The ‘Candidatus Phytoplasma ziziphi’ effectors SJP1/2 negatively control leaf size by stabilizing the transcription factor ZjTCP2 in jujube

Abstract

Phytoplasmas manipulate host plant development to benefit insect vector colonization and their own invasion. However, the virulence factors and mechanisms underlying small-leaf formation caused by jujube witches’ broom (JWB) phytoplasmas remain largely unknown. Here, effectors SJP1 and SJP2 from JWB phytoplasmas were identified to induce small-leaf formation in jujube (Ziziphus jujuba). In vivo interaction and expression assays showed that SJP1 and SJP2 interacted with and stabilized the transcription factor ZjTCP2. Overexpression of SJP1 and SJP2 in jujube induced ZjTCP2 accumulation. In addition, the abundance of miRNA319f_1 was significantly reduced in leaves of SJP1 and SJP2 transgenic jujube plants and showed the opposite pattern to the expression of its target, ZjTCP2, which was consistent with the pattern in diseased leaves. Overexpression of ZjTCP2 in Arabidopsis promoted ectopic leaves arising from the adaxial side of cotyledons and reduced leaf size. Constitutive expression of the miRNA319f_1 precursor in the 35S::ZjTCP2 background reduced the abundance of ZjTCP2 mRNA and reversed the cotyledon and leaf defects in Arabidopsis. Therefore, these observations suggest that effectors SJP1 and SJP2 induced small-leaf formation, at least partly, by interacting with and activating ZjTCP2 expression both at the transcriptional and the protein level, providing new insights into small-leaf formation caused by phytoplasmas in woody plants.

Effector, miRNA319, phytoplasmas, small leaves, Ziziphus jujuba, ZjTCP2

Introduction

Jujube (Ziziphus jujuba Mill.) is an economically and ecologically important perennial fruit tree in the family Rhamnaceae. The primary centre of jujube domestication is the Shanxi–Shaanxi area of China, with more than 7000 years of cultivation (Liu et al., 2020; Guo et al., 2021). Long-term selection by natural forces or human behaviour drove the selection of many horticulturally impactful traits that are specific to jujube (Guo et al., 2021; Zhang et al., 2022). Hundreds of primary and specialized bioactive metabolites in jujube contribute to its edible and medicinal properties (Hua et al., 2022; Zhang et al., 2022; Zhang et al., 2023). An arbuscular mycorrhizal symbiosis enhances the ability of jujube to adapt to salt stress (Ma et al., 2022). However, the majority of cultivated jujubes are susceptible to jujube witches’ broom (JWB) disease (Zhao et al., 2019), which is caused by ‘Candidatus Phytoplasma ziziphi’ (Jung et al., 2003). JWB disease is prevalent worldwide in jujube orchards and causes extremely destructive losses for the industry (Song et al., 2019; Zhao et al., 2019). The diseased plants show excessive proliferation of shoots and branches (witches’ broom) (Zhou et al., 2021; Ma et al., 2023), floral organs that turn into leaf-like structures (phyllody) (Ma et al., 2020; Deng et al., 2021), and small leaves that are not shed in winter, leading to the death of the tree (Zhao et al., 2019). To date, the pathogenesis of JWB phytoplasma-induced small leaves is still not fully understood in jujube.

The JWB phytoplasma is one of four members in the elm yellows group (16SrV) (Wei and Zhao, 2022) and has a reduced genome of approximately 750 kb (J. Wang et al., 2018). A set of genes involved in some essential metabolic pathways are completely missing in JWB phytoplasmas (J. Wang et al., 2018), making them highly dependent on jujube for their nutrition and difficult to culture in vitro. JWB phytoplasmas are restricted to vascular phloem sieve cells (Zhao et al., 2019) and are dynamically distributed in roots, trunks, developing branches, and bearing shoots during year-round growth (Jin et al., 2006). JWB phytoplasmas stimulate the excessive growth of vegetative tissues with small leaves and reverse flower development to prolong the lifespan in perennial woody jujube plants, which is beneficial to the phytoplasma for insect vector colonization and phytoplasma invasion (Sugio and Hogenhout, 2012). In addition to the typical witches’ broom and phyllody symptoms (Ermacora and Osler, 2019), small leaves occur in many host plants (Rao, 2021; Marcone et al., 2023), such as juniper (Juniperus occidentalis) (Davis et al., 2010), sugar beet (Beta vulgaris) (Thilagavathi et al., 2011), pomegranate (Punica granatum) (Abu Alloush et al., 2023), brinjal (Solanum melongena) (Snehi et al., 2021), and geranium (Pelargonium hortorum) (Amirmijani et al., 2020). These plants are infected by different phytoplasmas but show the same symptom of small leaves, indicating that the phytoplasmas have evolved conserved mechanisms to modulate leaf development.

The release of 47 phytoplasma genomes has provided a powerful tool to comprehensively understand the interaction of phytoplasmas and their host plants (Wei and Zhao, 2022). Phytoplasmas secrete effectors (such as TENGU from the onion yellows mild (OYM) phytoplasma strain and SAP05, SAP06, SAP11, and SAP54 from the aster yellows witches’ broom (AYWB) phytoplasma strain) that modulate host plant developmental processes, including architecture, leaf morphology, flowering time, sterility, and lifespan (Hoshi et al., 2009; Sugio et al., 2011; Maclean et al., 2014; Chang et al., 2018; Pecher et al., 2019; Huang et al., 2021; Correa Marrero et al., 2024). One or two homologues of SAP effectors exist in different phytoplasma genomes, while these homologues are completely missing in some other phytoplasmas (Huang et al., 2021). Among the well-identified effectors, SAP11 and its homologues are involved in regulating stem proliferation and leaf morphology (Sugio et al., 2011; Chang et al., 2018). SAP11_AYWB induces more severely crinkled leaves than its homologues (Chang et al., 2018) and AYWB phytoplasma-infected Arabidopsis leaves (Sugio et al., 2011).

Class II CINCINNATA (CIN)–TEOSINTE-LIKE1, CYCLOIDEA, and PROLIFERATING CELL FACTOR (TCP) transcription factors are plant-specific growth regulators that regulate leaf development (Wang et al., 2021) and are targeted by SAP11 effectors (Sugio et al., 2011; Chang et al., 2018; Pecher et al., 2019). Loss of function of AtTCP2, 3, 4, 10, and 24 in Arabidopsis results in large and crinkled leaves (Karidas et al., 2015; Koyama et al., 2017; Challa et al., 2019). SAP11_AYWB destabilizes most of the CIN-TCPs, while its homologues have little impact on the abundance of these proteins, such as SAP11_PnWB from the peanut witches’ broom (PnWB) phytoplasma, SAP11_OYM from the OY-M phytoplasma, and SAP11_MBSP from the maize bushy stunt phytoplasma (MBSP) (Tan et al., 2016; Chang et al., 2018; Pecher et al., 2019). A recent report shows that SAP11_PnWB and SAP11_OYM bind class I TCPs but not CIN-TCPs (Correa Marrero et al., 2024). It is still unknown how CIN-TCP transcription factors are destabilized by SAP11_AYWB. All these findings indicate that diversity in the number, virulence, and molecular mechanism of effectors contributes to species-specific symptoms in host plants.

Phytoplasma-induced small leaves have been recorded in different plant species, including industrial crops (Thilagavathi et al., 2011), fruit crops (Kirdat et al., 2019; Abu Alloush et al., 2023), vegetables (Kumari et al., 2019; Snehi et al., 2021), and ornamentals (Panda et al., 2019; Amirmijani et al., 2020). The virulence factors and their molecular mechanisms are still mysterious. In our previous studies, 43 candidate secreted JWB phytoplasma proteins (SJPs) were identified from the JWB phytoplasma genome (Deng et al., 2021). Two SAP11 homologues, SJP1 and SJP2, are involved in witches’ broom and small leaves in Nicotiana benthamiana (Zhou et al., 2021). Here, SJP1 and SJP2 were identified as the key regulators that negatively control leaf size in jujube, at least partly, by stabilizing ZjTCP2 in leaves at the post-transcriptional and protein levels. Our results provide new insights into the molecular mechanisms of small-leaf formation caused by JWB phytoplasmas in perennial woody plants.

Materials and methods

Plant materials

From late May to mid-June during the year following infection, the morphology of mature leaves in the healthy and JWB-infected jujube plants was investigated for two consecutive years. Leaves and three types of lateral main buds from healthy and JWB-infected jujube plants (Zhou et al., 2021) were used to investigate the expression of miRNAs and ZjTCP2. Samples of the healthy and JWB-infected jujube plants were confirmed by amplification of JWB phytoplasma 16S rDNA as previously described (Ma et al., 2020). Transgenic jujube calli and tobacco carrying 35S::GFP, 35S::SJP1-GFP, and 35S::SJP2-GFP have been described previously (Zhou et al., 2021). Seeds of tcp2 (SALK_144731C) were obtained from AraShare (https://www.arashare.cn/index/).

Bimolecular fluorescence complimentation

For a bimolecular fluorescence complementation (BiFC) assay, ZjTCP2 coding sequences were cloned into the pSPYNE vector to generate NE-ZjTCP2. The ∆SP (lacking the signal peptide) and ∆N (lacking the signal peptide and the entire N-terminal coiled coil) fragments of SJP1 and SJP2 were cloned into the pSPYCE vector to generate CE-SJP1^∆SP, CE-SJP1^∆N, CE-SJP2^∆SP, and CE-SJP2^∆N. Arabidopsis mesophyll protoplasts were co-transformed by the polyethylene glycol (PEG)–Ca²⁺-mediated transformation system (Ma et al., 2018) and imaged by laser scanning confocal microscopy (Olympus FluoView FV1000 system, Japan) (Zhou et al., 2021). The plasmid combination pSPYNE-ZjTCP2/pSPYCE was used as the negative control. All the specific primers for vector construction are listed in Supplementary Table S1.

Co-immunoprecipitation assays

The in vivo interactions between effectors SJP1/2 and ZjTCP2 were investigated by co-immunoprecipitation (Co-IP) assays. For the Co-IP assay, total proteins from leaves of the 35S::SJP1-GFP and 35S::SJP2-GFP transgenic jujube plants were extracted and suspended in IP buffer as described previously (Yoon and Kieber, 2013). The wild-type (WT) jujube plant and the 35S::GFP jujube calli were used as negative controls. Immunoprecipitation experiments were performed with protein A/G sepharose beads (7Sea Pharmatech Co., Ltd, Shanghai, China) as described previously (Yao et al., 2017). Briefly, green fluorescent protein (GFP)-fused SJP1 and SJP2 were immunoprecipitated using an anti-GFP antibody (TransGen Biotech, Beijing, China) (1:200), and the co-immunoprecipitated protein was then detected using an anti-ZjTCP2 antibody (Vapol Biosciences Co., Ltd, Wuhan, China) (1:2000).

Subcellular localization

The Rhizobium radiobacter strain GV3101 containing ZjTCP2–mCherry was infiltrated into leaves of the SJP1 and SJP2 transgenic tobacco lines (Zhou et al., 2021), and tobacco leaves were incubated at room temperature in the dark for 48 h. The 35S::GFP line was used as the control. For GFP/mCherry colocalization experiments using laser scanning confocal microscopy, tobacco leaves, different tissues of SJP1 and SJP2 transgenic jujube seedlings, and transgenic jujube calli carrying SJP1-GFP, SJP2-GFP, and ZjTCP2-GFP were observed according to a previous report (Ma et al., 2018).

Expression analysis and western blotting

Samples including lateral buds (healthy dormant buds, infected dormant buds, and infected growing buds) (Zhou et al., 2021) and leaves of healthy and diseased jujube plants, 10-day-old transgenic Arabidopsis seedlings, and transgenic jujube calli and lines were collected. Total RNA was extracted using the RNAprep Pure Plant Kit (Tiangen, Beijing, China). For miRNA expression analysis, total RNA (1 μg) was reverse-transcribed using a miRNA 1st strand cDNA synthesis kit (Accurate Biotechnology Co., Ltd, Hunan, China) by the poly(A) tailing method. For gene expression analysis, qRT-PCR analyses were performed as previously described (Zhou et al., 2021). ZjACT1 was used as the reference gene for data normalization. qRT-PCR primers are listed in Supplementary Table S1. Error bars show the SDs from three independent experiments (**P< 0.01 and *P< 0.05; Student’s t-test). For western blotting, total proteins were extracted from the leaves of healthy and diseased jujube plants and transgenic jujube lines using lysis buffer (Zhou et al., 2021). The extracted protein was subsequently subjected to western blot analysis using antibodies against GFP, actin (ABclonal), SJP1/2 (Ma et al., 2023), and jujube ZjTCP2.

Generation of transgenic Arabidopsis lines and jujube calli

PsRobot (Wu et al., 2012) and TargetFinder (Fahlgren and Carrington, 2010) were used to predict target genes of miRNAs. ZjTCP2 (LOC107412598) coding sequences and the miRNA319f_1 precursor were fused to the N-terminus of mCherry and the GFP gene under the control of the CaMV35S promoter, respectively. To produce a gain of repression activity, ZjTCP2 coding sequences were also fused to the N-terminus of a repressor domain, SRDX (Hiratsu et al., 2003). The recombinant plasmids ZjTCP2-mCherry, ZjTCP2-SRDX and miRNA319f_1-GFP were transformed into R. radiobacter strain GV3101. Transgenic Arabidopsis lines in the Columbia-0 (Col-0) or ZjTCP2-mCherry background were generated by agroinfiltration using the floral dip method (Clough and Bent, 1998). Arabidopsis tcp2 plants were used as controls. Moreover, ZjTCP2-mCherry and miRNA319f_1-GFP were transformed into jujube calli by the Agrobacterium-mediated transformation system (Zhou et al., 2021). All 10-day-old transgenic Arabidopsis lines and jujube calli were used for growth and gene expression analysis. All the specific primers for vector construction are listed in Supplementary Table S1.

Yeast one-hybrid assays

For the yeast one-hybrid assay, the promoters of ZjKNAT1, ZjKNAT2, ZjKNAT6 were cloned and inserted into the HindIII and SalI restriction sites of the pAbAi vector. The coding sequences of ZjTCP2 were subsequently cloned and inserted into the pGADT7 vector to generate AD-ZjTCP2. Co-transformation of AD-ZjTCP2 with pAbAi-proZjKNAT1, pAbAi-proZjKNAT2, and pAbAi-proZjKNAT6 was carried out as previously described (Zhou et al., 2021).

Growth analysis of bearing shoots, secondary shoots, leaf size, and cell number

From late May to mid-June at 2 years after JWB phytoplasma infection, the bearing shoots from the healthy and diseased jujube plants were collected and photographed. ImageJ software was used to determine the length of the secondary shoots (n=8) from the four-month-old WT, 35S::SJP1 and 35S::SJP2 transgenic jujube plants. For the leaf area measurements, the mature leaves of bearing shoots (n=20) and secondary shoots (n=12) from jujube plants, the top and basal leaves of the SJP1 and SJP2 transgenic tobacco (100 d after transplantation, n=20), and the mature first leaves of the ZjTCP2 and miRNA319f_1 transgenic Arabidopsis lines (14 d after transplantation, n=10–15) were flattened and measured in the photographs using ImageJ software. All the leaves were cleared to remove chlorophyll overnight as described earlier (Challa et al., 2019). Images of abaxial epidermal cells were taken using a light microscope with a ZEN universal imaging system (Zeiss, Germany) and analysed with ImageJ software. For cell size analysis, the average area of the abaxial epidermal cells was measured from different fields of 10–20 leaves. The total number of cells per leaf was calculated as described earlier (Omidbakhshfard et al., 2018).

Results

JWB phytoplasma infection reduced leaf size of jujube plants

Phytoplasma infection alters the morphology of host plants, such as tomato (Pracros et al., 2006), Arabidopsis (Hoshi et al., 2009; Sugio et al., 2011; Maclean et al., 2014), petunia (Himeno et al., 2011), and Paulownia (Cao et al., 2021). JWB phytoplasma is a member of the 16SrV-B group, and as a result JWB phytoplasma-infected jujube plants displayed species-specific changes during bearing shoot (BS) development. Compared with the healthy BSs, the infected BSs showed morphological alterations of the sepals, petals, and stamens into small ‘leaves’ in the current year (Fig. 1A). The vegetative meristem from every leaf axil repeatedly initiated and grew into branches instead of flowers, resulting in bushy symptoms and leaf yellowing with dark green leaf veins 2 years after infection (Fig. 1A). The length (Fig. 1B) and leaf size (Fig. 1C, D) of the infected BSs were extremely reduced. In addition, the size of serrations on each side of the margin was increased (Fig. 1E). Unlike the jigsaw-puzzle appearance of pavement cells in healthy leaves, the cell shape was small and irregular in the infected leaves with decreased cell number (Fig. 1F, G). The morphological alterations of flowers and leaf shape suggested that JWB phytoplasma infection might employ several effectors to disturb cell proliferation and expansion, resulting in a reduced leaf size.

Fig. 1.

Jujube witches’ broom (JWB) phytoplasma infection decreased leaf size. (A) Comparison of leaf phenotypes from healthy and JWB phytoplasma-infected jujube plants. YAI, year after infection. (B) Representative phenotypes of healthy and JWB phytoplasma-infected bearing shoots. (C) Representative mature leaves of healthy and JWB phytoplasma-infected bearing shoots. (D) Mature leaf area of healthy and JWB phytoplasma-infected bearing shoots. n=20; significant differences were determined using Student’s t-test: **P<0.01. (E) Comparison of serrations between healthy and JWB phytoplasma-infected leaves. (F) Epidermal cells of healthy and JWB phytoplasma-infected leaves. (G) The cell number of the healthy and JWB phytoplasma-infected leaves. n=6; significant differences were determined using Student’s t-test: **P<0.01.

JWB phytoplasma effectors SJP1 and SJP2 participated in negatively controlling leaf size in jujube

In previous research, JWB phytoplasma effectors SJP1 and SJP2 induced typical witches’ broom symptoms with smaller leaves on the increased lateral branches (Zhou et al., 2021). Here, leaf areas were measured between the top (Supplementary Fig. S1A) and basal positions (Supplementary Fig. S1B) of lateral branches 90 d after transplantation when the apices stopped growing. The areas of both the younger and the older leaves were significantly reduced in the 35S::SJP1 and 35S::SJP2 transgenic tobacco compared with the control (Supplementary Fig. S1C, D). These observations indicated that effectors SJP1 and SJP2 participated in regulating leaf development.

To enhance the function of effectors SJP1 and SJP2 in regulating leaf development, transgenic jujube calli carrying 35S::SJP1-GFP and 35S::SJP2-GFP were first induced to produce shoots and roots. After transfer onto root induction medium after 8 weeks on the shoot induction medium, GFP-tagged roots were produced for another 8 weeks of induction. In situ fluorescence imaging showed that SJP1 and SJP2 were distributed in the four detected tissues, especially in the cytoplasm and nucleus of cells from lateral buds and leaves (Fig. 2), which was consistent with the subcellular localization of SJP1 and SJP2 in jujube calli (Supplementary Fig. S2). Thus, these results showed that SJP1 and SJP2 were mainly expressed in lateral buds and leaves of the transgenic jujube plants.

Fig. 2.

The tissue distributions of jujube witches’ broom phytoplasma effectors SJP1 and SJP2 in the transgenic jujube seedlings. Localization of SJP1 and SJP2 in different tissues of the transgenic jujube plants. Isolated leaves, stems, lateral buds, and root tips from transgenic jujube seedlings were cultured on screening medium for 16 weeks and observed under a laser scanning confocal microscope. Scale bars: 50 μm.

Next, these seedlings were subsequently grown in the soil (Fig. 3A) and identified by western blot using anti-SJP2 antibodies (Fig. 3B), which recognized both SJP1 and SJP2 because of their high similarity in amino acid sequence (Zhou et al., 2021; Ma et al., 2023). The transgenic jujube plants showed typical witches’ broom symptoms with a bushy and dwarfing appearance compared with the WT at 20 weeks after transplantation (Fig. 3A). Furthermore, the length of secondary shoots (SS) and the size of mature leaves in the transgenic jujube plants were smaller than those in the WT (Fig. 3C–F). Effectors SJP1 and SJP2 decreased the number of abaxial epidermal cells in the transgenic jujube plants (Fig. 3G). Most of the abaxial epidermal cells in the WT had large jigsaw shapes, whereas the abaxial surface of mature leaves in SJP1 and SJP2 transgenic jujube plants were composed mostly of elongated small cells (Fig. 3H, I). Together, these results confirmed that the JWB phytoplasma effectors SJP1 and SJP2 participated in the formation of small leaves in jujube.

Fig. 3.

Jujube witches’ broom phytoplasma SJP1 and SJP2 effectors reduced leaf size in jujube. (A) The dwarf and bushy phenotypes in SJP1- and SJP2-overexpressing transgenic jujube plants. Scale bar: 5 cm. (B) The expression of SJP1 and SJP2 in the mature leaves of the transgenic jujube plants. The protein extract was subjected to western blotting with anti-SJP1/2 antibodies. Ponceau S was used as the loading control. (C, D) Phenotypic comparison of secondary shoots (C) and mature leaves (D) in wild-type (WT) and 35S::SJP1 and 35S::SJP2 transgenic jujube plants. The red arrowheads indicate the shoots that grew out. Scale bars: 2 cm in (C) and 1 cm in (D). (E) Length of secondary shoots in WT and 35S::SJP1 and 35S::SJP2 transgenic jujube plants. n=8. (F) Mature leaf area of the secondary shoots in wild-type and 35S::SJP1 and 35S::SJP2 transgenic jujube plants. n=12. (G–I) Number (G), cellular shape (H), and average area (I) of the abaxial pavement cells shown in (D). Scale bars: 100 μm in (H). n=100 cells per leaf. Different letters indicate significant differences between plant lines as determined using one-way ANOVA (P<0.05, Duncan’s test) in SPSS 22.0 software.

JWB phytoplasma effectors SJP1 and SJP2 interacted with and stabilized ZjTCP2 expression

Since SJP1 and SJP2 could interact with CYC/TB1-TCP transcription factors from the class II group in jujube (Zhou et al., 2021), we hypothesized that these effectors might also recruit other members of the CIN-TCP subgroup. No interaction was observed in yeast cells (results not shown). However, BiFC assays showed that nucleus-localized fluorescence signals were observed in the combination of SJP2^ΔSP (lacking the signal peptide) and ZjTCP2 in Arabidopsis mesophyll protoplasts (Fig. 4A). The coiled-coil domains impaired the ability of SJP1 and SJP2 to bind their TCP targets (Zhou et al., 2021). When the signal peptide (SP) and the entire N-terminal coiled coil were deleted, both SJP1^ΔN and SJP2^ΔN interacted with ZjTCP2 (Fig. 4A). The interactions were further confirmed by Co-IP assays (Fig. 4B). Total protein was extracted from the leaves of 35S::SJP1-GFP and 35S::SJP2-GFP transgenic jujube plants. WT jujube plant and the 35S::GFP jujube calli were used as the negative controls. The protein‒protein complexes were immunoprecipitated using anti-GFP antibodies, and ZjTCP2 protein was enriched using sequence-specific antibodies against ZjTCP2 in the presence of SJP1 and SJP2 but not in the controls (Fig. 4B). These results indicated that effectors SJP1 and SJP2 interacted with ZjTCP2 in vivo.

Fig. 4.

Jujube witches’ broom (JWB) phytoplasma SJP1 and SJP2 effectors interacted with ZjTCP2 and promoted its accumulation. (A) BiFC assays confirmed the interaction of SJP1 and SJP2 with ZjTCP2 in Arabidopsis mesophyll protoplasts. SJP1 and SJP2 with the deletions ΔSP (the N-terminal signal peptide) and ΔN (from the signal peptide to the entire N-terminal coiled coil) were co-transformed with ZjTCP2 in Arabidopsis mesophyll protoplasts. The plasmid combination pSPYNE-ZjTCP2/pSPYCE was used as a negative control. Scale bars: 10 μm. (B) The interactions of SJP1 and SJP2 with ZjTCP2 were confirmed by co-immunoprecipitation assays. The protein‒protein complexes were immunoprecipitated using anti-GFP antibodies and subsequently subjected to western blotting using sequence-specific antibodies against ZjTCP2. (C) Co-localization of JWB phytoplasma effectors SJP1 and SJP2 with ZjTCP2 in tobacco leaves. (D) The expression of ZjTCP2 protein in the SJP1 and SJP2 transgenic jujube lines. Ponceau S was used as the loading control.

Next, the impact of effectors SJP1 and SJP2 on the subcellular localization and expression of ZjTCP2 were investigated. Subcellular co-localization showed that the visible fluorescence signal of ZjTCP2–mCherry was distributed in the nucleus with or without the presence of SJP1–GFP and SJP2–GFP (Fig. 4C). Furthermore, both the mRNA and protein levels of ZjTCP2 were increased in SJP1 and SJP2 transgenic jujube calli (Supplementary Fig. S3), and ZjTCP2 protein was accumulated in leaves of SJP1 and SJP2 transgenic jujube plants (Fig. 4D). Taken together, the JWB phytoplasma effectors SJP1/2 interact with and stabilize ZjTCP2 expression at the transcriptional and protein levels.

ZjTCP2 was expressed in JWB phytoplasma-induced small leaves and reduced leaf size

To investigate the regulation of ZjTCP2 in controlling leaf morphogenesis, we first analysed the expression of ZjTCP2 during JWB phytoplasma-induced bud outgrowth and its function in leaf development. ZjTCP2 (LOC107412598) was present in chromosome 1 (locus 6 391 776–6 400 915) in the jujube genome (Fig. 5A). Phylogenetic analysis showed that ZjTCP2 was homologous to Arabidopsis AtTCP2, with a conserved TCP domain and a diverse N-terminal region (Supplementary Fig. S4A, B). Subcellular localization indicated that ZjTCP2 was distributed in the plant nucleus of the transgenic jujube calli (Fig. 5B). ZjTCP2 was highly expressed at the infected growing lateral main bud compared with the healthy dormant bud (Fig. 5C). Compared with the healthy leaves, ZjTCP2 protein was highly accumulated in the JWB phytoplasma-infected small leaves (Fig. 5D). Overexpression of ZjTCP2 in Arabidopsis resulted in small-leaf phenotypes 14 d after transplantation (Fig. 5E), which were similar to those in JWB phytoplasma-infected jujube plants (Fig. 1A). No obvious morphological changes were observed in ZjTCP2-SRDX (a chimeric repressor) lines and the tcp2 mutant compared with 35S::GFP (Fig. 5E). Furthermore, the mature first rosette leaves of the 35S::ZjTCP2 lines were significantly smaller than those of the 35S::GFP, tcp2, and ZjTCP2-SRDX lines (Fig. 5F, G). These results indicated that ZjTCP2 participated in JWB phytoplasma-induced leaf development and negatively regulated leaf size.

Fig. 5.

ZjTCP2 was induced in jujube witches’ broom (JWB) phytoplasma-infected small leaves and reduced leaf area. (A) Chromosome location and gene structure of ZjTCP2 (LOC107412598). The black rectangle indicates the exon. (B) Subcellular localization of ZjTCP2 in jujube calli. 4ʹ,6-Diamidino-2-phenylindole (DAPI) stain was used to indicate the nucleus. Scale bars: 10 μm. (C) The expression of ZjTCP2 in latera main buds of the healthy and JWB phytoplasma-induced plants. HDB, healthy dormant bud; IDB, infected dormant bud; IGB, infected growing bud. Different letters indicate significant differences between plant lines as determined using one-way ANOVA (P<0.05, Duncan's test) in SPSS 22.0 software. (D) The expression of ZjTCP2 in healthy and JWB phytoplasma-infected small leaves. The protein extraction was subjected to western blotting with anti-ZjTCP2 antibodies. Anti-actin was used as the loading control. (E) Phenotypic characterization of Arabidopsis plants expressing 35S::ZjTCP2 and 35S::ZjTCP2-SRDX. 35S::GFP and Arabidopsis tcp2 mutant lines were used as controls. All the transgenic lines were photographed 14 d after transplantation. Scale bars: 1 cm. (F, G) Mature first leaves (F) and leaf area (G) of the transgenic lines. Values are means ±SD; n=10–15; significant differences between means were determined using Student’s t-test: **P<0.01; ns, no significant difference.

Considering the control of TCP genes in the regulation of cotyledon development (Palatnik et al., 2003; Koyama et al., 2007, 2017), we next assessed the growth of cotyledons in the 35S::ZjTCP2 and ZjTCP2-SRDX transgenic lines (Fig. 6A). The fusion of cotyledons was not observed in either transgenic lines or the tcp2 mutant (Fig. 6A), but a number of ectopic leaves were generated on the adaxial side of the cotyledons from the 35S::ZjTCP2 transgenic lines (Fig. 6B). The CUP-SHAPED COTYLEDON (CUC) genes were associated with the regulation of the cotyledon boundary and the formation of ectopic shoots on cotyledons (Laufs et al., 2004; Koyama et al., 2007). qRT-PCR analysis showed that the mRNA levels of AtCUC1, AtCUC2, and AtCUC3 were increased in the 35S::ZjTCP2 lines and decreased in the ZjTCP2-SRDX lines except for AtCUC1 (Fig. 6C). The class I KNOTTED1-like homeobox (KNOX) genes SHOOT MERISTEMLESS (AtSTM), AtKNAT1, AtKNAT2, and AtKNAT6 played distinct but partially redundant functions in the regulation of meristem maintenance and leaf organogenesis (Ori et al., 2000; Li et al., 2012). The transcript levels were all significantly elevated in the rosette leaves of 10-day-old 35S::ZjTCP2 seedlings grown on MS medium, while they were maintained at a lower level in ZjTCP2-SRDX, tcp2, and the empty control (GFP) (Fig. 6D). Moreover, the expression of ZjSTM, ZjKNAT1, and KNAT6 was activated in 35S::ZjTCP2 transgenic jujube calli (Fig. 6E). Y1H assays showed that ZjTCP2 directly bound to the ZjKNAT1, ZjKNAT2, and ZjKNAT6 promoters (Fig. 6F). These results indicated that the small-leaf phenotypes of the 35S::ZjTCP2 lines might be associated with the derepression of KNOX gene expression to induce abnormal leaf development.

Fig. 6.

Overexpression of ZjTCP2 promoted the formation of cotyledon-derived small leaves. (A) Phenotypes of Arabidopsis transgenic seedlings carrying 35S::ZjTCP2 and 35S::ZjTCP2-SRDX. All seedlings were grown on MS medium for 10 d. The red arrowheads indicate the cotyledons. Scale bars: 1 cm. (B) Morphological comparison of the cotyledons in ZjTCP2 transgenic lines. Scale bars: 1 mm. (C, D) The expression of the plant organ boundary genes CUC1/2/3 (C) and class-I KNOX genes (D) in 10-day-old 35S::ZjTCP2 and 35S::ZjTCP2-SRDX transgenic seedlings. Different letters indicate significant differences between plant lines as determined using one-way ANOVA (P<0.05, Duncan's test) in SPSS 22.0 software. (E) Expression of KNOXI genes in ZjTCP2 transgenic jujube calli. The standard deviation was obtained from three independent biological replicates (n=3; significance differences were determined using Student’s t-test: *P<0.05, **P<0.01). (F) ZjTCP2 directly bound to the ZjKNAT1, ZjKNAT2 and ZjKNAT6 promoters according to Y1H assays.

miRNA319f_{_}1 suppressed ZjTCP2 activity to control leaf size

Plant miRNA319 has been confirmed to target five CINCINNATA (CIN)-TCP genes by a post-transcriptional mechanism (Palatnik et al., 2003; Koyama et al., 2017). In this study, 13 mature sequences of the jujube miRNA319 family were predicted (Supplementary Fig. S5A) and some of them were mapped to the same known precursor (Supplementary Fig. S5B). Moreover, qRT-PCR results showed that miRNA319f_{_}1 expression was significantly down-regulated in the mature leaves of both the JWB phytoplasma-infected small leaves and SJP1 and SJP2 transgenic jujube plants (Fig. 7A, B), which showed the opposite pattern to ZjTCP2 (Fig. 7B). The mature sequences of miRNA319f_{_}1 were different from other members in jujube as well as three Arabidopsis miRNA319s (Supplementary Fig. S5A) but were predicted to recognize the cleavage site in the genome regions of ZjTCP2, AtTCP2, and AtTCP24 (Supplementary Fig. S6). To determine whether jujube miRNA319f_{_}1 specifically guided the cleavage of ZjTCP2 mRNA in vivo, the precursor form of miRNA319f_{_}1 was expressed in jujube calli. qRT-PCR analysis showed that the expression of ZjTCP2 was increased in the 35S::ZjTCP2 jujube calli and decreased in the 35S::miRNA319f_{_}1 and 35S::miRNA319f_{_}1 35S::ZjTCP2 jujube calli (Fig. 7C). When the precursor form of miRNA319f_{_}1 was overexpressed in the Col-0 background and 35S::ZjTCP2 lines, miR319f_1 abolished ZjTCP2 expression (Fig. 7D) and the effects on cotyledon-derived small leaves (Fig. 7E) and leaf size (Fig. 7F). Taken together, these results confirmed that the JWB phytoplasma effectors SJP1 and SJP2 induced smaller leaves in jujube, partly by stabilizing ZjTCP2 at the post-transcriptional and protein levels (Fig. 8).

Fig. 7.

miR319f_1 repressed the expression of ZjTCP2 and abolished its effects on cotyledon-derived leaves and leaf size. (A) Expression analysis of miR319f_1 in the healthy and jujube witches’ broom (JWB) phytoplasma-infected leaves. (B) The expression of miR319f_1 and ZjTCP2 in leaves of the SJP1 and SJP2 transgenic jujube lines. (C) Expression of ZjTCP2 in miR319f_1 transgenic jujube calli. In (B, C) the standard deviation was obtained from three independent biological replicates (n=3; significance differences were determined using Student’s t-test: *P<0.05). (D) Expression of ZjTCP2 in transgenic Arabidopsis lines. Different letters indicate significant differences between plant lines as determined using one-way ANOVA (P<0.05, Duncan's test) in SPSS 22.0 software. (E) Cotyledon phenotypes of 35S::miR319f_1 transgenic seedlings in the Clo-0 and 35S::ZjTCP2 backgrounds. 35S::GFP lines were used as the control. All seedlings were grown on MS medium for 10 d. Scale bars: 1 cm (top row) and 1 mm (bottom row). (F) Morphological comparison of miR319f_1 and miR319f_1/ZjTCP2 transgenic lines. All the transgenic lines were photographed at 14 d after transplantation (DAT). Scale bars: 1 cm.

Fig. 8.

Model of jujube witches’ broom (JWB) phytoplasma effectors SJP1- and SJP2-mediated formation of small leaves in jujube. JWB phytoplasma effectors SJP1 and SJP2 induced small leaves, at least partly, by elevating ZjTCP2 accumulation at the post-transcriptional and protein levels.

Discussion

Phytoplasma diseases cause serious yield and economic losses in more than 1000 plant species worldwide (Rao, 2021). Since it is difficult to cultivate phytoplasmas in vitro, few effective strategies for controlling phytoplasma diseases have been established. Phytoplasma-infected host plants show many common and representative symptoms in particular organs based on direct in-field and/or laboratory diagnosis (Ermacora and Osler, 2019; Rao, 2021). In our previous observations, JWB phytoplasmas from the 16SrV-B group induced witches’ broom disease (Zhou et al., 2021) and the formation of leaf-like structures, preventing the plant from reproducing (Ma et al., 2020; Deng et al., 2021). The bushy appearance was accompanied by plenty of small leaves in SJP1/2 transgenic tobacco lines (Zhou et al., 2021). In this study, we provided new findings about the mechanism of formation of small leaves mediated by the effectors SJP1 and SJP2.

Effectors SJP1 and SJP2 are two key regulators that negatively control leaf size in jujube

Phytoplasma-induced small leaves, associated with a range of phylogenetically distant plant species, has spread worldwide (Rao, 2021). JWB phytoplasmas have been reported to cause small leaves in brinjal (Solanum melongena) (Snehi et al., 2021), which is a very important vegetable parasitized by ‘Ca. Phytoplasma asteris’ from the group 16SrI (Kumar et al., 2012) and ‘Ca. Phytoplasma trifolii’ from the group 16SrVI (Yadav et al., 2015). The pathogenesis induced by these phytoplasmas might be conserved in jujube and other host plants. The typical and common symptom of JWB phytoplasma-infected leaves is an extreme decrease in size (Fig. 1C, D). It was noticed that the diseased leaves could be asymptomatic in size at the initial stages of JWB phytoplasma infection. Most of the small leaves occurred when the sepals, petals, and stamens reversed into leaves and shoots with newly emanating leaves sprouting from the base of the carpel (Deng et al., 2021). These species-specific morphological changes offer wide opportunities to investigate the interaction of JWB phytoplasmas with jujube during leaf development.

Since the observation of phytoplasma-induced small leaves in field crops in 1937 (Schneider et al., 1999), the virulence factors involved in this phytoplasma disease have remained largely unknown. In this study, the effectors SJP1 and SJP2 were identified as two key regulators that reduced leaf size in tobacco (Supplementary Fig. S1) and in jujube (Fig. 3), which was consistent with those caused by JWB phytoplasma infection (Fig. 1). These results indicated that effectors SJP1 and SJP2 are two virulence factors that induce the formation of small leaves in jujube. In recent decades, the full genome sequencing of phytoplasma strains from different phylogenetic clusters has provided a comprehensive understanding of metabolism and virulence factors (Kube et al., 2019). Several SJP1/2 homologues have been identified to induce witches’ broom and dwarfism but different leaf morphologies. Overexpression of the effectors SAP11_AYWB and SAP11_CaPM in Arabidopsis produced severely and mildly crinkled leaves, while leaves of SAP11_PnWB, SAP11_OYM, and SAP11_MBSP lines showed no apparent differences from the WT (Chang et al., 2018; Pecher et al., 2019). However, the leaf phenotypes caused by the SAP11_AYWB effector did not precisely account for the changes in leaf morphology caused by AYWB phytoplasmas (Sugio et al., 2011). These differences in leaf morphology induced by SJP1/2 and their homologues were associated with the protein abundances of these effectors and their interactors from the host plants. Because the transgenic jujube plants had a less severe leaf phenotype than JWB-infected jujube (Figs 1, 3), we do not exclude the possibility that JWB phytoplasmas secreted some other effectors to manipulate leaf size.

Effectors SJP1/2 interacted with and stabilized ZjTCP2 at the post-transcriptional and protein levels

In previous research, JWB phytoplasma effectors SJP1/2 induced witches’ broom symptoms by destabilizing ZjBRC1 (Zhou et al., 2021). Here, effectors SJP1 and SJP2 interacted with ZjTCP2 (Fig. 4A, B) and stabilized its accumulation in jujube leaves overexpressing SJP1 and SJP2 (Fig. 4D), which was consistent with the JWB phytoplasma-infected leaves (Fig. 5D). ZjTCP2 was homologous to Arabidopsis CIN-TCP AtTCP2 but showed significant differences in the N-terminal and C-terminal regions (Supplementary Fig. S4). Overexpression of ZjTCP2 in Arabidopsis significantly reduced the leaf area compared with 35S::GFP, tcp2 mutant, and ZjTCP2-SRDX lines, which exhibited less obvious changes in leaf size (Fig. 5E–G). This phenomenon was not observed in AtTCP2 transgenic Arabidopsis lines under the control of the constitutive 35S promoter but was consistent with those lines overexpressing the miRNA319-resistant version of AtTCP2 (Palatnik et al., 2003). In addition, miRNA319f_1 was identified and significantly repressed in JWB phytoplasma-induced small leaves (Fig. 7A), which showed the opposite pattern to its target ZjTCP2 in SJP1 and SJP2 transgenic jujube plants (Fig. 7B). Overexpression of miRNA319f_1 in WT and 35S::ZjTCP2 calli and ZjTCP2 transgenic Arabidopsis lines reduced the abundance of ZjTCP2 mRNA (Fig. 7C, D), indicating that the expression of ZjTCP2 was controlled by miRNA-guided cleavage. Thus, down-regulation of miRNA319f_1, at least partly, contributed to the SJP1- and SJP2-activation of ZjTCP2 expression.

SJP1/2 and their homologues specifically target TCP transcription factors. There seems to be a consensus that these effectors destabilize class II CYC/TB1-TCPs, such as BRC1/TCP18 and BRC2/TCP12 (Sugio et al., 2011; Tan et al., 2016; Chang et al., 2018; N. Wang et al., 2018; Pecher et al., 2019). However, the protein destabilization of class II CIN-TCPs was diverse in the presence of these effectors. SAP11_AYWB exhibited a stronger ability to destabilize the majority of CIN-TCPs than its homologues, including SAP11_CaPM, SAP11_PnWB, SAP11_OYM, and SAP11_MBSP (Sugio et al., 2011; Tan et al., 2016; Chang et al., 2018; Pecher et al., 2019), and it phenocopied the crinkled leaf morphology of the cin mutant (Pecher et al., 2019). It is worth noting that SAP11 homologues, except for SAP11_MBSP, destabilize a narrow set of CIN-TCPs and show different abilities to mainly destabilize Arabidopsis AtTCP2 (Tan et al., 2016; Chang et al., 2018; Pecher et al., 2019), which is involved in leaf development (Palatnik et al., 2003). How SAP11 effectors destabilize AtTCP2 transcription factors remains unknown. These observations indicate that JWB phytoplasmas evolved different mechanisms to induce small leaves in jujube.

The miRNA319f_{_}1–ZjTCP2 module participated in leaf size control

In this study, ectopic leaf primordia were observed on the adaxial surface of cotyledons in 35S::ZjTCP2 seedlings (Fig. 6B), and the smaller-leaf phenotype was exacerbated by the ectopic growth of leaves in clusters arising from the adaxial side of cotyledons (Figs 5E, 6A). This might be caused by misexpression of the CUC1, 2, 3, and KNOXI genes in 35S::ZjTCP2 seedlings and jujube calli (Fig. 6C–E), leading to abnormal cotyledon and leaf development (Li et al., 2012; Wang et al., 2021). To confirm that the smaller-leaf phenotypes were caused by the accumulation of target mRNAs, the miRNA319f_1 precursor was constitutively expressed in the 35S::ZjTCP2 background. The resultant miR319f_1 35S::ZjTCP2 rescued the rosette leaf defects (Fig. 7F). The rescued cotyledon defects were also found on the adaxial surface of cotyledons in miR319f_1 35S::ZjTCP2 seedlings (Fig. 7E). These observations indicated that miRNA319f_1-directed cleavage of ZjTCP2 mRNA was required for normal leaf development.

Arabidopsis class I TCPs promoted leaf growth, while class II CIN-TCPs showed functional redundancy to negatively regulate leaf growth (Efroni et al., 2008; Schommer et al., 2008). Loss of function of class II CIN-TCPs AtTCP2, 4, 10, or overexpression of miR319a (jaw-D) could increase leaf size (Karidas et al., 2015; Koyama et al., 2017; Challa et al., 2019). The effectors SJP1 and SJP2 effectively utilized the miRNA319f_1–ZjTCP2 module to decrease leaf size by decreasing cell number and cell size. The mechanism by which JWB phytoplasmas impeded the cell proliferation and expansion remains unknown.

In conclusion, our results showed that JWB phytoplasma effectors SJP1 and SJP2 were not only responsible for witches’ broom but were also virulence factors that induced small leaves in jujube, at least partly, by stabilizing ZjTCP2 in leaves at the post-transcriptional and protein levels (Fig. 8). This discovery will greatly contribute to expanding the understanding of small-leaf formation during phytoplasma–host plant interactions and also facilitate genetic improvement in phytoplasma-parasitized leaf vegetables. Further research will focus on identifying new virulence factors and their corresponding host plant targets involved in the regulation of leaf size.

Supplementary data

The following supplementary data are available at JXB online.

Fig. S1. JWB phytoplasma SJP1 and SJP2 effectors participated in leaf development.

Fig. S2. Localization of JWB phytoplasma effectors SJP1 and SJP2 in jujube calli.

Fig. S3. JWB phytoplasma effectors SJP1 and SJP2 elevated ZjTCP2 expression at the transcript and post-translational levels.

Fig. S4. Phylogenetic and amino acid sequence alignment analysis of ZjTCP2.

Fig. S5. Phylogenetic analysis of miR319f family and the mapping of mature sequences from jujube miRNA319 family to the known precursors.

Fig. S6. The prediction of target sites of miR319f_1 in jujube CIN-TCP transcription factors.

Table S1. List of primers used in this study.

Author contributions

JS, QS, and FM planned and designed the research; YZ, NZ, FM, MZ, JH, MD, and JZ performed the experiments; YL, CG, and GF conducted the transformation; JS, QS, and FM analysed the data; FM wrote the manuscript; and JS and QS revised the manuscript.

Conflict of interest

The authors declare that there are no conflicts of interest.

Funding

This work was supported by the National Natural Science Foundation of China (32002007 and 31971687), Anhui Province Key Research and Development Program (202004a06020008), Natural Science Foundation of Anhui Province (2008085QC127) and the University Natural Science Research Project of Anhui Province (KJ2019A0186).

Data availability

All data supporting the findings of this study are available within the paper and within its supplementary materials published online. Further inquiries can be directed to the corresponding author (Jun Sun).

Author notes

Fuli Ma, Yunyan Zheng and Ning Zhang contributed equally to this work.