Genome-wide identification of ZjMAPKKKs
A total of 56 ZjMAPKKKs were defined. All of these kinases have the key S-TKc (serine/threonine protein kinase) domain and other conserved protein kinase domains (Additional file. 2). To clearly understand and discriminate between the MAPKKK genes, the locus of ZjMAPKKKs was designated according to the nomenclature suggestions for Arabidopsis, where Zj refers to Ziziphus jujuba, and the series numbers ZjMAPKKK1-56 are coded in terms of their chromosome locations (Table 1). The ZjMAPKKKs are distributed over all 12 pseudochromosomes, except for ZjMAPKKK44-56, which could not be matched to a corresponding chromosome.
Specific information for each CDS and amino acid sequence of the ZjMAPKKKs is listed in Additional files 1 and 7. Based on the specific conserved signature motif, all the ZjMAPKKKs could be grouped into one of the two subfamilies Raf and MEKK. No ZIK subfamily members were identified. As shown in Table 1, the length of the CDS sequences ranged from 762 bp (ZjMAPKKK36) to 4455 bp (ZjMAPKKK7) with an average length of 1804 bp. The amino acid sequence length of ZjMAPKKKs varied from 253 (ZjMAPKKK36) to 1,484 (ZjMAPKKK7) amino acids (aa); the average length was 600 aa. The predicted molecular weight (Mw) of these proteins ranged from 28.29 (ZjMAPKKK36) to 160.66 (ZjMAPKKK7), and the theoretical isoelectric points (pI) ranged from 4.78 to 9.34.
Phylogenetic analyses of ZjMAPKKK genes
To assess the phylogenetic relationships between Chinese jujube and Arobidopsis, a phylogenetic tree was constructed with all 136 protein sequences (56 ZjMAPKKKs and 80 AtMAPKKKs). As illustrated in Fig. 1, the members of AtMAPKKKs could be clustered into three categories, Raf, ZIK and MEKK, indicating that the method used to build the phylogenetic tree was reliable. However, the 56 members of ZjMAPKKKs could be clustered into only two subfamilies, Raf and MEKK. In addition, the largest Raf subfamily consisted of 41 members, with the remaining 15 members of ZjMAPKKKs belonging to the MEKK subfamily. None could be ascribed to the ZIK subfamily. Moreover, some ZjMAPKKKs located on the same chromosome showed little divergence but clustered into the same group. Examples are ZjMAPKKK36, -37 and -40; ZjMAPKKK15 and -16; and ZjMAPKKK38 and -39. These results indicate that some duplication of ZjMAPKKKs occurred during the evolution of jujube.
Conserved domains and gene structure analyses of ZjMAPKKKs
Within the analysis of MEME software, five main conserved motifs were identified in all 56 ZjMAPKKKs (Fig. 2). Motifs 1, 3 and 4 were found in all ZjMAPKKKs, while the other two motifs were observed in all Raf subfamily members. The MEKK subfamily members fell into two groups: one contained motifs 1-4, including ZjMAPKKK21, -56, -6, -31, -10 and -25. The remaining members contained only motifs 1, 3 and 4. These results illustrate that ZjMAPKKKs share the same conserved motifs, which further indicates that the protein structures for each subfamily are highly conserved.
For the analyses of the exon/intron contents, the differences among ZjMAPKKKs were significant. As shown in Fig. 3 and in Additional file 8, the number of exons in ZjMAPKKKs ranged from 1 (ZjMAPKKK9, -12, -29, -35, -36, -44 and -54) to 19 (ZjMAPKKK42). Interestingly, the members of ZjMAPKKKs containing only 1 exon all belong to the MEKK subfamily (47%). The highest number of exons in this subfamily was 17 (ZjMAPKKK25 and -10), and the average number was 5.6. This result demonstrates that in this subfamily, significant loss and gains of exons took place during evolution. For the Raf subfamily, the number of exons varied from 2 (ZjMAPKKK16 and -28) to 19 (ZjMAPKKK42) with an average number of 9.56. Although there was significant variation in the number of exons in the Raf and MEKK subfamilies, some exon structure patterns were clearly conserved in close paralogues. For instance, ZjMAPKKK24 and -49 have 12 exons, while ZjMAPKKK37 and -40 have 2 exons, and they are all closely clustered in the same phylogenetic tree. Collectively, the evolutionarily different organisations of the ZjMAPKKK gene structures between the Raf and MEKK subfamilies indicate that tandem and segmental duplication events may have occurred in ancient times and that diverse exon structures may function differently in the jujube genome.
Furthermore, with the multiple protein alignment of ZjMAPKKKs, the Raf-specific signature motif GTXX(W/Y)MAPE was found in the Raf subfamily, and the kinase domain was located at the N terminal or C terminal. In contrast, the less conserved MEKK-specific signature motif G(T/S)PX(W/F)MAPEV was observed in the MEKK subfamily, while the kinase domain was located at three positions: the N- or C-terminal or the central part of the proteins (Fig. 4). The features of the signature motifs of ZjMAPKKKs are consistent with other orthologues in other plant species, where they fulfil important roles in a diversity of signal transduction processes.
Synteny analysis of ZjMAPKKK genes
Tandem duplication events were first analysed according to the principle that two or more genes can be located on a chromosomal region within 200 kb [34] of one another. As shown in Fig. 5, one pair of ZjMAPKKKs (15/16) was the only tandem duplication event on LG5. In addition, 13 segmental duplication events with 22 ZjMAPKKKs were also identified. These results indicate that some ZjMAPKKKs may have been generated by gene duplication and that segmental duplication events were probably a major driving force in ZjMAPKKK evolution.
Phytoplasma detection in different tissues infected by phytoplasma
To characterize the functions of ZjMAPKKKs involved in phytoplasma infection, the expression levels of individual ZjMAPKKKs were detected by qPCR in two kinds of infected plant material. The first infected plant material was from diseased plants in the field (in vivo). This material showed three levels of symptoms: (a) witches’ broom leaves, (b) phyllody leaves and (c) apparently normal leaves (but from diseased plants). The other infected plant material was from sterile (in vitro) cultured tissues of JWB plantlets. The phytoplasma concentrations in the in vivo material with three levels of symptoms were measured by Xue et al. (2018) [22]. The phytoplasma determination in the in vitro tissues shows fluorescent spots forming a large circle in the phloem of the petiole (Additional file 5). These results confirm the subsequent tests on ZjMAPKKK function in response to phytoplasma infection.
Expression analysis of ZjMAPKKKs in witches’ broom leaves
In Additional file 9 and Fig. 6 (A), the heat map shows the expression levels of ZjMAPKKKs with significantly different patterns in the witches’ broom leaves from June to September. There were 42 candidates expressing at a detectable level, but the expression levels of the other 14 ZjMAPKKKs were either not expressed or were expressed at levels below our detection threshold. The ZjMAPKKKs genes with too low (or nonexistent) expression were rejected as candidates for further calculation and analysis. Among these genes, the most significant transcript induction took place in the early stage (June or July) when the concentration of witches’ broom began to increase. For example, ZjMAPKKK13, -14, -15, -23, -34, -42, -44, -47 and -56 were significantly induced in June or July, but induction later decreased from August to September, as shown in Fig. 6 (B). However, ZjMAPKKK3, -43 and -50 were downregulated from June to September. The expression levels of two ZjMAPKKK members (ZjMAPKKK26 and -45) remained high from June to September. This finding may indicate that these are key MAPKKKs in response to phytoplasma infection. However, the clustering of ZjMAPKKK expression profiles was not aligned with gene similarities, illustrating that gene function may not rely on gene structure.
Expression analysis of ZjMAPKKKs in phyllody leaves
The transcript abundance of ZjMAPKKKs was also investigated in phyllody leaves. The heat map of the expressing ZjMAPKKKs is shown in Fig. 7 (A). Several of the ZjMAPKKKs were highly expressed in June or July, but their expression levels then decreased from August to September. However, most ZjMAPKKKs showed no significant changes in expression level. Expression details for all ZjMAPKKKs are shown in Fig. 7 (B). ZjMAPKKK10, -14, -15 -34, -44 and -56 were all significantly upregulated in the early stage (June or July). However, ten of the ZjMAPKKKs (ZjMAPKKK3, -16, -18, -41, -43, -50, -51, -52, -53 and -55) were significantly downregulated. As in the witches’ broom leaves, in the phyllody leaves, ZjMAPKKK26 and -45 were highly upregulated from June to September.
Expression analysis of ZjMAPKKKs in apparently normal leaves
The apparently normal but asymptomatic infected leaves were used to test which ZjMAPKKKs play a role in the phytoplasma infection response. Interestingly, the heat map figure shows different expression patterns for ZjMAPKKKs in these leaves (Fig. 8B). A few genes were highly upregulated, but most showed downregulation. For example, ZjMAPKKK1, -3, -7, -16, -17, -18, -19, -41, -43, -50, and -52 were downregulated from June to September, while ZjMAPKKK28, -34 and -47 were significantly upregulated in June or July, while ZjMAPKKK27 and -54 were upregulated from August or September. However, ZjMAPKKK26 and -45 showed the same pattern of high expression in the asymptomatic infected leaves from June to September (Fig. 8 A).
To summarise, in the phytoplasma-infected tissues of the three symptomatic severities (apparently normal, phyllody and witches’ broom) and in the four months (June through September), ZjMAPKKK26 was significantly upregulated, and ZjMAPKKK45 was also highly induced. As the infection developed, the visible disease symptoms increased, becoming gradually more severe, from apparently normal leaves to phyllody leaves to witches’ broom leaves [21]. This progression occurred even though the concentration of JWB decreased gradually from August through September. The expression of ZjMAPKKK26 increased approximately six-fold in the phyllody leaves in June but not at the other two symptomatic levels. Then, as the infection developed, ZjMAPKKK26 was upregulated approximately three-fold in the witches’ broom leaves in July but downregulated in the phyllody. Meanwhile, in June, the induction of ZjMAPKKK45 increased approximately three-fold in the apparently normal (but phytoplasma-infected) leaves and approximately six-fold in the phyllody leaves. Then, during July, August and September, it was downregulated, but induction in the witches’ broom leaves remained approximately constant (Fig. 6, 7 and 8). These results show that ZjMAPKKK26 responds quickly in phyllody leaves and is highly induced in witch broom leaves, while ZjMAPKKK45 responds more rapidly than ZjMAPKKK26, as indicated by its high expression in apparently normal leaves in June. In contrast to ZjMAPKKK26 and ZjMAPKKK45, which were significantly upregulated, ZjMAPKKK3, ZjMAPKKK43 and ZjMAPKKK50 were significantly downregulated.
Expression analysis of the ZjMAPKKKs in sterile cultured JWB plantlets
In addition to the ZjMAPKKK expression profiles in the in vivo field tissues, we also examined expression levels in the in vitro cultured JWB plantlets, with uninfected plantlets being used as a control. As shown in Fig. 9, the in vitro ZjMAPKKK expression profiles differed significantly from the in vivo profiles. Only four of the ZjMAPKKKs were significantly induced in the diseased plants - ZjMAPKKK4, -10, -25 and -44. ZjMAPKKK6, -7, -17, -18, -30, -34, -35, -37, -40, -41, -43, -46, -52 and -53 were significantly downregulated. The other ZjMAPKKKs showed no significant change.