Integrated Transcriptome and Metabolome Dynamic Analysis of Galls Induced by the Gall Mite Aceria pallida on Lycium barbarum Reveals the Molecular Mechanism Underlying Gall Formation and Development

Galls have become the best model for exploring plant–gall inducer relationships, with most studies focusing on gall-inducing insects but few on gall mites. The gall mite Aceria pallida is a major pest of wolfberry, usually inducing galls on its leaves. For a better understanding of gall mite growth and development, the dynamics of the morphological and molecular characteristics and phytohormones of galls induced by A. pallida were studied by histological observation, transcriptomics and metabolomics. The galls developed from cell elongation of the epidermis and cell hyperplasia of mesophylls. The galls grew quickly, within 9 days, and the mite population increased rapidly within 18 days. The genes involved in chlorophyll biosynthesis, photosynthesis and phytohormone synthesis were significantly downregulated in galled tissues, but the genes associated with mitochondrial energy metabolism, transmembrane transport, carbohydrates and amino acid synthesis were distinctly upregulated. The levels of carbohydrates, amino acids and their derivatives, and indole-3-acetic acid (IAA) and cytokinins (CKs), were markedly enhanced in galled tissues. Interestingly, much higher contents of IAA and CKs were detected in gall mites than in plant tissues. These results suggest that galls act as nutrient sinks and favor increased accumulation of nutrients for mites, and that gall mites may contribute IAA and CKs during gall formation.


Introduction
Plant galls are among the most emblematic examples of the manipulation and reprogramming of plant development produced by the gall-inducers [1]. There are many kinds of gall inducers in nature, such as bacteria, fungi, nematodes, mites and insects, but the most common gall inducers are insects, followed by mites [2]. The induction of galls is based on DNA transfer in some species of galling bacteria; however, the precise process underlying gall formation by arthropods is still unclear [3]. Research on the mechanism of gall induction in the case of arthropods has always focused on the gall inducers and suggests that gall formation is triggered by the action of chemical substances secreted by them, such as phytohormones [4] and other compounds [5][6][7]. However, the mode of action of these chemical compounds, and the general mechanism by which the gall inducer manipulates plant growth and physiology, are still largely unknown [8]. Furthermore, the chemical hypothesis is highly unlikely to provide a comprehensive explanation of the mechanisms of induction and morphogenesis of these structures due to the intricacy of the induction process and growth of plant galls [4]. The interaction between a plant and an herbivore involves complex networks of molecular and physiological processes within and

Alteration to L. barbarum Morphology and Population Dynamics of A. pallida
Most of the galls induced by A. pallida were on the leaves and showed a cystic shape, the color of which varied from yellowish green to green and then purple at different stages ( Figure 2A-C). The average gall diameter varied from 0.74 mm at 3 d to 2.92 mm at 30 d, and the rapid gall growth period was days 3~9, with a growth rate of 86.08%, followed by 12.62% from 12 d to 18 d and finally by 5.84% from 21 d to 30 d ( Figure 2D).
A. pallida can induce the gall by one or multiple mites, but always by a single one ( Figure 2E), and then live and reproduce inside it ( Figure 2F,G). The number of mites grew rapidly from 1.33 per gall to 353.33 per gall in 30 d. The increase was mainly concentrated within the first 18 d, with a compound growth rate (CGR) of 142.81% per 3 days but, after that, the CGR was only 33.65% per 3 d ( Figure 2H). Moreover, the egg number peaked at 18 d, with an average of 128.87 per gall, and gradually decreased thereafter. The egg number at 18 d was 6.67-and 0.20-fold higher than that at 6 d and 30 d, respectively ( Figure  2H).
There was a significant positive correlation between gall size and mite population size (r = 0.848, p < 0.01). According to the variation in gall size and mite population size, the galls used in the following study were divided into three developmental stages: (1) the induction stage, representing 3 d after mite infestation; (2) the growth stage, representing 9 d after mite infestation; and (3) the maturation stage, representing 18 d after mite infestation.

Alteration to L. barbarum Morphology and Population Dynamics of A. pallida
Most of the galls induced by A. pallida were on the leaves and showed a cystic shape, the color of which varied from yellowish green to green and then purple at different stages ( Figure 2A-C). The average gall diameter varied from 0.74 mm at 3 d to 2.92 mm at 30 d, and the rapid gall growth period was days 3~9, with a growth rate of 86.08%, followed by 12.62% from 12 d to 18 d and finally by 5.84% from 21 d to 30 d ( Figure 2D).
A. pallida can induce the gall by one or multiple mites, but always by a single one ( Figure 2E), and then live and reproduce inside it ( Figure 2F,G). The number of mites grew rapidly from 1.33 per gall to 353.33 per gall in 30 d. The increase was mainly concentrated within the first 18 d, with a compound growth rate (CGR) of 142.81% per 3 days but, after that, the CGR was only 33.65% per 3 d ( Figure 2H). Moreover, the egg number peaked at 18 d, with an average of 128.87 per gall, and gradually decreased thereafter. The egg number at 18 d was 6.67-and 0.20-fold higher than that at 6 d and 30 d, respectively ( Figure 2H).
There was a significant positive correlation between gall size and mite population size (r = 0.848, p < 0.01). According to the variation in gall size and mite population size, the galls used in the following study were divided into three developmental stages: (1) the induction stage, representing 3 d after mite infestation; (2) the growth stage, representing 9 d after mite infestation; and (3) the maturation stage, representing 18 d after mite infestation.

Histologic Structures of Galls Induced by A. pallida
Both the upper epidermal cell area of galls (G-EC) and the surrounding ungalled epidermal cell area (UG-EC) ( Figure 3A) showed an increasing trend over time, although no significant differences were found between them at any developmental stages ( Figure 3D). However, the upper epidermal cell length/width of galled tissue (G) was significantly greater than that of ungalled tissue (UG) beginning at 3 d, and the cell length/width of G-EC was 1.86-fold greater than that of UG-EC at 18 d ( Figure 3E). With gall development, the palisade and spongy parenchyma within the gall were gradually replaced by nutritive tissue (NT) near the mite feeding site and storage tissue (ST) far from the feeding site ( Figure 3B,C). The cells of NT were smaller than those of ST, their cytoplasm appeared to be denser than that of other cells within galls and their number showed a significant increase over time. The cell layers of galls (G-CL) increased significantly beginning at 9 d ( Figure 3B), while the cell layers of the surrounding ungalled tissues (UG-CL) showed no obvious change; compared with those in UG-CL, the cell layers in G-CL were 1.72-fold greater at 9 d and 2.82-fold greater at 18 d ( Figure 3F).

Alteration of L. barbarum Transcripts and Metabolites by A. pallida
In total, 97,302 transcript contigs were obtained with a total length of 6.57 G base pairs (bp) and an average length of 1003 bp. The GC content was 43.72%. After multiple filtering steps (see Section 4.4), 9246 unigenes were retained for final analysis. Principal component analysis (PCA) divided transcripts from the three infestation stages into two groups, the ungalled group and the galled group ( Figure 4A), indicating that the transcript profiles of ungalled and galled tissues were significantly different. The number of differentially expressed genes (DEGs) between ungalled and galled tissues increased gradually with gall development, with 3527, 4675 and 4958 DEGs identified between ungalled and galled tissues at 3 d, 9 d and 18 d, respectively ( Figure 4B). The number of upregulated DEGs was greater than that of downregulated DEGs at all developmental stages; the highest proportion of upregulated DEGs occurred at 9 d, and the ratio of upregulated/downregulated DEGs at 3 d, 9 d and 18 d was 1.68, 1.88 and 1.24, respectively ( Figure 4B). These DEGs were mainly involved in photosynthesis, cellular respiration, substance transportation, phytohormone action, cell division and cell wall organization, carbohydrates, amino acids, fatty acids and secondary metabolism pathways (Table 1). Most of the DEGs involved in photosynthesis were downregulated, while most of the DEGs involved in cellular respiration and transportation, and the metabolism of materials such as carbohydrates, amino acids and fatty acids, were upregulated (Table 1)  tissues at the same stage at the level of p < 0.05.

Alteration of L. barbarum Transcripts and Metabolites by A. pallida
In total, 97,302 transcript contigs were obtained with a total length of 6.57 G base pairs (bp) and an average length of 1003 bp. The GC content was 43.72%. After multiple filtering steps (see Section 4.4), 9246 unigenes were retained for final analysis. Principal component analysis (PCA) divided transcripts from the three infestation stages into two groups, the ungalled group and the galled group ( Figure 4A), indicating that the transcript profiles of ungalled and galled tissues were significantly different. The number of differentially expressed genes (DEGs) between ungalled and galled tissues increased gradually with gall development, with 3527, 4675 and 4958 DEGs identified between ungalled and galled tissues at 3 d, 9 d and 18 d, respectively ( Figure 4B). The number of upregulated DEGs was greater than that of downregulated DEGs at all developmental stages; the highest proportion of upregulated DEGs occurred at 9 d, and the ratio of upregulated/downregulated DEGs at 3 d, 9 d and 18 d was 1.68, 1.88 and 1.24, respectively ( Figure 4B). These DEGs were mainly involved in photosynthesis, cellular respiration, substance transportation, phytohormone action, cell division and cell wall organization, carbohydrates, amino acids, fatty acids and secondary metabolism pathways (Table 1). Most of the DEGs involved in photosynthesis were downregulated, while most of the DEGs involved in cellular respiration and transportation, and the metabolism of materials such as carbohydrates, amino acids and fatty acids, were upregulated (Table 1). A total of 1735 known compounds from 16 classes were detected in L. barbarum leaves by LC-MS/MS, of which 1111 compounds were detected under positive ion mode and 624 compounds were detected under negative ion mode. PCA divided metabolites from the three infestation stages into two groups, the ungalled group and the galled group, indicating that the metabolites from ungalled and galled tissues were significantly different; in addition, the difference between the ungalled group and galled group increased over time, indicating that the metabolic differences between ungalled and galled tissues widen with gall development ( Figure 4C). The numbers of differential metabolites (DMs) at 3 d and 9 d were similar, but both were smaller than that at 18 d; 612, 596 and 940 DMs were identified between ungalled and galled tissues at 3 d, 9 d and 18 d, respectively ( Figure 4D). The number of upregulated DMs was larger than that of downregulated DMs at all developmental stages; the highest proportion of upregulated DMs was observed at 18 d, and the ratio of upregulated/downregulated DMs was 1.10 at 3 d, 1.18 at 9 d and 1.76 at 18 d ( Figure 4D). These DMs could mainly be classified into eight groups associated with carbohydrates, amino acids, fatty acids, lipids, alkaloids, organic acids, terpenes and phenolics ( Table 2). The DMs from the carbohydrate, fatty acid, lipid, and organic acid groups were significantly upregulated at the three stages, the DMs from the amino acid, alkaloid and phenolic groups were significantly upregulated at 18 d, while the DMs from the terpene group were mainly downregulated (    In the column charts, red indicates the number of upregulated genes or metabolites, and blue indicates the number of downregulated genes or metabolites. The fold change threshold for the transcriptome and metabolome is log 2 > 1 or log 2 < −1.  Photosynthesis  152  159  185  7  12  13  145  147  172  Cellular respiration  69  70  65  68  68  61  1  2  4  Vesicle trafficking  35  63  61  35  63  58  0  0  3  Protein translocation  31  46  39  29  38  25  2  8  14  Phytohormone action  23  44  53  18  25  28  5  19  25  Cell division  50  72  77  46  62  57  4  10  20  Cytoskeleton organization  25  35  26  24  33  21  1  2  5  Cell wall organization  14  25  26  9  16  9  5  9  17  Carbohydrate metabolism  28  45  45  21  36  32  7  9  13  Amino acid metabolism  24  41  44  14  27  25  10  14  19  Protein biosynthesis  184  277  199  180  222  139  4  55  60  Lipid metabolism  54  77  95  39  53  60  15  24  35  Secondary metabolism  19  21  29  11  13 16 8 8 13 Combining the results for the transcriptome and metabolome, it could be seen that (1) the ungalled tissues and galled tissues differed distinctly and could be separated easily at both the gene and metabolite level and, (2) overall, the influence of the gall mite A. pallida on wolfberry strengthened with time, peaking at 18 d after infestation.

Manipulation of Hormone Balance and Cell Development in Galls by A. pallida
Cell division plays a significant role in gall development and the determination of gall shape [46]. A typical cell division cycle includes four phases: the G1 stage, S stage, G2 stage and M stage. In addition, cell wall formation is urgently needed for cytokinesis in plants [47], with the cell wall consisting mainly of polysaccharides such as cellulose, hemicellulose, pectin and several classes of proteins [30]. In the present study, the genes related to cell division were significantly upregulated in galled tissues over time ( Figure 5A), especially genes involved in the S and M stages. Cyclin-dependent kinase (CDK), which regulates cyclin synthesis, was also markedly upregulated during gall development ( Figure 5B). Furthermore, most of the genes involved in the components of the cell wall and its organization, such as lignin, cellulose, hemicellulose, pectin and cell wall protein synthesis, were distinctly upregulated in galled tissues at 3 d and 9 d ( Figure 5B).
Cell hyperplasia (cell division) and hypertrophy (cell growth) are the two main processes of gall initiation and development, which points toward the implication of AUXs (specifically IAA) and CKs during gall formation [4]. A limited number of IAA and CKs metabolism genes were annotated in this study, based on the known pathways for them. Most of the genes that regulate IAA synthesis were significantly downregulated in galled tissues ( Figure 5A) [48], including amidase 1 (AMI1), which was significantly downregulated at 9 d and 18 d, and aldehyde oxidase (AO) and flavin-dependent monooxygenase (YUC), which were significantly downregulated at all stages of development. Furthermore, the genes regulating IAA-amino acid conjugate synthase (GH3), which respond to a high concentration of IAA, were downregulated or showed no significance at 3 d and 9 d ( Figure 5C). The genes involved in CKs synthesis were not annotated in galled tissues, but genes regulating CKs interconversion [49], such as adenosine kinase (AK) and zeatin-O-glucosyltransferase (ZOGT), and genes encoding cytokinin dehydrogenase (CKX), which is involved in CKs degradation at high concentrations of CKs, were significantly upregulated in galled tissues ( Figure 5C). The pattern of GH3 and CKX indicated that IAA synthesis weakened and CKs degradation strengthened in galled tissues. Int. J. Mol. Sci. 2023, 24, x FOR PEER REVIEW 8 of 21 Cell hyperplasia (cell division) and hypertrophy (cell growth) are the two main processes of gall initiation and development, which points toward the implication of AUXs (specifically IAA) and CKs during gall formation [4]. A limited number of IAA and CKs metabolism genes were annotated in this study, based on the known pathways for them. Most of the genes that regulate IAA synthesis were significantly downregulated in galled tissues ( Figure 5A) [48], including amidase 1 (AMI1), which was significantly downregulated at 9 d and 18 d, and aldehyde oxidase (AO) and flavin-dependent monooxygenase (YUC), which were significantly downregulated at all stages of development. Furthermore, the genes regulating IAA-amino acid conjugate synthase (GH3), which respond to a high concentration of IAA, were downregulated or showed no significance at 3 d and 9 d ( Figure 5C). The genes involved in CKs synthesis were not annotated in galled tissues, but genes regulating CKs interconversion [49], such as adenosine kinase (AK) and zeatin- The AUX and CKs contents in ungalled tissues showed a decreasing trend over time, but most such contents in galled tissues were higher than those in ungalled tissues, and higher levels of AUXs and CKs were detected in the gall mite A. pallida than in plant tissues ( Figure 6). Only one type of AUX, IAA, was detected in both ungalled and galled tissues. The IAA concentration in ungalled and galled tissues showed a decreasing trend with gall development, but its concentration in galled tissues was significantly higher than that in ungalled tissues at 3 d (1.42-fold) and 18 d (1.82-fold). Six types of CKs were detected in ungalled and galled tissues, including IP, IPR, tZR, cZ, cZR and BAPR. No significant difference in the IP, IPR or cZ content was found between ungalled and galled tissues at 3 d and 9 d, but their contents in galled tissues were significantly (2.48-fold and 34.46-fold, respectively) higher than those in ungalled tissues at 18 d. Interestingly, cZ was not detected in ungalled tissues at 18 d. The cZR content in galled tissues was always higher than that in ungalled tissues, and significance was observed at 9 d and 18 d, being 1.61-fold and 36.11-fold higher than in ungalled tissues, respectively. The tZR content in galled tissues was also higher than that in ungalled tissues at all developmental stages; it was 4.45-fold higher than that in ungalled tissues at 3 d and was not detected in ungalled tissues at 9 d and 18 d. In addition, BAPR in galled tissues was detected only at 9 d and 18 d, when it was 0.07 and 2.84 ng·g −1 , respectively. Two types of AUXs (IAA, IBA) and ten types of CKs (IP, IPR, cZ, cZR, tZ, tZR, DHZ, BAP, BAPR, kinetin (K)) were detected in the gall mites. Most of their contents in mites were significantly higher than those in galled tissues except for BAPR, of which the IAA content was 1.69-fold, the IP content was 4826.56-fold, the IPR content was 28.99-fold and the cZ content was 592.93-fold higher than that in plant tissues. sues ( Figure 6). Only one type of AUX, IAA, was detected in both ungalled and galled tissues. The IAA concentration in ungalled and galled tissues showed a decreasing trend with gall development, but its concentration in galled tissues was significantly higher than that in ungalled tissues at 3 d (1.42-fold) and 18 d (1.82-fold). Six types of CKs were detected in ungalled and galled tissues, including IP, IPR, tZR, cZ, cZR and BAPR. No significant difference in the IP, IPR or cZ content was found between ungalled and galled tissues at 3 d and 9 d, but their contents in galled tissues were significantly (2.48-fold and 34.46-fold, respectively) higher than those in ungalled tissues at 18 d. Interestingly, cZ was not detected in ungalled tissues at 18 d. The cZR content in galled tissues was always higher than that in ungalled tissues, and significance was observed at 9 d and 18 d, being 1.61-fold and 36.11-fold higher than in ungalled tissues, respectively. The tZR content in galled tissues was also higher than that in ungalled tissues at all developmental stages; it was 4.45-fold higher than that in ungalled tissues at 3 d and was not detected in ungalled tissues at 9 d and 18 d. In addition, BAPR in galled tissues was detected only at 9 d and 18 d, when it was 0.07 and 2.84 ng·g −1 , respectively. Two types of AUXs (IAA, IBA) and ten types of CKs (IP, IPR, cZ, cZR, tZ, tZR, DHZ, BAP, BAPR, kinetin (K)) were detected in the gall mites. Most of their contents in mites were significantly higher than those in galled tissues except for BAPR, of which the IAA content was 1.69-fold, the IP content was 4826.56-fold, the IPR content was 28.99-fold and the cZ content was 592.93-fold higher than that in plant tissues.

Modification of Energy Metabolism in Galls by A. pallida
Chloroplasts and mitochondria are the two powerhouses of plant cells; the former is responsible for photosynthesis, and the latter is responsible for energy conversion [50]. In the present study, the genes related to photosynthesis were significantly downregulated in galled tissues at all developmental stages and showed an increasing inhibitory trend ( Figure 7A), including genes regulating chlorophyll synthesis and photosystem II (PSII), photosystem I (PSI), the cytochrome b 6/f complex (Cytb 6/f), ATPase and the Calvin cycle ( Figure 7B). However, genes regulating the tricarboxylic acid (TCA) cycle and oxidative phosphorylation were significantly upregulated in galled tissues during gall development ( Figure 7A), with citrate synthase (CSY), aconitase (ACO), oxoglutarate dehydrogenase (ODH) and succinate dehydrogenase (SDH), as well as four oxidoreductase complexes (I, III, IV, and V), being significantly upregulated ( Figure 7C). In addition, the genes that regulate transmembrane transportation were also significantly upregulated in galled tissues at all stages ( Figure 7A), including the genes related to carbohydrate, protein and ion transport ( Figure 7D). These results suggested that photosynthetic activity was suppressed, but cellular respiration and substance transport were enhanced, during gall development.
( Figure 7B). However, genes regulating the tricarboxylic acid (TCA) cycle and oxidative phosphorylation were significantly upregulated in galled tissues during gall development ( Figure 7A), with citrate synthase (CSY), aconitase (ACO), oxoglutarate dehydrogenase (ODH) and succinate dehydrogenase (SDH), as well as four oxidoreductase complexes (I, III, IV, and V), being significantly upregulated ( Figure 7C). In addition, the genes that regulate transmembrane transportation were also significantly upregulated in galled tissues at all stages ( Figure 7A), including the genes related to carbohydrate, protein and ion transport ( Figure 7D). These results suggested that photosynthetic activity was suppressed, but cellular respiration and substance transport were enhanced, during gall development.

Changes in Nutrient Metabolism in Galls Induced by A. pallida
Gall inducers can manipulate the source-sink dynamics of plants, manifesting as galled tissues usually containing high concentrations of nutritive substances such as sugars, proteins, lipids and other nutrients [9]. In this study, the genes involved in carbohydrate synthesis were significantly upregulated in galled tissues at all stages ( Figure 8A), including the genes regulating sucrose, glucose, mannose and starch synthesis and the genes involved in gluconeogenesis ( Figure 8B). The genes involved in nonaromatic amino acid synthesis (glutamate and aspartate groups) and the genes encoding aminotransferases in galled tissues were markedly upregulated during gall development ( Figure 8B). However, the genes involved in aromatic amino acid synthesis (shikimate group) in galled tissues were mainly downregulated ( Figure 8B). For metabolism, the abundance of carbohydrates and amino acids, and their derivatives in galled tissues, also significantly increased at all developmental stages (Figure 8A), and the alteration peaked at 18 d ( Figure 8C). Both the transcriptome and metabolome results showed that the nutrients in galled tissues increased with gall development.
including the genes regulating sucrose, glucose, mannose and starch synthesis and the genes involved in gluconeogenesis ( Figure 8B). The genes involved in nonaromatic amino acid synthesis (glutamate and aspartate groups) and the genes encoding aminotransferases in galled tissues were markedly upregulated during gall development ( Figure 8B). However, the genes involved in aromatic amino acid synthesis (shikimate group) in galled tissues were mainly downregulated ( Figure 8B). For metabolism, the abundance of carbohydrates and amino acids, and their derivatives in galled tissues, also significantly increased at all developmental stages ( Figure 8A), and the alteration peaked at 18 d ( Figure  8C). Both the transcriptome and metabolome results showed that the nutrients in galled tissues increased with gall development.

Defensive Responses of L. barbarum to A. pallida
Host plants initiate various defense responses against gall inducers when infested, including direct defenses (the accumulation of secondary metabolites) and indirect defenses (the release of volatile components) [9]. In the present study, most of the genes involved in the PPP pathway and defense response were significantly upregulated in galled tissues ( Figure 9A). The genes related to the PPP pathway were significantly upregulated in galled tissues at 9 d and 18 d, such as genes encoding phenylalanine ammonia

Defensive Responses of L. barbarum to A. pallida
Host plants initiate various defense responses against gall inducers when infested, including direct defenses (the accumulation of secondary metabolites) and indirect defenses (the release of volatile components) [9]. In the present study, most of the genes involved in the PPP pathway and defense response were significantly upregulated in galled tissues ( Figure 9A). The genes related to the PPP pathway were significantly upregulated in galled tissues at 9 d and 18 d, such as genes encoding phenylalanine ammonia lyase (PAL) at 18 d, cinnamate 4-hydroxylase (C4H) at 9 d and 18 d and 4-coumaroyl-CoA ligase (4CL) at 9 d ( Figure 9B). In addition, the genes that regulate flavonoid synthesis were markedly upregulated in galled tissues at three stages, including flavanone 3-hydroxylase (F3H) and anthocyanin 3-O-glucosyltransferase (3GT) ( Figure 9B). For metabolism, the abundance of flavonoids and coumarins significantly enhanced in galled tissues ( Figure 9A), and the alteration peaked at 18 d ( Figure 9C). lyase (PAL) at 18 d, cinnamate 4-hydroxylase (C4H) at 9 d and 18 d and 4-coumaroyl-CoA ligase (4CL) at 9 d ( Figure 9B). In addition, the genes that regulate flavonoid synthesis were markedly upregulated in galled tissues at three stages, including flavanone 3-hydroxylase (F3H) and anthocyanin 3-O-glucosyltransferase (3GT) ( Figure 9B). For metabolism, the abundance of flavonoids and coumarins significantly enhanced in galled tissues ( Figure 9A), and the alteration peaked at 18 d ( Figure 9C).

Discussion
This study systematically and comprehensively revealed the morphological dynamics of galls induced by the gall mite A. pallida, as well as the gall mite's population dynamics and impacts on host transcripts, metabolites and phytohormones during gall formation. The gall induced by A. pallida is a closed cystic gall with a structure of storage tissues outside and nutritive tissues inside, developed by the cell elongation of epidermal cells and hyperplasia of mesophyll cells. The galls grew quickly within 9 d, and the mite population size increased rapidly within 18 d. Although most of the genes involved in phytohormone synthesis were distinctly downregulated, IAA and CKs were greatly enhanced in galled tissues, which may be mainly due to the mites' contribution to phytohormones. The gall mite greatly influenced the chlorophyll synthesis and photosynthetic capacity of wolfberry but could enrich nutrients by transmembrane transportation, which was supported by the significant upregulation of most genes involved in the TCA cycle, oxidative phosphorylation process, transmembrane transportation and carbohydrate and amino acid synthesis, and the significant increase in the abundance of carbohydrates and amino acids, and their derivatives, in galled tissues. As a defense against the gall mite, wolfberry activated a defensive response and accumulated secondary metabolites, which were supported by the significant upregulation of most of the genes involved in the

Discussion
This study systematically and comprehensively revealed the morphological dynamics of galls induced by the gall mite A. pallida, as well as the gall mite's population dynamics and impacts on host transcripts, metabolites and phytohormones during gall formation. The gall induced by A. pallida is a closed cystic gall with a structure of storage tissues outside and nutritive tissues inside, developed by the cell elongation of epidermal cells and hyperplasia of mesophyll cells. The galls grew quickly within 9 d, and the mite population size increased rapidly within 18 d. Although most of the genes involved in phytohormone synthesis were distinctly downregulated, IAA and CKs were greatly enhanced in galled tissues, which may be mainly due to the mites' contribution to phytohormones. The gall mite greatly influenced the chlorophyll synthesis and photosynthetic capacity of wolfberry but could enrich nutrients by transmembrane transportation, which was supported by the significant upregulation of most genes involved in the TCA cycle, oxidative phosphorylation process, transmembrane transportation and carbohydrate and amino acid synthesis, and the significant increase in the abundance of carbohydrates and amino acids, and their derivatives, in galled tissues. As a defense against the gall mite, wolfberry activated a defensive response and accumulated secondary metabolites, which were supported by the significant upregulation of most of the genes involved in the defensive response process and PPP pathway, and by a notable increase in the relative abundance of flavonoids and coumarins in galled tissues. Therefore, it could be seen that the gall mite A. pallida manipulated its host plant via morphology and metabolism, with the potential to contribute to phytohormones (IAA and CKs) during gall induction.
The formation and development of galls usually involve a series of cytological changes, such as cell hyperplasia, cell hypertrophy and cell redifferentiation [2]. In the current study, the upper epidermal cell length/width and cell layers of galls significantly increased, and genes related to the cell cycle were distinctly upregulated in galled tissues, indicating that the formation of galls induced by the gall mite A. pallida involved cell division. Similarly, cell hyperplasia was also observed in galls induced by the psyllid Pseudophacopteron longicaudatum [30]. During gall information, cell wall fortification is commonly inhibited, and cell expansion is promoted [24]. The genes involved in cell wall organization were strikingly upregulated in galled tissues; this was conducive to cell expansion in galled tissues induced by A. pallida. Cell wall modification was consistently detected in galls induced by the nematode Meloidogyne incognita [51]. Since the stylets of gall mites mouthparts can insert into the cell wall mechanically, cut, drill and produce salivary secretions [37], the degradation of the cell wall may be closely related to the salivary secretions of A. pallida.
The common traits of galls with negative effects on photosynthetic capacity have been observed in several galling models [41,52,53]. In the present study, the genes involved in chlorophyll synthesis and photosynthesis were notably downregulated in galled tissues at each infestation stage, suggesting that the gall mite A. pallida markedly inhibited photosynthesis efficiency (especially the process of photophosphorylation). Similarly, the strong suppression of the genes encoding enzymes involved in the biosynthesis of photosynthetic pigments (carotenoids and chlorophylls) was also detected in galls formed by Daktulosphaira vitifoliae [33]. Obviously, the photoassimilates of galled tissues are barely sufficient to support the galls, and the nutrition required by the gall mite A. pallida might thus be translocated directly or indirectly from the adjacent ungalled tissues around galled tissues. This hypothesis was supported by the genes involved in transmembrane transportation being significantly upregulated in galled tissues; thus, the gall mite A. pallida realized a shift from autotrophy to heterotrophy during gall induction. This phenomenon was in accordance with the characteristics of gall induction by the psyllid Nothotrioza myrtoidis and the gall midge Bruggmanniella sp., although there was less chlorophyll and more polysaccharide, or soluble sugar, accumulation in galled tissues [14,41]. A few genes involved in the Calvin cycle were found to be significantly upregulated in galled tissues compared with ungalled tissues. Considering that there were no stomas in the inner galled tissues, this promotion of the Calvin cycle may be attributable to the CO 2 produced during gall mite respiration inside the gall, as shown in the galls of D. vitifoliae and Schlechtendalia chinensis [15,33].
AUXs and CKs are classes of phytohormones that induce cell growth and division, so the most general hypothesis is that gall formation is triggered by the action of phytohormones such as IAA and CKs [4]. Many recent works have shown that galled tissues are rich in IAA and CKs [28,40]. Exogenous CKs + IAA injections have been confirmed to lead to gall-like growth in Capsicum annuum [54]. In the current study, most of the genes involved in IAA synthesis were significantly downregulated in galled tissues, while genes regulating CKs conversion were significantly upregulated. In addition, the contents of IAA, tZR and cZR in galled tissues were dramatically higher than those in ungalled tissues at most infestation stages, indicating that higher contents of IAA and CKs existed in galled tissues during gall development. A similar pattern was also found in galls induced by Eurosta solidagini and Cicadulina bipunctata and leaves infested by the leaf-mining insect Phyllonorycter blancardella [11,27,55,56]. Active IAA and CKs have been identified in several gall inducers or their salivary secretions [35,57,58]. For example, Yamaguchi et al. [42] demonstrated that a galling insect was able to synthesize IAA from tryptophan. Thus, we reasoned that the enhanced content of IAA and CKs in galled tissues might have originated from the gall mite A. pallida. IAA can promote cell elongation and the maintenance of apical dominance by inhibiting the formation of lateral buds, and CKs can lead to cell division and promote the growth of plant tissues [59]; therefore, they might jointly contribute to the gall formation induced by A. pallida, in which IAA promotes epidermal cell elongation and CKs promote mesophyll cell division.
Gall inducers commonly change the chemical profile of their host plant tissues during gall induction and development [25]. In this study, the gall mite A. pallida needed a large amount of plant resources to develop and reproduce, and this demand for energy and carbon was reflected in the numerous genes involved in carbohydrate and amino acid synthesis being distinctly upregulated in the galled tissues at all developmental stages, which was also supported by the abundance of carbohydrates and amino acids and their derivatives being significantly enhanced in galled tissues compared with ungalled tissues. The same pattern was detected in flower galls induced by fig wasps, with dramatic upregulation of the transcripts associated with carbohydrate metabolism [17]. The genes involved in the PPP pathway were significantly upregulated in galled tissues at 9 d and 18 d, especially PAL, C4H and F3H enzymes, which play major roles in the biosynthesis of phenolic acids, such as flavonoids and coumarins [60], suggesting that the defense of wolfberry against A. pallida strengthened with infestation time, which was supported by the metabolism results: the abundance of flavonoids and coumarins prominently increased in galled tissues. Correspondingly, there were also much higher sugar, amino acid and phenol levels in the galls induced by the psyllid Pauropsylla depressa [61]. Some researchers have confirmed that tissues near the center of galls are usually rich in nutrient compounds, but that high levels of phenolics frequently accumulate outside of the gall [18,62], which could reduce the chemical defense of the host plants and facilitate the growth and development of gall inducers [25,31]. Phenolic acids might play roles in gall formation; for instance, flavonoids can act as transport regulators of IAA during gall formation [63], and coumarins (scopoletin) can act as IAA oxidase inhibitors [64]. Hence, we speculated that the elevated levels of phenolics in galled tissues of A. pallida not only play a defensive role but are also involved in the promotion of gall formation.

Plant and Mite Resources and Their Rearing Conditions
Seeds of L. barbarum were collected from fruits in Zhongning (37 • 29 N, 105 • 42 E), Ningxia Hui Autonomous Region, China, in 2020. The wolfberry seeds were sown in plastic trays (54 cm × 28 cm × 11 cm) containing potting soil media (Meihekou, Jilin Province, China) for germination and cultivated in growth chambers (25 ± 2 • C, 60 ± 5% relative humidity (RH), 14 L:10 D photoperiod, illumination: 18,000 l×). Approximately 6 weeks after sowing, every seedling with approximately 10 leaves was transplanted into a plastic cup (6.5 cm × 6.5 cm × 9 cm) and watered once every three days for continued cultivation under the abovementioned conditions. Approximately 2 weeks after transplanting, uniform seedlings with 15~20 leaves were selected for subsequent experiments.
The A. pallida used in this study were originally collected from L. barbarum in Zhongning (37 •

Preparation of Experimental Samples
A certain number of seedlings were prepared and divided into a control group and a treatment group. In the control group, no gall mites were inoculated. In the treatment group, the infestation of plants by gall mites was performed as follows: mature A. pallida galls (approximately 3 mm in diameter) cultured in the chamber were cut into 2 mm long and 0.05 mm wide slices (30~50 adult mites in each slice) under a stereomicroscope (Leica EZ4, Leica Microsystems, Wetzlar, Germany) [44], and then one slice was carefully attached to the tip of a seedling. Then, both the control and treatment groups were kept under the above rearing conditions.
The galls selected in this study were divided into 3 developmental stages according to the infestation time, gall size and mite population: (1) the induction stage, representing labeled D3C, D9C and D18C; the galls of leaves were dug out and labeled D3M, D9M and D18M.
was labeled D3C, D9C and D18C; the galls of leaves were dug out and labeled D3M, D9M and D18M.
Mites were collected from galls at 18 d after infestation. The galls were dissected under a stereomicroscope (Leica EZ4, Leica Microsystems, Wetzlar, Germany) and packed in a 50 mL centrifuge tube with sterile gauze, with approximately 15 g dissected galls per tube. To decrease the possibility of contamination and degradation, the tubes were injected with purified water, oscillated in an ice and water bath by ultrasound for 1 h, and subsequently centrifuged for 30 s at room temperature. Then, the upper water was absorbed by a micropipette (Eppendorf, Hamburg, Germany), and the lower sediment were the gall mites ( Figure 10D).
The cell area and cell layers were calculated and counted by Digimizer 6.0 software. There were 3 replicate slices for each treatment at each time point (3, 9 and 18 d after mite Mites were collected from galls at 18 d after infestation. The galls were dissected under a stereomicroscope (Leica EZ4, Leica Microsystems, Wetzlar, Germany) and packed in a 50 mL centrifuge tube with sterile gauze, with approximately 15 g dissected galls per tube. To decrease the possibility of contamination and degradation, the tubes were injected with purified water, oscillated in an ice and water bath by ultrasound for 1 h, and subsequently centrifuged for 30 s at room temperature. Then, the upper water was absorbed by a micropipette (Eppendorf, Hamburg, Germany), and the lower sediment were the gall mites ( Figure 10D).
The cell area and cell layers were calculated and counted by Digimizer 6.0 software. There were 3 replicate slices for each treatment at each time point (3, 9 and 18 d after mite infestation). To reduce the differences between individual leaves, the ungalled tissues surrounding galled tissues were taken as the control.

RNA Extraction, Transcriptome Sequencing and Analysis
Total RNA was isolated from frozen samples of plant tissues (3 replicates, 100 mg each) and gall mites (3 replicates, 30 mg each) using a TRIzol extraction kit (Tiangen Biotechnology, Beijing, China) following the manufacturer's instructions. The integrity and purity of RNA samples were assessed using an Agilent Bioanalyzer 2100 system (Agilent Technologies, Santa Clara, CA, USA) and a NanoPhotometer spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), respectively. RNA samples of acceptable quality were used to construct non-strand-specific sequencing libraries with the TruSeq RNA Sample Prep Kit (Illumina, San Diego, CA, USA). These libraries were sequenced using an Illumina NovaSeq 6000 by Novogene Co., Ltd., in Beijing, China (https://cn.novogene.com/ (accessed on 25 September 2022)).
Raw RNA-Seq reads were processed to remove adapters, reads containing more than two N bases and low-quality sequences using Trimmomatic [65]. De novo assembly of the clean sequence reads was carried out by Trinity v2.4.0 [66]. The transcript obtained by Trinity was taken as the reference sequence (Ref), the clean reads of each sample were mapped to Ref using RESM v1.2.15 [67], and then raw read counts for each transcript were derived. The read counts were imported into the R package DESeq2 v1.6.3 [68] for fragments per kilobase exon per million mapped fragments (FPKM) normalization. To exclude possible mite tissue contaminants from the galled tissue transcript files, a blast search of the protein sequences of galls and mites was carried out; subsequently, any matching sequences in the gall transcript file were removed. Finally, any transcript contigs with three replicates averaging <10 FPKM across all groups were excluded. Unigenes were annotated against NCBI nonredundant protein sequences (Nr) for function and against the Gene Ontology (GO) and Mercator4 protein function mapping databases for pathways [69].
DEGs were identified by the following criteria: absolute FPKM value more than 10, fold change (FC) between the treated group and control group greater than 2 (log2 ratio > 1 or < −1) and p value less than 0.05.

Metabolic Analysis by Liquid Chromatography Coupled to Tandem Mass Spectrometry (LC-MS/MS)
An amount of 100 mg plant tissue powder was individually ground with liquid nitrogen, and the homogenate was resuspended with prechilled 80% methanol (MeOH) by well vortex. The samples were incubated on ice for 5 min and then centrifuged at 15,000 r·s −1 and 4 • C for 20 min. Some of the supernatant was diluted to a final concentration containing 53% MeOH by LC-MS grade water. The samples were subsequently transferred to a fresh tube and then centrifuged at 15,000 r·s −1 and 4 • C for 20 min. Finally, the supernatant was injected into the LC-MS/MS system for analysis.
UPLC-MS/MS analyses were performed using a Vanquish UHPLC system coupled with an Orbitrap Q Exactive TM HF mass spectrometer (Thermo Fisher Scientific, Bremen, Germany). Samples were injected into a Hypesil Gold column (100 mm × 2.1 mm, 1.9 µm, Thermo Fisher Scientific, Bremen, Germany) using a 12 min linear gradient at a flow rate of 0.2 mL·min −1 . The eluents for the positive polarity mode were eluent A (0.1% formic acid in water) and eluent B (MeOH). The eluents for the negative polarity mode were eluent A (5 mM ammonium acetate, pH 9.0) and eluent B (MeOH). The solvent gradient was as follows: 0~1.5 min, from 2% to 2% B; 1.5~3 min, from 2% to 85% B; 3~10 min, from 85% to 100% B; 10~10.1 min, from 100% to 2% B; 10.1~12 min, from 2% to 2%. The mass spectrometer was operated in positive/negative polarity mode with a spray voltage of 3.5 kV, a capillary temperature of 320 • C, a sheath gas flow rate of 35 psi, an auxiliary gas flow rate of 10 L·min −1 , an S-lens RF level of 60, and an auxiliary gas heater temperature of 350 • C.
The raw data files generated by UPLC-MS/MS were processed using Compound Discoverer 3.1 (CD3.1, Thermo Fisher Scientific, Bremen, Germany) to perform peak alignment, peak selection, and quantitation for each metabolite. The main parameters were set as follows: retention time tolerance, 0.2 min; actual mass tolerance, 5 ppm; signal intensity tolerance, 30%; signal/noise ratio, 3; and signal intensity, minimum. After that, peak intensities were normalized to the total spectral intensity. The normalized data were used to predict the molecular formula based on additive ions, molecular ion peaks and fragment ions. Then, peaks were matched with the mzCloud (https://www.mzcloud.org/ (accessed on 20 September 2022)), mzVault and MassList databases to obtain accurate qualitative and relative quantitative results. Statistical analyses were performed using the statistical software R (version 3.4.3), Python (version 2.7.6) and CentOS (version 6.6). When data were not normally distributed, normal transformations were attempted using the area normalization method. using the MapMan4 visualization toolkit (https://mapman.gabipd.org (accessed on 10 September 2022)).

Conclusions
This paper provides a first glimpse into the dynamic responses of host plants to gall mites during gall induction and development at the transcriptomic, metabolic and phytohormonal levels simultaneously, based on laboratory samples. The current results indicate that: (1) the gall mite A. pallida induces galls via cell elongation of the epidermis and the division of mesophyll cells; (2) it transforms the photo assimilative capacity of L. barbarum leaves into a nutrient sink by transmembrane transport and by enhancing nutrient synthesis; and (3) it does not significantly alter the expression of the genes involved in IAA and CKs synthesis in L. barbarum, but has the potential ability to synthesize these compounds. The significant economic importance of gall mites is associated with their ability to transmit phytopathogens and induce galls on plants, but their minute size, and the problems related to protracted maintenance of these mites in culture, make them rather inconvenient objects for examination [37]. We argue that A. pallida + L. barbarum is an ideal model for studying the gall induction mechanism of gall mites due to the convenience of long-term rearing in the laboratory. In conclusion, this study will pave the way for future studies of gall induction by mites and plant biotic interactions.
Author Contributions: M.Y. performed research, analyzed the data and wrote the main manuscript text, H.L. helped rear the gall mites and wolfberry seedlings, H.Q. and K.G. contributed the idea of omics data analysis, R.X. and H.W. provided suggestions for the discussion's organization, J.W. provided ideas for abstract writing and title drafting and S.L. and C.X. designed the research and revised the manuscript. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement:
The raw data of transcripts that support the findings of this study are openly available in the NCBI Sequence Read Archive (SRA) under project PRJNA961943.