MdAIL5 overexpression promotes apple adventitious shoot regeneration by regulating hormone signaling and activating the expression of shoot development-related genes

Abstract Adventitious shoot (AS) regeneration is a significant factor in the genetic transformation of horticultural plants. It is also a noteworthy approach to their vegetative propagation. AS regeneration remains highly dependent on the genotype or maturity of explants. We here found that the AS regeneration abilities of apple leaves were positively correlated with MdAIL5 expression. MdAIL5 overexpression dramatically increased AS regeneration efficiency. Notably, MdAIL5 overexpression could restore the AS formation ability of explants to a certain extent, which was lost with an increase in maturity. Endogenous hormone detection revealed that MdAIL5 overexpression changed the contents of auxin, cytokinin (CK), and other hormones in apple leaves. Transcriptome analysis revealed that many genes related to auxin, CK, and brassinolide signaling pathways were significantly and differentially expressed between MdAIL5-overexpressing transgenic apple and wild-type apple plants. Yeast one-hybrid assays, the electrophoretic mobility shift assay, and the dual-luciferase reporter assay revealed that MdAIL5 directly binds to MdARF9 and MdHB14 promoters and positively affects their expression. We here established a model of MdAIL5 regulating AS formation, which acts as a theoretical basis for facilitating genotype- or explant maturity-independent AS regeneration in the future.


Introduction
Plants undergo self-repair or their damaged tissues or structures are replaced during regeneration, and this process allows plants to adapt to the environment [1][2][3].Plant regeneration is based on totipotency or pluripotency, which ref lects the high f lexibility or plasticity of plant cells [4][5][6].Plant regeneration can be divided into somatic embryogenesis and de novo organogenesis [7].Among de novo organogenesis, adventitious shoots (ASs) play a crucial role in genetic transformation.Therefore, increasing research attention has been directed toward AS regeneration [8][9][10].
After nearly 2000 years of domestication and cultivation, apple has been cultivated in most countries and regions globally and is popular with consumers [11,12].With the release of several highquality apple genomes, molecular-assisted selection (MAS) technology has been gradually applied to apple breeding to increase its efficiency [11,13,14].Gene function research is the basis of MAS technology, and apple genetic transformation is the main method used in gene function research.The leaf disk method is mainly used to genetically transform apple, and therefore an efficient AS regeneration system dictates the success of this transformation [10,15].For many plants, including apple, AS formation efficiency is highly dependent on the genotype or explant maturity [10,[15][16][17][18][19].For example, 'Gala' and 'Jonagold', which are widely used in genetic transformation of apple, have relatively high AS regeneration efficiencies [20], whereas 'Fuji' has an extremely low AS regeneration efficiency, which makes successful genetic transformation difficult [21].Mao et al. [10] clarified the AS regeneration ability of several common apple rootstocks.They found that the AS regeneration efficiencies of Malus prunifolia (MP) and M26 was >4 times of that of T337 under the same culture conditions.Leaf maturity is another decisive factor for AS formation in apple.Generally, the greater the leaf maturity, the poorer the ability of cells to divide, the slower the growth, and the greater the difficulty of AS regeneration [18,19,22].The low transformation efficiency (0-3%) limited the application of transgenic and gene editing technologies in apple.
The interaction between auxin and cytokinin (CK) determines cell fate transitions during plant regeneration [23].A high CK concentration mediates the loss of root meristem characteristics in the callus, enabling the callus to gain shoot regeneration ability [23,24].Consistently, many genes related to CK biosynthesis and signaling pathways are involved in AS regeneration.The receptors of CK signaling, Arabidopsis cytokinin-receptor histidine kinases (AHKs), regulate AS regeneration by activating WUS expression while inhibiting WOX5 expression [25,26].Type-B ARR (B-ARR) and Type-A ARR (A-ARR) positively and negatively regulate the CK response, respectively [27].Correspondingly, inhibiting the expression of B-ARRs and overexpressing A-ARRs can reduce the AS regeneration efficiency [28].Cytokinin dehydrogenase/oxidase (CKX) enzymes irreversibly degrade the excess of CK to regulate its level [29].Accordingly, CKX1-overexpressing plants exhibit lower CK content and AS regeneration efficiency [30].In apple, MdWOX11 can downregulate MdCKX5 expression to suppress AS formation [10].Auxin also plays crucial roles in AS regeneration.The main mediators of auxin signaling, such as indole-3-acetic acids (IAAs), small auxin-upregulated RNAs (SAURs), and AUXIN RESPONSE FACTORs (ARFs), are key drivers of AS regeneration.For example, the IAA14 gain of function mutant is defective in callus formation and AS regeneration [31].ARF5 can directly bind SHOOT MERISTEMLESS (STM) and CYTOKININ RESPONSE FACTOR 2 (CRF2) promoters and activate their expression to promote callus formation and AS regeneration [32].The AS regeneration efficiency of zmsaur15 (maize SAUR15 deletion mutant) was five times higher than that of the wild-type (WT) plant [33].In addition, Brassinolide (BR), ethylene (ETH), and gibberellin (GA) signaling-related genes can also regulate AS formation, probably in connection with auxin and CK signaling [8,15,34].
Conventionally, hormone-based de novo organogenesis is cumbersome and time-consuming as well as relying heavily on the experience of operators [16].Many studies have recently unraveled the molecular basis of plant organogenesis and identified several regeneration-promoting key genes, thereby significantly improving the regeneration efficiency of many plant species.For instance, AtWUS overexpression greatly improved the AS regeneration efficiency of Coffea canephora, Gossypium hirsutum, and tobacco [35][36][37][38][39].The KNOX homeodomain transcription factor STM prevents the differentiation of meristematic cells [40].The transformation efficiency of maize STM homolog Knotted1 (ZmKn1)-overexpressing citrus increased 3-15 times.The transformation efficiency of ZmKn1-overexpressing tobacco increased 3 times.BBM, an AP2/ERF family transcription factor, can promote cell proliferation and ectopic embryo formation in plant somatic embryogenesis and organogenesis [41].Overexpression of native and heterologous BBM genes improves the transformation efficiency of various plant species [42][43][44][45].BBM belongs to the AIL family of transcription factors, which plays a critical role in plant regeneration.In a previous study, we analyzed the expression change of AIL family members during AS regeneration in apple, and the results revealed that AINTEGUMENTA-LIKE 5 (AIL5) (MD13G1252700) was most significantly upregulated, exceeding BBM1 [21].Phylogenetic analysis of Arabidopsis and apple AILs revealed that MdAIL20 (MD13G1252700) was the predicted ortholog of AtAIL5.To facilitate the functional analysis of MdAIL20 (MD13G1252700), we renamed MdAIL20 (MD13G1252700) as MdAIL5 in the current study.In Arabidopsis, AIL5 can induce the expression of the characteristic genes PLT1 and PLT2 of the root meristem and the characteristic factors CUP-SHAPED COTYLEDON 1 (CUC1) and CUC2 of AS regeneration, thus endowing the callus with AS regeneration ability [46].However, the effect of AIL5 on AS regeneration in apple needs to be studied further.
For the current study, we classified the AS regeneration ability of different apple cultivars and rootstocks.We further demonstrated that MdAIL5 overexpression promotes AS formation in apple.Changes were observed in the transcription level of MdAIL5-overexpressing (MdAIL5-OE) apple lines through RNA-seq and in the hormone content of these lines through liquid chromatography-tandem mass spectrometry (LC-MS/MS).Finally, we demonstrated that MdARF9 and MdHB14 act downstream of MdAIL5 to promote AS formation.Our results revealed the mechanism through which MdAIL5 promotes AS formation in apple and provide a theoretical basis for the further application of MdAIL5 to help overcome the recalcitrant transformation of apple.

Adventitious shoot regeneration ability varies for different apple genotypes
In the current study, we tested the AS regeneration ability of four major apple cultivars and seven apple rootstocks in the same regenerative medium (2 mg/l TDZ and 0.5 mg/l NAA).The AS increment coefficient, AS regeneration efficiency, and number of ASs per explant for 'Gl-3' and M26 were significantly higher than those for other cultivars and rootstocks.The AS regeneration ability of the tested apple cultivars was in the following order (high to low): 'Gl-3' > 'Gala' > GD > HF > 'Fuji' (Fig. 1a and b).The AS regeneration ability of the tested apple rootstocks was in the following order (high to low): M26 > T337 > Gm256 > B9 > Bp > 54-118 > 71-3-150 (Fig. 1a-c).In our previous study, we found that many MdAILs were upregulated during AS regeneration [21].We here detected the expression of the six most significantly upregulated MdAILs in the leaves of these cultivars and rootstocks after 5 weeks of regenerative culture.The results showed that MdAIL5 was most significantly upregulated during AS regeneration and its mRNA level and AS regeneration ability were positively correlated (Fig. 1b and c, Supplementary Data Fig.S1).Therefore, MdAIL5 was speculated to play a crucial role in apple AS regeneration.

MdAIL5 overexpression enhances adventitious shoot regeneration in apple
The subcellular localization of MdAIL5 was analyzed through transient expression in tobacco leaf epidermal cells.The f luorescence signal of MdAIL5-GFP was completely fused with that of the DAPI-stained nucleus, indicating that MdAIL5 was expressed and functional in the nucleus (Fig. 2a).To test the role of MdAIL5 in AS regeneration, we obtained seven MdAIL5-OE lines by using an Agrobacterium-mediated method and detected MdAIL5 mRNA overexpression in the leaves of the transgenic apple lines (Fig. 2d and e).
Lines L11, L12, and L19 with the highest MdAIL5 expression were selected for subsequent experiments (Fig. 2c).On the commonly used regeneration medium (2 mg/l TDZ and 0.5 mg/l NAA), visual evaluation indicated that the AS regeneration efficiency of the transgenic apple lines was clearly increased.However, because all MdAIL5-OE lines and WT plants had regenerated many shoots and these shoots were constantly propagating, conducting quantitative statistics was difficult.To perform quantitative statistics on AS regeneration ability, we further tested MdAIL5-OE lines and WT plants on MS medium with a low CK concentration (0.1 mg/L TDZ), and the AS regeneration phenotypes are presented in Fig. 2g.The AS increment coefficient, AS regenerative efficiency, and number of ASs per explant for L11, L12, and L19 were significantly higher than those for 'GL-3' apple (Fig. 2f and g).
AIL5 expression was significantly higher in young tissues than in mature tissues of Arabidopsis [47].Correspondingly, MdAIL5 expression in young apple tissues (f lower bud, 30-day-old leaf, current-year shoot, and fruit core) was significantly higher than that in mature apple tissues (f lower, 90-day-old leaf, perennial branch, and fruit f lesh) (Fig. 2b).Further research found that MdAIL5 expression in leaves decreased with an increase in subculture days of apple tissue culture seedlings (Fig. 2c).In addition to the genotype, the maturity of leaf explants is a crucial inf luencing factor for AS regeneration in apple [17].In the current study, the AS regeneration ability of apple leaf explants decreased with an increase in the maturity of the explants.AS regeneration was difficult for 'GL-3' tissue cultured seedlings at 75 days after subculture (DAS).Conversely, the inhibitory effect of increased maturity of leaf explants was slightly less in the MdAIL5-OE lines.The AS increment coefficient, AS regenerative efficiency, and number of ASs per explant for the MdAIL5-OE lines were clearly higher than those for the WT plants (Fig. 3).The data indicated that MdAIL5 overexpression can restore the AS regeneration ability of apple leaf explants to a certain extent, which was lost with an increase in maturity.
No remarkable morphological difference was observed between the MdAIL5-OE lines and non-transformed WT plants.However, some L11, L12, and L19 tissue-cultured seedlings produced numerous calli at the base of the seedlings and  spontaneously formed adventitious roots (ARs) at nearly 60 DAS (Supplementary Data Fig.S2a).We further tested the effect of MdAIL5 on AR formation.The AR increment coefficient, AR regenerative efficiency, and number of ARs per explant for the MdAIL5-OE lines were higher than those for the WT plants (Supplementary Data Fig.S2b).
Among CK compounds, the cZ and DL_DZ content was not different between the MdAIL5-OE lines and WT.The cZR, N6_iPR, and tZR content was lower in WT than in the MdAIL5-OE lines, and the cZR and tZR content was higher in L19 than in L12.Among auxin compounds, the IAA_Glu and I3CA contents were lower in WT than in the MdAIL5-OE lines, and the IAA_Glu content was higher in L19 than in L12.The IAA content was higher in L19 than in L12 and WT.The OxIAA content was not clearly different between the MdAIL5-OE plants and WT.The BR content was slightly lower in WT than in L12 and L19.The ABA content was higher in the MdAIL5-OE lines than in WT.Among GA compounds, the GA19 content was higher in WT than in the MdAIL5-OE lines.By contrast, the GA53 content was lower in WT than in the MdAIL5-OE lines.In conclusion, MdAIL5 overexpression causes significant changes in hormone content in apple.

MdAIL5 overexpression changes expression levels of hormone-and shoot development-related genes
Genome-wide gene expression changes between the WT plants and MdAIL5-OE lines (L12 and L19) were analyzed through transcriptome analysis (RNA-seq) (Supplementary Data Fig.S3).In total, 1965 differentially expressed genes (DEGs) were detected when the L12 and WT libraries were compared, including 1154 upregulated DEGs and 811 downregulated DEGs.The 1965 DEGs were also detected when the L19 and WT libraries were compared, including 2218 upregulated DEGs and 1489 downregulated DEGs.In total, 602 DEGs upregulated and 149 DEGs downregulated in L12 were also upregulated and downregulated, respectively, in L19.Then, we performed GO classification and KEGG enrichment analysis to compare the functional differences in these DEGs with the same expression trend in L12 versus WT and L19 versus WT (Supplementary Data Fig.S4).The GO term 'cellular anatomical entity' belonged to the maximum number of genes among both upregulated and downregulated DEGs.Moreover, the KEGG enrichment analysis revealed that these DEGs were significantly enriched in hormone-and shoot development-related pathways.
ARF genes in auxin signaling are crucial regulators of plant AS regeneration [32].In the present study, 12 upregulated ARF genes were detected in L12 versus WT and L19 versus WT.The expression levels of other auxin signaling-related genes, including AUX1, TIR1, AUX/IAA, GH3, and SAUR, were unchanged between the MdAIL5-OE lines and WT plants (Fig. 5a).In Arabidopsis, AIL/PLTs can induce YUCCA-mediated auxin accumulation.Song et al. [48] isolated 20 YUCCAs from the apple genome.Our transcriptome analysis revealed that three MdYUCCAs (MdYUCCA10c, MdYUCCA4b, and MdYUCCA3a) were significantly upregulated in both MdAIL5-OE lines (Supplementary Data Fig.S5).YUCCA encodes the rate-limiting enzyme for auxin biosynthesis.The increased auxin level in MdAIL5-OE line leaves may be due to the upregulation of these YUCCA genes.A-ARRs and B-ARRs positively and negatively regulate the CK response, respectively, and are involved in cell division and shoot initiation [27,49].Only upregulated B-ARRs were detected in transcriptome data, whereas no differentially expressed A-ARR was detected between the MdAIL5-OE lines and WT plants (Fig. 5b).ETR and CTR1 are negative and positive regulators of AS regeneration, respectively [50].The transcriptome analysis revealed that three ETR genes (MD13G1292200, MD12G1245100, and MD16G1212500) identified in the apple reference genome were clearly upregulated in the MdAIL5-OE lines.In contrast to ETR genes, five CTR1 genes (MD02G1122500, MD13G1266700, MD14G1121000, MD15G1236700, and MD16G1153700) were upregulated in the MdAIL5-OE lines (Fig. 5c).CYCDs are vital regulators of the plant BR response and positively regulate plant regeneration [51].Six upregulated CYCD genes (MD02G1178200, MD05G1087300, MD08G1092900, MD10G1095900, MD15G1077100, and MD15G1288100) were detected in the two MdAIL5-OE lines (Fig. 5d).Pluripotency acquisition, shoot promeristem formation, shoot progenitor formation, and shoot outgrowth are the four main stages of AS formation [34,[52][53][54].PLTs, WOXs, and SCRs act as main pluripotency factors.In the current study, the expression levels of MdAIL1 and 6, WOX4, 8, and 11, and SCR14, 22, 28, and 29 were clearly upregulated in both MdAIL5-OE lines.PINs are indispensable for shoot promeristem formation.The expression levels of PIN3, 4, 5, and 6 were clearly upregulated in both MdAIL5-OE lines compared with the WT plants.AIL/-PLTs regulate shoot-promoting CUC genes to establish the shoot promeristem in Arabidopsis [55].We identified 13 CUCs from the apple reference genome (Supplementary Data Fig.S6b).The transcriptome analysis demonstrated that the expression of all CUC genes was unchanged in L12 and L19 compared with WT (Supplementary Data Fig.S6a and c).HD-ZIP IIIs, SPLs, and WUS are the main regulators in the shoot progenitor stage.HB14, WUS, and SPB6 and 9 were significantly upregulated in the MdAIL5-OE lines.TCPs are indispensable for shoot outgrowth.TCP9 was upregulated in both MdAIL5-OE lines, whereas TCP4 and TCP20 were downregulated (Fig. 5e).Moreover, based our published RNAseq data, we analyzed the expression of hormone-and shoot development-related DEGs in apple leaves on 3, 7, 14, and 21 days after inoculation in the regeneration medium [21].The results showed that these DEGs exhibited significant differential expression during AS regeneration (Supplementary Data Fig.S7).

MdARF9 and MdHB14 are direct targets of MdAIL5 and are positively regulated by MdAIL5
To determine the target genes of MdAIL5, we analyzed whether the promoter regions of the aforementioned hormone-and shoot development-related DEGs contained MdAIL5-binding sites [57].The promoter regions of MdARF9 and MdHB14 respectively contained two different MdAIL5-binding sites (Fig. 6a).The binding of MdAIL5 to the MdARF9 and MdHB14 promoters was determined through the Y1H assay (Fig. 6b).The electrophoretic mobility shift assay (EMSA) analysis confirmed this direct binding (Fig. 6d and e).The effector (MdAIL5) and reporter (Luc controlled by the MdARF9 or MdHB14 promoter) were co-injected into tobacco leaves, and the result showed that MdAIL5 clearly enhanced Luc/Ren activity (Fig. 6c).

MdARF9 and MdHB14 overexpression enhances tobacco adventitious shoot regeneration
To analyze the function of MdARF9 and MdHB14 in AS regeneration, we generated MdARF9-OE and MdHB14-OE tobacco lines.No remarkable morphological changes were observed between the transgenic tobaccos and non-transformed WT plants.We detected MdARF9 and MdHB14 expression in eight randomly selected MdARF9-OE and MdHB14-OE tobacco lines, respectively (Fig. 7b and c).We then selected two transgenic tobacco lines with the highest expression levels of MdARF9 (Fig. 7b) and MdHB14 (Fig. 7c) to compare their AS regeneration efficiencies with those of the WT tobacco lines.The results showed that the number of ASs per explant, AS increment coefficient, and AS regenerative efficiency were clearly higher in the MdARF9-OE and MdHB14-OE tobaccos than in the WT plants (Fig. 7a and d-f).

Discussion
AS regeneration is a multistage developmental process.It includes the perception and transmission of plant hormone signals by somatic cells, initiation of cell division and proliferation, dedifferentiation with organ regeneration ability, and redifferentiation of organ formation [10,34,54].An efficient AS regeneration system is the main factor leading to successful genetic transformation [7,15].Furthermore, the improvement in the AS regeneration efficiency of apple has a potential application in improving the asexual propagation of apple seedlings.Based on the crucial applied values, many studies have analyzed the AS regeneration mechanism of model plants such as Arabidopsis.However, the genetic basis for apple regeneration remains unclear.
Genotype is considered among the most critical factors affecting AS regeneration [7,10,16,34].However, unified identification of the AS regeneration efficiency of different apple genotypes is lacking.Therefore, we here conducted a unified identification of the AS regeneration ability of several apple genotypes.Among the identified cultivars and rootstocks, 'Gala' and M26 exhibited the highest AS regeneration efficiencies (Fig. 1).In addition, the expression of MdAIL5, which belongs to the AIL family, positively correlated with AS regeneration ability (Fig. 1b and c).Maturity of leaf explants is another main factor affecting AS regeneration efficiency [22].The AS regeneration efficiency of apple decreased with increasing number of subculture days of tissue culture seedlings, and the MdAIL5 mRNA levels were negatively correlated with leaf explant maturity (Fig. 2c and g).Studies have shown that AIL mRNA levels in the young leaves and stems of Arabidopsis are considerably higher than those in mature tissues [47,58].This expression characteristic of AIL5 is consistent between apple and Arabidopsis (Fig. 2b).The aforementioned findings indicated that the MdAIL5 mRNA level is affected by genotype and explant maturity.MdAIL5 may be a key regulator of AS regeneration in apple.
Studies on AIL5 have focused on its function in the regulation of germination and seed maturation [59][60][61], and little is known of its role in AS formation.To explore the role of MdAIL5 in AS regeneration, we transformed the MdAIL5 overexpression vector into the 'GL-3' apple.Then, the AS regeneration efficiencies of higher-expressing MdAIL5-OE lines (L11, L12, and L19) and WT plants on a regenerative medium with a low CK concentration were calculated.The results revealed that the AS formation efficiencies of the MdAIL5-OE lines were clearly higher than those of the 'GL-3' apple (Fig. 2f and g), which indicated that MdAIL5 positively regulates AS regeneration in apple.Notably, MdAIL5 overexpression can restore the AS regeneration ability of apple leaf explants to a certain extent, which was lost with an increase in explant maturity (Fig. 3a and b).Studies have found that an increase in CK concentrations can restore the defect of the decline in the regeneration ability of adult plant leaves [22].The recovery of AS regeneration ability of MdAIL5-OE highly mature explants may be related to the increase in the endogenous CK content.
The balance and crosstalk of auxin and CK determine the developmental fate of plant cells during de novo organogenesis [7,8,23,34].LC-MS/MS revealed that the CK and auxin levels increased in the MdAIL5-OE lines (Fig. 4).Transcriptome analysis revealed that MdAIL5 overexpression changed the expression of hormone signaling pathway genes (Fig. 5a-d).ARFs, crucial mediators of auxin signaling, have key roles in AS formation.For example, ARF5 can transactivate STM and CRF2 promoters, thereby promoting callus formation and further improving the AS regeneration efficiency [32].ARF3 mutant explants could form a callus normally on CIM medium, but the number of ASs formed was significantly reduced when the calli were transferred to SIM [23].Moreover, miR160 regulates ARF10 expression, and the expression of main shoot regeneration genes, such as WUS, in transgenic explants expressing miR160-resistant ARF10 is significantly increased; the AS regeneration ability is significantly enhanced compared with the WT plants [34].Upregulated expression of ARFs in the MdAIL5-OE lines suggests that there may be downstream genes of MdAIL5 that promote AS regeneration in apple.B-ARRs, key regulators of the plant CK response, can directly bind to WUS and induce its expression, thereby significantly improving plant regeneration [27,49].In the current study, the mRNA levels of multiple B-ARRs in the MdAIL5-OE lines were clearly increased, indicating that MdAIL5 can enhance the sensitivity of leaf explants to CK.Overall, the increased auxin and CK biosynthesis and sensitivity may play a major role in promoting AS regeneration of MdAIL5-OE lines.
Additional phytohormones, including ETH and BRs, are also vital for AS formation.Three upregulated ETR genes and five downregulated CTR1 genes were detected in the MdAIL5-OE plants.Studies have found that ETH-related genes can positively and negatively regulate AS formation, determined by their role in ETH signaling.ETH-overproducing mutants and constitutive ETH response mutants exhibited an increased number of ASs, whereas ETH-insensitive mutants exhibited a decreased number of ASs in Arabidopsis [50].The upregulated ETR genes and downregulated CTR1 genes in the MdAIL5-OE lines were consistent with their positive and negative regulatory roles in AS formation, respectively.The CYCDs positively regulate cell division to promote plant regeneration [51].The upregulated CYCDs detected in the present study indicated that they are key regulators of AS formation.Studies in model plants have shown that AS regeneration can be classified into pluripotency acquisition, shoot promeristem formation, establishment of the confined shoot progenitor, and shoot outgrowth, and specific family genes have a major regulatory role in each stage [34,[52][53][54].Several genes involved in these stages, including MdAIL6, MdPIN6, MdHB14, and TCP9, were significantly and differentially expressed in the WT and MdAIL5-OE plants (Fig. 5e), which indicated that these genes act downstream of MdAIL5 to regulate AS regeneration at a specific stage.Arabidopsis AIL/PLTs promote AS regeneration through the activation of CUC genes [46].We detected no CUC DEGs between the MdAIL5-OE lines and WT plants (Supplementary Data Fig.S6).Therefore, differences are observed in the mechanism of AIL5-mediated regulation of AS regeneration between apple and Arabidopsis.
AR formation is necessary for the vegetative propagation of horticultural crops such as apple [9,62].The differentiation and elongation of phloem parenchyma cells around the vascular bundles in the stem are the basis for the formation of AR primordia, and high IAA concentrations can induce this process [63,64].In Arabidopsis, the wounding-induced IAA peak regulates AR formation by activating WOX11 expression [65].In apple, MdWOX11 promotes AR formation by inducing MdLBD29 expression [62].In the current study, the increase in IAA content in the MdAIL5-OE lines was detected through LC-MS/MS (Fig. 4).In addition, transcriptome sequencing analysis found that MdAIL5 overexpression induced WOX11 expression (Fig. 4e, Supplementary Data Fig.S6).These two points may be the main reasons for the enhanced AR regeneration ability in MdAIL5-OE plants.
Y1H, EMSA, and dual-luciferase assays suggested that MdAIL5 could directly bind to the MdARF9 and MdHB14 promoters and trigger their expression to promote AS regeneration in apple (Fig. 6).Furthermore, MdARF9 and MdHB14 significantly improved the AS regeneration efficiency through stable genetic transformation of tobacco (Fig. 7).ARFs can activate or repress auxin response elements on their promoters [66].Most studies on ARF9 have focused on the regulation of embryo and hypocotyl AR (HAR) formation.For example, Boutilier et al. [42] found that ARF9 can induce LBD16 and LBD28 expression by forming a protein complex with JMJ30, thereby promoting embryo development.ARF9 participates in the biogenesis of darkness-induced HARs by regulating ARF7 and ARF19 expression [67].This is the first study to demonstrate that ARF9 plays a crucial role in promoting AS regeneration in plants.MdHB14 belongs to HD-ZIP III.HD-ZIP III can bind to B-ARRs to form a transcriptional complex that upregulates WUS, thereby determining the spatial specificity of WUS expression [27].Based on these results, we speculate that MdAIL5 promotes AS formation through two pathways (Fig. 8).(I) MdAIL5 activates the expression of ARF, B-ARR, ETR, and CYCD3 and inhibits CTR1 expression to regulate crossing among different hormone signaling pathways, thus ensuring signal integration for proper AS regeneration.(II) MdAIL5 can upregulate the expression of some shoot development-related genes, including MdAIL6, MdSCR29, TCP9, WOX11, WUS, and PIN5, to regulate AS formation at different stages.The genotype and maturity of leaf explants are critical factors affecting AS regeneration in apple.In addition, studies have found that B-ARRs can regulate WUS expression with the HD-ZIP III transcription factor forming a protein complex [27].Thus, a relationship may exist between these two different regulation pathways, and this needs to be proven in our subsequent studies.Because AS regeneration is the crucial step of genetic transformation, clarifying the mechanism of MdAIL5-mediated promotion of AS regeneration lays a theoretical foundation for encouraging the application of genetic transformation in apple and other rosaceous fruit trees.

Quantification parameters of adventitious shoot regeneration ability of apple leaves
AS regeneration ability was quantified by estimating the AS increment coefficient, regenerative efficiency, and number of ASs per explant.The results of three experiments were analyzed, with each experiment conducted using 60 explants.Quantification parameters of AS regeneration ability were calculated as follows: AS increment coefficient = number of ASs/number of explants; AS regenerative efficiency (%) = (number of explants that regenerated ASs/number of explants) × 100%; and average number of ASs per explant = number of ASs/number of explants that regenerated ASs.

Construction of vectors and transgenesis
The 1668-, 3107-, and 2000-bp coding sequences (CDSs) of MdAIL5, MdARF9, and MdHB14 were amplified from the cDNA of 'Gala' leaves for the construction of the overexpression vector.Gene cloning was validated through sequencing.The CDSs were recombined into the plant expression vector PRI101 by using the ABclonal MultiF Seamless Assembly Mix (ABclonal Technology).Then, recombinant PRI101-MdAIL5, PRI101-MdARF9, and PRI101-MdHB14 were independently transformed into Agrobacterium tumefaciens EHA105.The primers are listed in Supplementary Data Table S1.
The genetic transformations of tobacco [70] and apple [12] are described elsewhere.'NC89' and 'GL-3' were used as the genetic backgrounds for tobacco and apple transformations, respectively.The regenerated transgenic shoots of tobacco and apple were screened with 100 and 25 mg/l kanamycin monosulfate (Kan), respectively, after coculturing with A. tumefaciens EHA105 containing the specific recombinant plasmid.To maintain the selection pressure, the Kan-resistant shoots obtained from tobacco and apple were transferred to fresh medium every 3 weeks.

Expression analysis by qRT-PCR
The primers for qRT-PCR were designed by Primer-BLAST software (Supplementary Data Table S1).cDNA was synthesized using the PrimeScript RT Regent Kit (TaKaRa).qRT-PCR was performed using TB Green Premix DimerEraser (TaKaRa) as described previously [21].MdActin (MD12G1140800) was used as the endogenous reference gene, and gene relative expression level was calculated using the 2 − method [71].

Hormone detection and analysis
The 100-mg samples for hormone extraction were harvested from the leaves of L11, L19, and WT.After the harvested sample was pretreated using a previously described procedure [72], the hormone was quantitatively detected through LC-MS/MS.In Multi-Quant software (Sciex, USA), the default parameters were used for automatic identification and integration of each MRM transition (ion pair), and manual inspection was assisted [73].

RNA-seq analysis
For transcriptome analysis, the first four apical expanding leaves of the tissue culture seedlings were collected for RNA extraction.Total RNA was extracted from three groups of leaves (WT, L12, and L19) by using the ethanol precipitation method and CTAB-PBIOZOL reagent, following the manufacturer's instructions.Illumina RNA sequencing was then performed using the BGISEQ500 platform (BGI-Shenzhen, China), following previously described methods.The gene expression levels were calculated using the FPKM method [74].Based on the criteria of fold change ≥2 and Q ≤ .05(adjusted P ≤ .05),DEGs were identified using the DEGseq package [56].The GO and KEGG enrichment analyses were performed using Phyper.

Subcellular localization
The CDS of MdAIL5 was recombined into PRI101-GFP by using the ABclonal MultiF Seamless Assembly Mix.The recombinant MdAIL5-pRI101-GFP was then transformed into A. tumefaciens GV3101 through heat shock.The resuspended solution was injected into the tobacco leaves by using a needle-free syringe and was cultured for 2 days before being placed on a glass slide.The resuspended solution was observed and photographed using a confocal scanning laser microscope (Leica TCS SP8, Leica, Germany).The primers are listed in Supplementary Data Table S1.

Figure 2 .
Figure 2. MdAIL5 overexpression enhanced AS regeneration in apple.a Subcellular localization of MdAIL5.Scale bar = 100 μm.b MdAIL5 expression levels in different apple tissues.c MdAIL5 expression levels of apple tissue culture seedlings on different subculture days.Error bars = standard deviation.The same lowercase letter indicates no significant difference at P < .05,Duncan's multiple range test.d DNA detection of MdAIL5-OE lines.e MdAIL5 expression levels in MdAIL5-OE lines and WT.f Quantification parameters of AS regeneration ability of MdAIL5-OE lines and WT.Error bars = standard deviation.* * P < .01(Student's t-test).g Morphology of AS regenerated from leaves of MdAIL5-OE lines cultured on MS + 0.1 mg/l TDZ compared with untransformed WT plants.Scale bar = 0.5 cm.

Figure 3 .
Figure 3. a Morphology of ASs regenerated from the leaves of MdAIL5-OE lines and WT plants on different subculture days.Scale bar = 0.5 cm.b Quantification parameters of AS regeneration ability of MdAIL5-OE lines and WT plants on different subculture days.Error bars = standard deviation.The same lowercase letter indicates no significant difference at P < .05(Duncan's multiple range test).

Figure 5 .
Figure 5. Expression of the phytohormone signal pathway and shoot development-related genes in MdAIL5-OE lines and WT. a Twelve ARF genes involved in auxin signaling were significantly downregulated in L12 and L19 versus WT. b Four B-ARR genes in the CK signaling pathway were significantly upregulated in L12 and L19 versus WT. c Three ETR genes and four CTR1 genes in ETH signaling pathways were significantly upregulated and downregulated, respectively, in L12 and L19 versus WT. d Six CYCDs in BR signaling were upregulated in L12 and L19 versus WT.Error bars = standard deviation.* Q < .05,* * Q < .01;Q value (adjusted P value) generated from DEG-seq [56].Red and blue represent upregulation and downregulation, respectively.e Heat maps of DEGs (FPKM ≥2) related to different AS development stages.Red and blue represent upregulation and downregulation, respectively.

Figure 6 .
Figure 6.MdARF9 and MdHB14 are target genes of MdAIL5.a Schematic representation of the MdAIL5 binding site in MdARF9 and MdHB14 promoters.b Yeast-one-hybrid assay.MdAIL5-pJG4-5 constructs introduced with MdARF9pro-pLacZi and MdHB14pro-pLacZi separately into yeast strain EGY48.c Luciferase reporter assay.Luciferase intensity was measured.Error bars = standard deviation.* * P < .01(Student's t-test).d MdAIL5 binds to the CACGCATCCCGAAG motif in the MdARF9 promoter.e MdAIL5 binds to the CACAATTGCCTATG motif in the MdHB14 promoter.

Figure 7 .
Figure 7. MdARF9 and MdHB14 overexpression enhanced tobacco AS formation.a Morphology of AS regenerated from the leaves of MdARF9-OE and MdHB14-OE tobaccos cultured on MS + 0.1 mg/l TDZ compared with non-transformed WT. b MdARF9 mRNA levels in MdARF9-OE tobaccos and WT.c MdHB14 mRNA levels in MdHB14-OE tobaccos and WT.d-f AS increment coefficient (d), AS regenerative efficiency (e), and number of ASs per explant (f) of MdARF9-OE tobaccos, MdHB14-OE tobaccos, and WT plants.Error bars = standard deviation.* P < .05,* * P < .01(Student's t-test).

Figure 8 .
Figure 8.A hypothetical model of MdAIL5 regulating AS.