 in leaves was further confirmed by the semi-quantitative RT-PCR analysis. Among 6 stages of developmental leaves (i.e. bronze, color-change, pale-green, mature, early- and mid-senescent) examined, the expression of HbSAG12H1 was limited to senescent leaves and the transcript level gradually increased during leaf senescing. In addition, the transcripts of HbSAG12H1 increased about 10 folds in ET-induced senescent leaves than that in mid-senescent leaves (Fig 5). Since mid-senescent and ET-induced senescent leaves contain the similar chlorophyll content (i.e. 45–55% relative to mature leaves), the expression of HbSAG12H1 is more likely to be induced by ET, which is consistent with the presence of several putative ET-responsive elements in its 2000-bp promoter region (data not shown). Similar expression patterns were also reported for other SAG12 orthologues [1822]. For examples, among 5 tissues (i.e. leaf, flower, stem, root and tuber) tested, the transcripts of SPG31 were detected only in senescent leaves, and the transcript level can be highly induced after treatment of mature leaves with ET for 3 days [19]; both BnSAG12-1 and BnSAG12-2 transcripts were detected only in senescent cotyledons and leaves [18]. Although the mechanism of the transcript abundance of HbSAG12H genes in male flowers and roots is yet to be elucidated, group I and III members tend to be expressed in male flowers, whereas group II and IV members prefer to be expressed in roots.


In this study, survey of the rubber tree genome resulted in 17 SAG12 orthologues, and the gene number is considerably larger than 8 in both castor and physic nut, another two Euphorbiaceae plants. These orthologues can be divided into four groups based on the phylogenetic analysis. Genome-wide comparative analysis indicated that the diversification of SAG12 orthologues can date back to the ancestor of core eudicot plants: the unique AtSAG12 in Arabidopsis is resulted from gene loss; the duplicated SAG12 orthologues in castor and physic nut were mainly resulted from tandem duplications; whereas the duplicated HbSAG12H genes were derived from tandem duplications as well as the recent WGD. Furthermore, HbSAG12H1 exhibits a strictly senescence-associated expression pattern in rubber tree leaves, and can be used as a marker gene for the study of senescence mechanism in Hevea.

Supporting information

S1 File. PCR conditions used in this study.


S1 Table. PCR primers used for gene cloning in this study.



The authors appreciate those contributors who make the related genome and transcriptome data accessible in public databases. They also thank two anonymous reviewers for their helpful suggestions.

Author Contributions

  1. Conceptualization: ZZ.
  2. Data curation: ZZ.
  3. Formal analysis: ZZ JL.
  4. Funding acquisition: GX LY.
  5. Investigation: ZZ JL.
  6. Methodology: ZZ.
  7. Project administration: ZZ GX LY.
  8. Resources: ZZ GX LY.
  9. Software: ZZ.
  10. Supervision: ZZ GX LY.
  11. Validation: ZZ.
  12. Visualization: ZZ.
  13. Writing – original draft: ZZ.
  14. Writing – review & editing: GX LY.


  1. 1. Schippers JH. Transcriptional networks in leaf senescence. Curr Opin Plant Biol. 2015;27:77–83. pmid:26190740
  2. 2. Kim J, Woo HR, Nam HG. Toward systems understanding of leaf senescence: an integrated multi-omics perspective on leaf senescence research. Mol Plant. 2016;9(6):813–825. pmid:27174403
  3. 3. Guo Y, Gan S. Convergence and divergence in gene expression profiles induced by leaf senescence and 27 senescence-promoting hormonal, pathological and environmental stress treatments. Plant Cell Environ. 2012;35(3):644–655. pmid:21988545
  4. 4. Jibran R, Hunter D, Dijkwel P. Hormonal regulation of leaf senescence through integration of developmental and stress signals. Plant Mol Biol. 2013;82(6):547–561. pmid:23504405
  5. 5. Lim PO, Kim HJ, Nam HG. Leaf senescence. Annu Rev Plant Biol. 2007;58:115–136. pmid:17177638
  6. 6. Avila-Ospina L, Moison M, Yoshimoto K, Masclaux-Daubresse C. Autophagy, plant senescence, and nutrient recycling. J Exp Bot. 2014;65(14):3799–3811. pmid:24687977
  7. 7. Ougham H, Hörtensteiner S, Armstead I, Donnison I, King I, Thomas H, et al. The control of chlorophyll catabolism and the status of yellowing as a biomarker of leaf senescence. Plant Biol (Stuttg). 2008;10 Suppl 1:4–14.
  8. 8. Roberts IN, Caputo C, Criado MV, Funk C. Senescence-associated proteases in plants. Physiol Plant. 2012;145(1):130–139. pmid:22242903
  9. 9. Hensel LL, Grbić V, Baumgarten DA, Bleecker AB. Developmental and age-related processes that influence the longevity and senescence of photosynthetic tissues in Arabidopsis. Plant Cell. 1993;5(5):553–564. pmid:8518555
  10. 10. Lohmen KN, Gen S, John MC, Amasino RM. Molecular analysis of natural leaf senescence in Arabidopsis thaliana. Physiol Plant. 1994;92:322–328.
  11. 11. Noh YS, Amasino RM. Identification of a promoter region responsible for the senescence-specific expression of SAG12. Plant Mol Biol. 1999;41(2):181–194. pmid:10579486
  12. 12. Weaver LM, Gan S, Quirino B, Amasino RM. A comparison of the expression patterns of several senescence-associated genes in response to stress and hormone treatment. Plant Mol Biol. 1998;37(3):455–469. pmid:9617813
  13. 13. Otegui MS, Noh Y-S, Martinez DE, Petroff AGV, Staehelin LA, Amasino RM, et al. Senescence-associated vacuoles with intense proteolytic activity develop in leaves of Arabidopsis and soybean. Plant J. 2005;41(6):831–844. pmid:15743448
  14. 14. Miao Y, Laun T, Zimmermann P, Zentgraf U. Targets of the WRKY53 transcription factor and its role during leaf senescence in Arabidopsis. Plant Mol Biol. 2004;55:853–867. pmid:15604721
  15. 15. Jiang Y, Liang G, Yang S, Yu D. Arabidopsis WRKY57 functions as a node of convergence for jasmonic acid- and auxin mediated signaling in jasmonic acid- induced leaf senescence. Plant Cell. 2014;26:230–245. pmid:24424094
  16. 16. Gan S, Amasino RM. Inhibition of leaf senescence by autoregulated production of cytokinin. Science. 1995;270:1986–1988. pmid:8592746
  17. 17. Guo Y, Gan SS. Translational researches on leaf senescence for enhancing plant productivity and quality. J Exp Bot. 2014;65(14):3901–3913. pmid:24935620
  18. 18. Noh YS, Amasino RM. Regulation of developmental senescence is conserved between Arabidopsis and Brassica napus. Plant Mol Biol. 1999;41(2):195–206. pmid:10579487
  19. 19. Chen GH, Huang LT, Yap MN, Lee RH, Huang YJ, Cheng MC, et al. Molecular characterization of a senescence-associated gene encoding cysteine proteinase and its gene expression during leaf senescence in sweet potato. Plant Cell Physiol. 2002;43:984–991. pmid:12354916
  20. 20. Shen F, Yu S, Han X, Fan S. Cloning and characterization of a gene encoding cysteine proteases from senescent leaves of Gossypium hirsutum. Chin Sci Bull. 2004;49(24):2601–2607.
  21. 21. Beyene G, Foyer CH, Kunert KJ. Two new cysteine proteinases with specific expression patterns in mature and senescent tobacco (Nicotiana tabacum L.) leaves. J Exp Bot. 2006;57(6):1431–1443. pmid:16551685
  22. 22. Carrión CA, Costa ML, Martínez DE, Mohr C, Humbeck K, Guiamet JJ. In vivo inhibition of cysteine proteases provides evidence for the involvement of 'senescence-associated vacuoles' in chloroplast protein degradation during dark-induced senescence of tobacco leaves. J Exp Bot. 2013;64(16):4967–4980. pmid:24106291
  23. 23. Leitch AR, Lim KY, Leitch IJ, O’Neill M, Chye ML, Low FC. Molecular cytogenetic studies in rubber, Hevea brasiliensis Muell. Arg. (Euphorbiaceae). Genome. 1998;41:464–467.
  24. 24. Zou Z, Yang LF, Wang ZH, Yuan K. Biosynthesis and regulation of natural rubber in Hevea. Plant Physiol J. 2009;45:1231–1238.
  25. 25. Zou Z, Yang LF, Wang ZH, Yuan K. Advances and prospects in genetic modification of Hevea brasiliensis. J Chin Biotechnol. 2010;30:85–92.
  26. 26. Hao BZ, Wu JL. Laticifer differentiation in Hevea brasiliensis: induction by exogenous jasmonic acid and linolenic acid. Ann Bot. 2000;85:37–43.
  27. 27. Yeang HY. The kinetics of latex flow from the rubber tree in relation to latex vessels plugging and turgor pressure. J Rubber Res. 2005;8:160–181.
  28. 28. Zou Z, Gong J, An F, Xie GS, Wang JK, Mo YY, et al. Genome-wide identification of rubber tree (Hevea brasiliensis Muell. Arg.) aquaporin genes and their response to ethephon stimulation in the laticifer, a rubber-producing tissue. BMC genomics. 2015;16:1001. pmid:26606923
  29. 29. Zhu J, Zhang Z. Ethylene stimulation of latex production in Hevea brasiliensis. Plant Signal Behav. 2009;4(11):1072–1074. pmid:20009550
  30. 30. Zou Z, Yang LF, Wang ZH, Yuan K. Strategies on prevention and treatment of tapping panel dryness syndrome of rubber (Hevea brasiliensis Muell. Arg.). J Chin Biotechnol. 2012;9:8–15.
  31. 31. Chen SC, Peng SQ, Huang GX, Wu K, Fu X, Chen Z. Association of decreased expression of a Myb transcription factor with the TPD (tapping panel dryness) syndrome in Hevea brasiliensis. Plant Mol Biol. 2003;51(1):51–58. pmid:12602890
  32. 32. Venkatachalam P, Thulaseedharan A, Raghothama K. Identification of expression profiles of tapping panel dryness (TPD) associated genes from the latex of rubber tree (Hevea brasiliensis Muell. Arg.). Planta. 2007;226(2):499–515. pmid:17356851
  33. 33. Peng SQ, Wu KX, Huang GX, Chen SC. HbMyb1, a Myb transcription factor from Hevea brasiliensis, suppresses stress induced cell death in transgenic tobacco. Plant Physiol Biochem. 2011;49(12):1429–1435. pmid:22078380
  34. 34. Li D, Deng Z, Chen C, Xia Z, Wu M, He P, et al. Identification and characterization of genes associated with tapping panel dryness from Hevea brasiliensis latex using suppression subtractive hybridization. BMC Plant Biol. 2010;10:140. pmid:20618931
  35. 35. Li D, Wang X, Deng Z, Liu H, Yang H, He G. Transcriptome analyses reveal molecular mechanism underlying tapping panel dryness of rubber tree (Hevea brasiliensis). Sci Rep. 2016;6:23540. pmid:27005401
  36. 36. Liu JP, Xia ZQ, Tian XY, Li YJ. Transcriptome sequencing and analysis of rubber tree (Hevea brasiliensis Muell.) to discover putative genes associated with tapping panel dryness (TPD). BMC Genomics. 2015;16:398. pmid:25994052
  37. 37. Putranto RA, Herlinawati E, Rio M, Leclercq J, Piyatrakul P, Gohet E, et al. Involvement of ethylene in the latex metabolism and tapping panel dryness of Hevea brasiliensis. Int J Mol Sci. 2015;16(8):17885–17908. pmid:26247941
  38. 38. Tang C, Yang M, Fang Y, Luo Y, Gao S, Xiao X, et al. The rubber tree genome reveals new insights into rubber production and species adaptation. Nat Plants. 2016;2(6):16073. pmid:27255837
  39. 39. Lamesch P, Berardini TZ, Li D, Swarbreck D, Wilks C, Sasidharan R, et al. The Arabidopsis Information Resource (TAIR): improved gene annotation and new tools. Nucleic Acids Res. 2012;40(Database issue):D1202–1210. pmid:22140109
  40. 40. Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997;25(17):3389–3402. pmid:9254694
  41. 41. Lomsadze A, Ter-Hovhannisyan V, Chernoff YO, Borodovsky M. Gene identification in novel eukaryotic genomes by self-training algorithm. Nucleic Acids Res. 2005;33:6494–6506. pmid:16314312
  42. 42. Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012;9(4):357–359. pmid:22388286
  43. 43. Zou Z, Yang L, Gong J, Mo Y, Wang J, Cao J, et al. Genome-wide identification of Jatropha curcas aquaporin genes and the comparative analysis provides insights into the gene family expansion and evolution in Hevea brasiliensis. Front Plant Sci. 2016;7:395. pmid:27066041
  44. 44. Chan AP, Crabtree J, Zhao Q, Lorenzi H, Orvis J, Puiu D, et al. Draft genome sequence of the oilseed species Ricinus communis. Nat Biotechnol. 2010;28(9):951–956. pmid:20729833
  45. 45. Wu P, Zhou C, Cheng S, Wu Z, Lu W, Han J, et al. Integrated genome sequence and linkage map of physic nut (Jatropha curcas L.), a biodiesel plant. Plant J. 2015;81(5): 810–821. pmid:25603894
  46. 46. Goodstein DM, Shu S, Howson R, Neupane R, Hayes RD, Fazo J, et al. Phytozome: a comparative platform for green plant genomics. Nucleic Acids Res. 2012;40(Database issue):D1178–1186. pmid:22110026
  47. 47. Edgar RC. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res. 2004;32(5):1792–1797. pmid:15034147
  48. 48. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S. MEGA6: Molecular evolutionary genetics analysis version 6.0. Mol Biol Evol. 2013;30(12):2725–2729. pmid:24132122
  49. 49. Bannai H, Tamada Y, Maruyama O, Nakai K, Miyano S. Extensive feature detection of N-terminal protein sorting signals. Bioinformatics. 2002;18(2):298–305. pmid:11847077
  50. 50. Petersen TN, Brunak S, von Heijne G, Nielsen H. SignalP 4.0: discriminating signal peptides from transmembrane regions. Nat Methods. 2011;8(10):785–786. pmid:21959131
  51. 51. Bailey TL, Boden M, Buske FA, Frith M, Grant CE, Clementi L, et al. MEME SUITE: tools for motif discovery and searching. Nucleic Acids Res. 2009;37(Web Server issue):W202–208. pmid:19458158
  52. 52. Hua YW, Huang TD, Huang HS. Micropropagation of self-rooting juvenile clones by secondary somatic embryogenesis in Hevea brasiliensis. Plant Breeding. 2010;129:202–207.
  53. 53. Tang C, Huang D, Yang J, Liu S, Sakr S, Li H, et al. The sucrose transporter HbSUT3 plays an active role in sucrose loading to laticifer and rubber productivity in exploited trees of Hevea brasiliensis (para rubber tree). Plant Cell Environ. 2010;33(10):1708–1720. pmid:20492551
  54. 54. Mortazavi A, Williams BA, McCue K, Schaeffer L, Wold B. Mapping and quantifying mammalian transcriptomes by RNA-seq. Nat Methods. 2008;5(7):621–628. pmid:18516045
  55. 55. Triwitayakorn K, Chatkulkawin P, Kanjanawattanawong S, Sraphet S, Yoocha T, Sangsrakru D, et al. Transcriptome sequencing of Hevea brasiliensis for development of microsatellite markers and construction of a genetic linkage map. DNA Res. 2011;18(6):471–482. pmid:22086998
  56. 56. Chow KS, Mat-Isa MN, Bahari A, Ghazali AK, Alias H, Mohd-Zainuddin Z, et al. Metabolic routes affecting rubber biosynthesis in Hevea brasiliensis latex. J Exp Bot. 2012;63(5):1863–1871. pmid:22162870
  57. 57. Chow KS, Ghazali AK, Hoh CC, Mohd-Zainuddin Z. RNA sequencing read depth requirement for optimal transcriptome coverage in Hevea brasiliensis. BMC Res Notes. 2014;7:69. pmid:24484543
  58. 58. Duan C, Argout X, Gébelin V, Summo M, Dufayard JF, Leclercq J, et al. Identification of the Hevea brasiliensis AP2/ERF superfamily by RNA sequencing. BMC Genomics. 2013;14:30. pmid:23324139
  59. 59. Mantello CC, Cardoso-Silva CB, da Silva CC, de Souza LM, Scaloppi Junior EJ, de Souza Gonçalves P, et al. De novo assembly and transcriptome analysis of the rubber tree (Hevea brasiliensis) and SNP markers development for rubber biosynthesis pathways. PLoS One. 2014;9(7):e102665. pmid:25048025
  60. 60. Wang Z, Gerstein M, Snyder M. RNA-Seq: a revolutionary tool for transcriptomics. Nat Rev Genet. 2009;10(1):57–63. pmid:19015660
  61. 61. Wang Y, Wang X, Paterson AH. Genome and gene duplications and gene expression divergence: a view from plants. Ann N Y Acad Sci. 2012;1256:1–14. pmid:22257007
  62. 62. Jiao Y, Leebens-Mack J, Ayyampalayam S, Bowers JE, McKain MR, McNeal J, et al. A genome triplication associated with early diversification of the core eudicots. Genome Biol. 2012;13(1):R3. pmid:22280555
  63. 63. Bowers JE, Chapman BA, Rong J, Paterson AH. Unravelling angiosperm genome evolution by phylogenetic analysis of chromosomal duplication events. Nature. 2003;422(6930):433–438. pmid:12660784
  64. 64. Blanc G, Wolfe KH. Functional divergence of duplicated genes formed by polyploidy during Arabidopsis evolution. Plant Cell. 2004;16(7):1679–1691. pmid:15208398
  65. 65. Arabidopsis Genome Initiative. Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature. 2000;408(6814):796–815. pmid:11130711
  66. 66. Tuskan GA, Difazio S, Jansson S, Bohlmann J, Grigoriev I, Hellsten U, et al. The genome of black cottonwood, Populus trichocarpa (Torr. & Gray). Science. 2006;313(5793):1596–1604. pmid:16973872
  67. 67. Ming R, Hou S, Feng Y, Yu Q, Dionne-Laporte A, Saw JH, et al. The draft genome of the transgenic tropical fruit tree papaya (Carica papaya Linnaeus). Nature. 2008;452(7190):991–996. pmid:18432245
  68. 68. Wiederanders B. Structure-function relationships in class CA1 cysteine peptidase propeptides. Acta Biochim Pol. 2003;50(3):691–713. pmid:14515150
  69. 69. Thakurta PG, Biswas S, Chakrabarti C, Sundd M, Jagannadham MV, Dattagupta JK. Structural basis of the unusual stability and substrate specificity of ervatamin C, a plant cysteine protease from Ervatamia coronaria. Biochemistry. 2004;43(6):1532–1540. pmid:14769029
  70. 70. Than ME, Helm M, Simpson DJ, Lottspeich F, Huber R, Gietl C. The 2.0 A crystal structure and substrate specificity of the KDEL-tailed cysteine endopeptidase functioning in programmed cell death of Ricinus communis endosperm. J Mol Biol.2004;336(5):1103–1116. pmid:15037072