ABA-CsABI5-CsCalS11 module upregulates Callose deposition of citrus infected with Candidatus Liberibacter asiaticus

Abstract Huanglongbing (HLB) primarily caused by Candidatus Liberibacter asiaticus (CLas) has been threatening citrus production globally. Under HLB conditions, an excessive accumulation of the polysaccharide callose in citrus phloem occurs, leading to phloem blockage and starch accumulation in leaves. The callose production is controlled by callose synthases (CalS), which have multiple members within plants. However, the knowledge of callose production in the citrus upon infection with CLas is limited. In this study, we firstly identified 11 CalSs in the Citrus sinensis genome through bioinformatics and found the expression pattern of CsCalS11 exhibited a positive correlation with callose deposition in CLas-infected leaves (correlation coefficient of 0.77, P ≤ 0.05). Knockdown of CsCalS11 resulted in a reduction of callose deposition and starch accumulation in CLas-infected citrus. Interestingly, we observed significantly higher concentrations of abscisic acid (ABA) in HLB-infected citrus leaves compared to uninfected ones. Furthermore, the expressions of CsABI5, CsPYR, and CsSnRK2 in the ABA pathway substantially increased in citrus leaves upon CLas infection. Additionally, the expression of CsCalS11 was significantly upregulated in citrus leaves following the application of exogenous ABA. We confirmed that CsABI5, a pivotal component of the ABA signaling pathway, regulates CsCalS11 expression by binding to its promoter using yeast one-hybrid assay, dual luciferase assay, and transient expression in citrus leaves. In conclusion, our findings strongly suggest that the CsABI5-CsCalS11 module plays a crucial role in regulating callose deposition through the ABA signaling pathway during CLas infection. The results also revealed new function of the ABA signaling pathway in plants under biotic stress.


Introduction
Huanglongbing (HLB) represents one of the most devastating diseases of citrus worldwide [1], decimating the citrus industry in over 60 citrus-growing countries and regions (https://www.cabi.org).Since the disease was first reported in Florida, the United States in 2005, the citrus production in the state has decreased from 174.8 million boxes in 2005-2006 to 45.1 million boxes in 2021-2022 (https://www.nass.usda.gov).In Ganzhou, Jiangxi province, China, around 50 million affected trees were removed and destroyed from 2013 to 2018 due to an outbreak of HLB [2].The causal pathogen of HLB are three species of phloem-limited, non-cultured, and Gram-negative α-Proteobacteria, namely Candidatus Liberibacter asiaticus (CLas), Candidatus L. africanus L. americanus (CLam), and Candidatus (CLaf).Among them, CLas is the primary pathogen of HLB and is predominantly distributed through the Asian citrus psyllid, Diaphorina citri, and grafting with diseased scion [1,3].CLas is capable of infecting almost all commercial citrus species and scions, and still now, no resistant cultivars have been identified.
HLB symptoms in CLas-infected plants predominantly comprise chlorosis and asymmetrical blotchy mottling leaves, "red-nose" fruits, and rotted fibrous roots [1,3,4].A growing number of research strongly suggests that the phloem transport malfunction in HLB-affected citrus is responsible for these disease symptoms.The dysfunctional phloem obstructs the transport of photosynthetic products from source to sink, resulting in starch accumulation in the leaves and nutrient deficiency in the roots [5].This eventually leads to HLB symptoms and even plant death [6,7].Callose, a 1-3 linked β-glucan polymer, is a crucial contributor to phloem dysfunction in citrus affected by HLB.In symptomatic citrus plants, excessive amounts of callose deposits are visible around the sieve pore, which is usually accompanied by phloem collapse and systemic cell death [8].Numerous callose accumulation can be also observed in symptomless leaves affected by HLB [9], while phloem cell collapse is not visible [6].Callose deposition serves not only as a symptom of HLB, but also as an indicator of sensitivity and resistance to the disease.HLBtolerant citrus, with lower levels of callose deposition and starch accumulation compared to susceptible citrus, can regenerate new phloem to mitigate obstacles, maintain vascular function, and sustain new shoot production [6,[10][11][12].Furthermore, gene-overexpressed susceptible citrus with SAMT1 and NPR1 demonstrated a significant reduction in callose accumulation upon CLas infection and exhibited mild symptoms [13].
Callose synthase (CalS), also known as β-1,3-glucan synthases (GSL), directly catalyze callose biosynthesis utilizing UDP-glucose as substrate.CalS members have been identified and analyzed in various plants, including model plant Arabidopsis, dicot plants, and monocot plants [14][15][16].These CalSs catalyzed callose biosynthesis in phloem, cell wall, cell plate, pollen tube, and other tissues and positions, playing an important role in systematic transportation, intercellular permeability, and plant protection from biotic and abiotic stresses [17].Arabidopsis AtCalS12 and its homologous TaGSL22 of wheat were significantly increased by infection of fungal pathogens.Overexpression of AtCalS12 (PMR4) enhanced callose deposition in the cell wall and conferred complete resistance to both the non-adapted and the virulent powdery mildew pathogens in Arabidopsis and barley [18,19], while knockdown or knockout of CalS in plants could decrease callose accumulation and increase susceptibility of pathogens.Silencing of AtCalS12 ortholog in wheat and barley increased cell wall penetration by Blumeria graminis [20,21].T-DNA insertion within AtCalS10 caused high susceptibility to Pseudomonas syringae in Arabidopsis, and reduced expression of ClCalS1 weakens resistance to Xanthomonas citri in lemon [9,22].The Atcals7ko mutation leads to increased susceptibility to Chrysanthemum Yellows phytoplasma [23].
The expression of various CsCalSs was modified in citrus trees affected by HLB in comparison to the healthy ones [12,24,25].However, it remains unclear which key CsCalS enhances callose deposition in citrus phloem under HLB stress and its regulation mechanism.Here, eleven CsCalSs were screened and identified in the citrus genome.Among the CsCalS family, CsCalS11 had been demonstrated to possess significantly higher expression in leaves of healthy citrus and was greatly upregulated in response to HLB stress.We analyzed the relationship between the expression of CsCalS11 and the callose deposition during CLas infection and also identified the transcriptional factor CsABI5 that regulated CsCalS11.As a result, the molecular mechanism of callose deposition regulated by CsCalS11 through the ABA pathway under HLB stress was unveiled.
In addition, CsCalS11 expression level was higher than CsCalS9 in CLas-free leaves of the above three citrus varieties, which maybe have more important role in plant growth.Therefore, CsCalS11 was chosen for further research.

CsCalS11 expression had a positive correlation with callose deposition under HLB stress
To analysis of the relationship between CsCalS11 expression and callose deposition in response to HLB stress, one-yearold "Wanjincheng" seedlings in the greenhouse were subjected to CLas inoculation through leaf disc grafting.At seven week post-inoculation (wpi), CLas was detected to be positive for the first time.Hence, the relative expression of CsCalS11 was quantified by RT-qPCR between HLB plants and healthy ones every 4 weeks since 7 wpi, along with counting callose spots.No significant variations in CsCalS11 expression or callose deposits were observed between HLB and healthy leaves from 7 to 15 wpi.However, starting at 19 wpi, a significant increase in both the CsCalS11 expression and callose deposits was observed in the HLB citrus.At 39 wpi, the fold change in CsCalS11 expression and callose deposits in HLB citrus reached a peak, with 4.53-and 3.25fold expression respectively, compared with those in the healthy control (Figure 2A-2C).Furthermore, statistical analysis results revealed a positive correlation between the CsCalS11 expression and callose deposition under HLB stress, with a Pearson correlation coefficient of 0.77 (P ≤ 0.05) (Figure 2D).
To compare the expression of CsCalS11 in citrus leaves treated with different phytohormones, RT-qPCR was performed.The results showed that CsCalS11 was significantly induced by 100 μmol•L −1 abscisic acid up to 6.54 times higher than in the control.Its expression could not be changed by 10 μmol•L −1 salicylic acid, 100 μmo•L −1 methyl jasmonate, and 10 μmol•L −1 ethephon (Figure 2E).
CsCalS11 was difficultly cloned into the overexpression plasmid pLGNL in Escherichia coli at one time for its long sequence with nearly 6000 bp, which was larger than most plant genes.The pLGNL-CsCalS11 was successfully constructed through cloning fragment 3 (2477 bp), fragment 2 (1689 bp), fragment 1 (1693 bp) onto vector pLGNL (Figure 3A).For confirming the function of CsCalS11 in callose biosynthesis, it was transiently overexpressed in "Wanjincheng" (C.sinensis) leaves.As expected, the expression of CsCalS11 was significantly up-regulated and the reference gene GUS was not changed.Furthermore, callose content were significantly increased in transiently overexpressed leaves in three separate experiments (Figure 3).

CsCalS11 knock-down decreased callose deposition in sweet orange
Four CsCalS11-RNAi plant strains (I-33, I-38, I40, and I-41) were generated by Agrobacterium tumefaciens carrying the vector pLGNL-CsCalS11-RNAi (Figure 4A).The propagation was subsequently carried out using the grafting technique, and three plantlets were obtained from each of the strains.Compared to the non-transgenic controls, CsCalS11 expression was significantly down-regulated in the range of 38.02% to 48.72% through RNA interference (Figure 4B).The callose content in leaves of CsCalS11-RNAi citrus seedlings was significantly decreased, from 3.23 ± 0.05 ng•g −1 in non-transgenic ones down to 2.70 ± 0.16 ng•g −1 (Figure 4C).Compared with the non-transgenic control, the average value of leaf number and plant height were reduced in the CsCalS11-RNAi scions, being down-regulated with 16.67% to 57.14% and 13.94% to 67.41%, respectively (Figure 4D-4E).At 10 months after being grafted with CLas-infected leaf discs, the symptomatic leaves of the CsCalS11-RNAi plants only exhibited chlorosis, while the non-transgenic controls presented serious symptoms with both leaf vein burst and chlorosis (Figure 4F).The callose concentration in the leaves of CsCalS11-RNAi plants was further reduced by 31.04% to 43.70% compared with that of the non-transgenic citrus.Furthermore, the titer of CLas was extremely significantly decreased in the CsCalS11-RNAi plants, especially the bacteria population in I-38 was 0.51% of ones in

CsCalS11 transcription was activated by CsABI5 via directly binding on its promoter
The CsCalS11 promoter (CsCalS11p), 1500-bp upstream of the starting codon, were analyzed with PlantCARE software and found several hormone-responsive cis-elements in the promoter sequence.They were an ABRE motif (involved in the ABA responsiveness), a TCA-element (involved in salicylic acid responsiveness), and two GARE-motif (involved in gibberellin responsiveness).Moreover, cis-acting elements involved in abiotic stress responsiveness were identified, including a WUN motif (involved in wound responsiveness) and MBS (involved in droughtinducibility) (Supplemental Table S3).
To screen candidate proteins that might directly regulate CsCalS11, a cDNA library was constructed using equally mixed RNA extracted from leaves of both HLB "Wanjincheng" and healthy ones.The titer of the cDNA library is 9.26 × 10 7 cfu•mL −1 , and the length of 80% cDNA insertion was more than 750 bp.The cDNA library was then subjected to yeast one-hybrid assay with pAbAi-CsCalS11p as bait.Twenty-four candidate proteins were identified, which showed potential binding ability with the CsCalS11 promotor (Supplemental Fig. S2 and Supplemental Table S5).To further identify the binding and activating ability, the open reading frames (ORFs) of a transcriptional factor CsABI5 in the ABA signal pathway was cloned into the pGADT7 vector.Subsequently, one-on-one bait/prey interactions were conducted and revealed that CsABI5 could bind to the CsCalS11 promoter (Figure 5A).At the same time, a dual-luciferase expression assay was conducted in tobacco leaves.The results demonstrated that CsABI5 could enhance the transcription of the luciferase gene driven by the promoter of CsCalS11 to 2.81 times, with a significant increase in luciferase activity using REN as a reference (Figure 5C).
For examining CsCalS11p activation in citrus, the promoter was inserted upstream of the reporter gene GUS in the pGNGM1300 vector (Figure 6A).Through A. tumefaciens transformation, three independent transgenic citrus lines containing CsCalS11p::GUS (P23, P57, and P58) were generated without significant differences to those non-transgene plants in vegetative growth patterns (Figure 6B-6D).After 42 weeks post-infection with CLas, the GUS expression in the transgenic citrus lines was increased compared with that of the healthy control (Figure 6E).On the other way, the expression of GUS was significantly up-regulated to 2.09, 2.30, and 1.93 times respectively in leaves of P23, P57, and P58 treated with 100 μmol•L −1 ABA (Figure 6F).When CsABI5 was transiently expressed in P23 leaves transformed with CsCalS11p::GUS.The overexpression of the transcriptional factor significantly upregulated the expression of GUS driven by the CsCalS11 promoter to 1.86 times (Figure 6D).Therefore, these findings suggested that the transcription of CsCalS11 could be activated by CLas, ABA, and CsABI5 in vivo.

ABA concentration and genes in the ABA signal pathway were induced by CLas
The results of exogenous phytohormone experiments showed that the expression of CsCalS11 was positively induced by ABA (Figure 2E and 6F).Therefore, we investigated the concentration of ABA and the expression of genes related to the ABA signaling pathway in citrus leaves under HLB stress.It was found that ABA concentration presented a significant increase to 8.31 ng•g −1 in HLB citrus, 1.73 times greater than the phytohormone in healthy one (Figure 7A).Moreover, some important genes in the ABA signal pathway were up-regulated in citrus affected by HLB.The expression of a PYR/PYL/RCAR-like receptor binding with ABA (CsPYR), protein kinase CsSnRK2, and CsABI5 was significantly increased in citrus leaves affected by HLB compared to the healthy leaves.A negative regulator in the ABA pathway, CsPP2C, was downregulated in HLB-affected citrus.These results indicated that ABA accumulation and its signal pathway could be induced in citrus by CLas infection.

Multiple CsCalSs were stimulated in response to CLas
Callose strengthens the cell wall and regulates the permeability of intercellular junction via plasmodesmata.Therefore, the biosynthesis of callose to combat infections induced by fungal, viral, and bacterial pathogens has garnered considerable attention.AtCalS12 is the main biosynthetic enzyme responsible for the callose deposition at the invasion site of fungal pathogens, and its regulation hinges on salicylic acid and NPR1 dependence [26,27].Its homologous TaGSL22 in wheat is the only CalS gene induced by Blumeria graminis [20].Additionally, AtCalS1, AtCalS5, AtCalS9, and AtCalS10 can be induced by Hyaloperonospora arabidopsis infection [27].AsCalS1-like gene was up-regulated by Pseudoperonospora cubensis (Downy mildew), and AsCalS5-like and AsCalS10-like were increased by Sphaerotheca fuliginea (powdery mildew) in cucumber as well [28].Otherwise, the level of AtCalS7like and AtCalS5-like was significantly increased in Arabidopsis infected with Chrysanthemum Yellows phytoplasma and cotton fed on by aphids respectively [23,29].
CLas is a phloem-restricted bacterium that is unevenly distributed throughout citrus plants.Even within the same CLaspositive plant, various leaves contained varying concentrations of the pathogen and some of the leaves are not colonized by the bacterium at all.These variations present significant challenges when evaluating the expression of CsCalSs in citrus affected by HLB.Regardless of the citrus variety and timing of CLas infection, the expression of AtCalS2, AtCalS7, AtCalS8, AtCalS9, and AtCalS12-like gene members were positively stimulated in response to HLB stress.AtCalS2-like and AtCalS7-like genes are potentially major regulators of callose accumulation in the HLBaffected phloem as they were significantly increased in diseased C. sinensis, C. grandis, and C. reticulate in both our lab and Granato' lab [25].

CsCalS11 served an important role in the callose deposition during CLas infection
Although the expression of multiple AtCalSs can be regulated under biotic stresses, only a few of them have be revealed function.AtCalS12 catalyzes callose biosynthesis in papillae of the cell wall to strengthen host resistance when the plant is infected with fungal pathogens [18].The atcals7ko mutant of Arabidopsis is sensitive to phytoplasma due to enhanced plasmodesmata permeability [23].Little is known about AtCalS2, except that it physically interacts with a Phytophthora effector protein RxLR3 [30].
Several CsCalS genes are up-regulated in citrus leaves under HLB stress.However, the major gene contributing to callose deposition remains unclear.CsCalS11, homologous with AtCalS2, was significantly increased in the HLB-positive citrus not only in the greenhouse but also in the field, compared with the healthy trees (Figure 1D and Supplemental Fig. S1).Additionally, its fold change between CLas-infected and healthy leaves possessed a positive correlation with the ratio of callose deposits in the phloem of leaf midrib (Figure 2D).CsCalS11 knock-down could significantly reduce callose deposition in citrus under HLB stress.Moreover, the marker gene GUS driven by the CsCalS11p promoter could be induced by CLas in CsCalS11p::GUS-transformed seedlings (Figure 6E).Thus, it could be concluded that CsCalS11 plays a vital role in callose deposition in the phloem of CLas-infected citrus.
Callose in the phloem of CLas-infected citrus hinder the transportability of sieve elements, consequently leading to HLB symptoms [5,6,24].Its accumulation is regulated by two CLas-secreted polypeptides in opposing ways, as the function of CLas-effectors has been increasingly investigated.A Sec-dependent polypeptide SECP8 of CLas localized in nucleus, cytoplasm, and cytoplasmic membrane and depressed callose deposition in Nicotiana benthamiana [31].While the Las5315mp effector elicits significant callose deposition, cell death and starch accumulation in tobacco upon transient expression [32].It is necessary to analyze whether these effectors can physically interact with CsCalS11 and their regulatory role in the activity of callose synthase for the future research.

CsCalS11 was increased by CsABI5 via ABA signal pathway
Callose biosynthesis is known to be promoted by abscisic acid (ABA) [33].In the dormant bud of Hybrid aspen (Populus tremula × tremuloides), ABA can induce the expression of CalS1 by repressing PICKLE, a chromodomain protein, and enhancing the expression of the transcription factor of SVL (short vegetative phase-like) [34].This process was dependent on PP2C family protein ABI1 interacting with the ABA receptor PYR/PYL/RCAR [35].In plant resistance to fungal pathogens, it is also found that exogenous ABA can induce CalS activity and promote callose deposition [33].However, it has not been revealed which of CalSs would be dependent on ABA in plant resistance to biotic stress and its regulating mechanism.In CLas-infected citrus leaves and fruit f lavodo, a notable increase in ABA levels was observed compared with the healthy ones [36,37].The ABA signaling positively modulates plant resistance against various pathogens by facilitating the accumulation of callose in host plants [38].Our study revealed that CsCalS11 expression could be induced by exogenous ABA (Figure 2E).Moreover, the expression of marker gene GUS driven by the CsCalS11 promoter was upregulated in leaves of CsCalS11p::GUS transformed citrus treated with ABA (Figure 6F).This suggested that CsCalS11 can be regulated via the ABA signal pathway during CLas infection.
ABI5 is a bZIP transcription factor that derives its name from the insensitivity to ABA in mutant form [39].In response to biotic and abiotic stresses, ABA binds to its PYR/PYL/RCAR-type receptors, and the latter combines and deactivates ABA repressor PP2C.
Subsequently, the SnRK2 is dissociated from PP2C and activated through self-phosphorylation, resulting in the phosphorylation and activation of ABI5 by SnRK2 [40].ABI5 can enhance the expression of ABA-responsive genes by binding to the ABRE motif in the promoter [39][40][41].In addition to preventing seed germination and post-germinative growth under unfavorable conditions, ABI5 also regulates physiological processes under abiotic stress during vegetative growth [42].In this study, CsABI5 was found to activate the expression of CsCalS11 by directly binding to its promoter (Figure 5 and 6G).Moreover, ABA concentration in HLBaffected leaves was significantly 1.73 times higher than that in healthy leaves (Figure 7A).CsABI5 and other genes in ABA signal pathway were up-regulated by CLas infection (Figure 7B).Therefore, it is reasonable to speculate that the regulation of callose deposition in HLB citrus phloem occurs through an ABA-CsABI5-CsCalS11 module (Figure 8).

Conclusion
This study revealed novel insights into the critical functions of CsCalS11 in mediating callose deposition in citrus under HLB stress and presented a comprehensive regulatory model.Specifically, CLas infection stimulated the accumulation of ABA, which was followed by enhanced expression of the transcriptional factor CsABI5.CsABI5 activated CsCalS11 by directly binding to its promoter.The resulting upregulation of CsCalS11 contributed to an abnormal callose deposition in the phloem.

Plant material and treatments
One-year "Wanjincheng" (C.sinensis) plants grafted onto "Ziyang Xiangcheng" (C.junos), RNA-interference plantlets, and transgenic plants with CsCalS11p::GUS were cultivated in a greenhouse with an average temperature of 26 • C.These trees were infected with CLas through leaf disc grafting with restricted access, and the infection was subsequently confirmed by PCR [43].All primers are listed in Supplemental Table S4 and synthesized in Tsingke Biotechnology Company.Leaves of C. grandis and C. reticulata were harvested from a commercial citrus grove in Guangxi province, China.Three HLB-affected citrus trees and three healthy ones were chosen for biological replicates in greenhouse and in field.Four fully mature leaves were collected per citrus and mixed for experiments.

Identification and bioinformatics analysis of the callose synthase enzyme family in citrus
The coding DNA sequences encoding callose synthase in C. sinensis were retrieved from NCBI (http://www.ncbi.nlm.nih.gov) and the citrus genome databases CPBD (http://citrus.hzau.edu.cn) and Phytozome V12 (https://phytozome.jgi.doe.gov).Following the removal of redundant sequences, all putative CsCalS protein sequences were subjected to the Simple Modular Architecture Research Tool (SMART, http://smart.embl-heidelberg.de)analysis, whereby the sequences with the Glucan synthase domain, FSK1 domain, as well as N-terminal and C-terminal transmembrane regions, were considered as true CsCalS protein in citrus.The exon-intron organization was determined with TBtools [44].Phylogenetic reconstruction was accomplished through the maximum likelihood method with MEGA V11 [45].

RNA extraction and RT-qPCR
Total RNA was extracted from leaf tissues with an EASY spin plant RNA extraction kit (Aidlab, Beijing, China), as well as with subsequent elimination of genomic DNA using DNase (Takara, Beijing, China).First, strand cDNA was synthesized from 0.5 μg of total RNA using the 5× PrimeScript RT Master Mix (TaKaRa, Beijing, China) according to the manufacturer's instructions.qPCR was performed using ChamQ™ Universal SYBR qPCR Master Mix (Vazyme, Jiangsu, China) on CFX96™ Real-Time System (BIO-RAD, California).Amplifications were conducted in triplicate for each sample, accompanied by the appropriate negative controls, using the following conditions: 95 • C for 1 minute, 40 cycles of 95 • C for 15 seconds, and 60 • C for 1 minutes.The housekeeping gene actin (GenBank No. GU911361.1)was applied as a reference gene for RT-qPCR analysis.All experiments were performed on three biological replicates.The relative expression values were calculated with the 2 − Ct method unless differently indicated.

Phytohormone treatments
Citrus leaves were treated with exogenous phytohormones according to the description from Yang et al [46].Mature leaves of Wanjincheng were collected and cut into leaf discs using a punch.The mixed leaf discs were divided into triangular bottles with 100 μmol•L −1 abscisic acid, 10 μmol•L −1 salicylic acid, 100 μmol•L −1 methyl jasmonate, or 10 μmol•L −1 ethephon and cultured at room temperature for 48 hours.Hormone treatments were repeated three times and leaves treated with distilled water served as controls.

Observation of callose deposition
Leaf midribs were cut into a length of 2 to 3 mm and immersed in a 4% paraformaldehyde universal tissue fixation solution, followed by dehydration using acetone solutions and being embedded in LR White Resin according to the description from Zou et al. [13].Subsequently, 0.8 μm slices were obtained from the embedded midribs through a Leica EM UC7 Ultramicrotome (Leica Microsystems, Wetzlar, Germany) and stained with 0.05% aniline blue for 5 seconds.The resulting blue-stained callose deposition in the phloem was captured by an Olympus BX51 microscope (Olympus, Tokyo, Japan) and counted with ImageJ software.

CsCalS11-RNAi plasmid construction
The CsCalS11 RNAi vector was constructed and transformed into citrus via the method described by Yao et al [47] with a few modifications.A 398-bp fragment was amplified by PCR with CsCalS11-i3F and CsCalS11-i3R primers.The restriction enzyme sites of SwaI and XbaI were present in the CsCalS11-i3F primer, while AscI and BamHI were present in the CsCalS11-i3R primer.The 398-bp PCR products were digested with SwaI and AscI to acquire forward fragments and with XbaI and BamHI for reverse fragments.Both fragments were then inserted into the corresponding sites of the pUCRNAi plasmid, respectively.The CsCalS11-RNAi fragment was digested with KpnI and SalI from the reconstituted pUCRNAi and integrated into the pLGNL.

Plasmid construction of CsCalS11p::GUS
A 1500 bp region upstream of the translation initiation codon of CsCalS11 (Cs_ont_7g029230.1) was amplified from C. sinensis genomic DNA with specific primers CsCalS11P-F and CsCalS11P-R.The CsCalS11p PCR products were ligated with a pUCm-T vector for sequencing, and the correct sequence was predicted for cis-acting elements through PlantCARE [48].Subsequently, the CsCalS11p fragment was digested using BamHI/PstI and inserted into the BamHI/PstI site of the pGNGM1300.The resulting recombinant vector including CsCalS11p::GUS was generated and transformed into the A. tumefaciens strain EHA105.

Measurement of callose concentration
The callose concentrations were quantified through the application of a plant callose ELISA kit (Boshen, Jiangsu, China), according to the manufacturer's protocols.The detection was conducted at 450 nm wavelength with JS-THERMO Varioskan Flash (Thermo, Massachusetts, USA).

Abscisic acid (ABA) quantification
Three mature leaves were collected per line and ground into a fine powder using liquid nitrogen.The endogenous ABA content of three HLB-citrus lines and healthy lines was determined by extraction and detection with LC-MS/MS (AB Sciex TripleTOF 5600+) through multiple reaction monitoring as described previously [49].

Starch quantification
Starch was extracted according to the protocol of the Starch Assay Kit (G-clone, Beijing, China).Brief ly, the diluted starch samples were incubated with reaction buffer at 95 • C for 10 min and measured at 620 nm using glucose as a standard.The starch content in fresh leaves was calculated using the formula: Starch content (mg•g −1 fresh weight) = 13.51×•F,where x represents the starch concentration calculated with glucose standard, and F represents the dilution rate of the sample.

Yeast one-hybrid assay
The yeast one-hybrid assay was conducted according to the instructions provided by the Matchmaker ® Gold Yeast One-Hybrid Library Screening System (Clontech, California, USA).CsABI5 (synthesized by Tsingke, Chongqing, China) were inserted into the prey plasmid pGADT7 through in-fusion cloning.pGADT7-CsABI5 and the pAbAi reporter vector ("bait") containing the CsCalS11 promoter were co-transformed into the Y1HGold strain.The transformants were cultured on SD/-Leu/AbA 200 medium for 3 days at 30 • C.

Dual luciferase assay
The CsCalS11p promoter was amplified using eCsCalS11p-F/R primer pairs and inserted into the pGreenII 0800-LUC plasmid after being digested with SalI and BamHI.This promoter was utilized to regulate the expression of the firef ly luciferase gene (LUC) in pGreenII 0800-LUC, with the Renilla luciferase gene (REN) under CaMV 35S promoter serving as an internal control of the transient expression.CsABI5 was cloned into the pLGNL vector for protein expression and served as effectors.The reconstituted vectors were separately transformed into A. tumefaciens strain GV3101 via the freeze-thaw method.The agroinfiltrated N. benthamiana leaves were grown at 25 • C for 4 days and analyzed using the dual-glo ® luciferase assay system (Promega, Wisconsin, USA) in a 96-well plate [50].The LUC/REN ratio was calculated.

Citrus transformation
The cut epicotyls were employed as explants for citrus transformation with A. tumefaciens following the methodology described by Zou et al. [13].The transformed seedlings were firstly selected through GUS staining or green f luorescence with a LUYOR-3415RG hand-held lamp, and then confirmed by PCR with DNA as template.When the transformed citrus strains grew to 30 cm, they were propagated three to five seedlings for each strain through grafting onto the "Ziyang Xiangcheng" rootstock.

Transient expression in plant leaves
A. tumefaciens strain GV3101 harboring the reconstructed vectors were injected into the citrus leaves, with a method that combined the protocols of Li et al. [51] and Acanda et al. [52].The overnight grown GV3101 was centrifuged at 5000 rpm for 10 min and resuspended to an A 600 of 0.8 with MMA solution (10 mmol•L −1 MgCl 2 , 15 mmol•L −1 2-(N-morpholino) ethane sulfonic acid, 200 μmol•L −1 acetosyringone, pH = 5.6).The three fully expanded leaves were subjected to surface sterilization and infiltrated until 1/3 of the leaf was saturated.Subsequently, they were placed on AIM solid medium (0.4% (w/v) MT, 0.4% (w/v) sucrose, 15 mmol•L −1 2-(N-morpholino) ethanesulfonic acid, 200 μmol•L −1 acetosyringone, pH = 5.6) and incubated at 22 • C in the dark for 6 days.Following this, the leaves were collected for RT-qPCR analysis.As a negative control, leaves infiltrated with GV3101 containing the empty pLGNL plasmid were used.This experiment was replicated three times.

Statistical analysis
The experiments were performed with three replicates.Graph-Pad Prism version 8.0.2 (Graphpad Prism Inc., USA) was used for statistical analysis.All comparisons in the study were made using Student's t-test.P ≤ 0.05 was considered statistically significant, P ≤ 0.01 was very significant, and P ≤ 0.001 was extremely significant.Results were expressed as mean values ± standard deviation (SD).

Figure 3 .
Figure 3. Transient expression of CsCalS11 in citrus leaves.A: Schematic representation of the CsCalS11 overexpressing vector based on the pLGNL binary plasmid employing the In-fusion strategy.11-3F/11-3R, 11-2F/11-2R, and 11-1F/11-1R were primer pairs for cloning fragments 3, 2, and 1 of the CsCalS11 respectively.pLGNL, 11-3F, and 11-2F include the SmaI cleavage site CCCGGG.The primer sequences from CsCalS11 are represented in capital letters and lowercase sequences from pLGNL in black letters.The expression of CsCalS11 (B) and GUS (C) and callose content (D) were quanlitified in transiently overexpressed citrus.The experiments were repeated for three times with the similar results.* : P ≤ 0.05.

Figure 5 .
Figure 5.The CsCalS11p was bound and activated by CsABI5 in yeast and tobacco.A: Physical interactions of CsABI5 with CsCalS11p in yeast one hybrid (Y1H) assays.The pGADT7-Rec-53 was introduced into yeast cells carrying pAbAi-P53 as a positive control.The empty vector pGADT7 was included as a negative control.Another negative control was yeast containing pGADT7-CsABI5 and pAbAi-P53.AbA 200 means 200 ng•nL −1 Aureobasidin A. B: Reporter and effector constructs for dual LUC assays.LUC: firef ly luciferase, REN: Renilla luciferase, 35S-P: CaMV35S promoter, 35S-T: CaMV35S terminator.CsCalS11p::LUC was constructed based on pGreenII 0800-LUC.The control effector construct was from the pLGNL plasmid.C: Dual luciferase activity assay performed by transient expression in Nicotiana benthamiana leaves.* P ≤ 0.05.

Figure 6 .
Figure 6.The CsCalS11 promoter (CsCalS11p) was activated under conditions of Huanglongbing (HLB) stress, exogenous abscisic acid (ABA) exposure and transient overexpression of CsABI5.A: Schematic representation of the CsCalS11p::GUS constructs T-DNA of the pGNGM1300 binary vector.B: The transgenic seedlings with CsCalS11p::GUS.C: PCR analysis with primer pair GUS-F/GUF-R using leaf DNA from (B) seedlings.D: GUS staining of leaf discs from (B) seedlings.E: The relative expression of GUS in CsCalS11p::GUS seedlings affected by HLB compared with the healthy one.F: The relative expression of GUS in CsCalS11p::GUS leaves treated with 100 μmol•L −1 ABA compared with H 2 O. G: GUS expression in the citrus leaves transformed with CsCalS11p::GUS through transient expression of CsABI5.NT: non-transgene seedling; P23/P57/P58: three citrus strains transformed with CsCalS11p::GUS.

Figure 7 .
Figure 7.The concentration of abscisic acid (ABA) and the expression of genes in its signal pathway in healthy and Huanglongbing (HLB) citrus.A: ABA concentration.B: Fold change of CsABI5, CsPYR, CsPP2C, and CsSnRK2 expression in citrus under HLB stress.* P ≤ 0.05.