Abstract

Background

   The tea plant Camellia sinensis (L.) O. Kuntze is a perennial crop,
   invaded by diversity of insect pest species, and pink tea mite is one
   of the most devastating pests for sustainable tea production. However,
   molecular mechanism of defense responses against pink tea mites in tea
   is still unknown. In this study, metabolomics and transcriptome
   profiles of susceptible and resistant tea varieties were compared
   before and after pink tea mite infestation.

Results

   Metabolomics analysis revealed that abundance levels of
   polyphenol-related compounds changed significantly before and after
   infestation. At the transcript level, nearly 8 GB of clean reads were
   obtained from each sequenced library, and a comparison of infested
   plants of resistant and susceptible tea varieties revealed 9402 genes
   with significant differential expression. An array of genes enriched in
   plant pathogen interaction and biosynthetic pathways of
   phenylpropanoids showed significant differential regulation in response
   to pink tea mite invasion. In particular, the functional network
   linkage of disease resistant proteins, phenylalanine ammonia lyase,
   flavanone -3-hydroxylase, hydroxycinnamoyl-CoA shikimate transferase,
   brassinosteroid-6-oxidase 1, and gibberellin 2 beta-dioxygenase induced
   dynamic defense signals to suppress prolonged pink tea mite attacks.
   Further integrated analyses identified a complex network of transcripts
   and metabolites interlinked with precursors of various flavonoids that
   are likely modulate resistance against to pink tea mite.

Conclusions

   Our results characterized the profiles of insect induced metabolic and
   transcript reprogramming and identified a defense regulatory network
   that can potentially be used to fend off pink tea mites damage.

Supplementary Information

   The online version contains supplementary material available at
   10.1186/s12864-024-10877-z.

   Keywords: Pink tea mite, Flavonoids, Resistance, Differential genes,
   Tea production

Background

   Tea, made from the young leaves, is the most popular non-alcoholic
   beverage in the world. The economies of several countries are based on
   perennial monoculture tea plantation [[38]1]. China, the first
   originator as well as the world'sleading tea producer, has been
   consuming tea as a medicine and beverage since at least 2700 BC
   [[39]2]. When fermented with variousmedicinal minerals, the
   biologically active components of tea leaves, such as polyphenols,
   minimize the detrimental effects on human health [[40]3, [41]4]. Tea
   began to be cultivated in other regions of the world in the
   mid-seventeenth century. It is currently cultivated commercially
   throughout Asia, Africa, and South America in tropical, subtropical,
   and temperate climates, as well as in a few selected locations in North
   America, Australia, and Europe [[42]5]. However, China, India, Kenya,
   Sri Lanka, Vietnam, Turkey, Indonesia, and Japan are the top 8
   producers of tea in the world, accounting for 88.9% of the total
   [[43]6]. This intensive tea cultivation leads to pressure from many
   diseases, insects, and pests. In addition, the distinctive
   characteristics of the perennial tea crop have a highly specific impact
   on the ecology of its native insect pests [[44]7]. In most tea-growing
   countries, changing temperature and rainfall patterns, together with
   the perennial growth pattern of the tea plant, results in a relatively
   consistent microclimate that provides an ideal habitat for several
   insect pests [[45]8]. It has been reported that the tea plant is
   affected by several hundred bacterial, fungal, and nematode diseases
   and 250 insect species insect pests [[46]7, [47]9]. Like other crops,
   tea leaves are frequently damaged by a wide range of herbivorous mites.
   The major species of tea mites include the tea kanzawai mite
   (Tetranychus kanzawai (Kishida), the pink tea mite (Acaphylla theae)
   (Watt), the scarlet tea mite (Brevipalpus obovatus) (Donnadieu), the
   red spider tea mite (Oligonychus coffeae) (Nietner), and the scarlet
   mite (Brevipalpus phoenicis) (Geijskes). Among these, Acaphylla theae
   (Watt) and Polyphagotarsonemus latus (Banks) consume the tender leaves
   of tea plants [[48]9, [49]10]. Severe infestation of the pink tea mite
   causes rusting on the back of the leaves, while P. latus causes brown
   streaking and roughens of the underside of the leaves [[50]11].

   The first outbreak of the pink tea mite was observed in India and then
   reported in various cultivated regions of China, Japan, and the United
   States [[51]12]. The infestation of this mite is now found in the four
   major tea producing areas of China especially in cool regions located
   at high altitudes [[52]13]. The previous field experiments showed that
   the most suitable host of pink tea mite is tea, so it can be considered
   as the most destructive insect pest that worsens the physiological and
   biochemical changes in tea leaves [[53]14]. However, the extent of pink
   tea mite infestation is probably related with weather type, percentage
   of rainfall, and light penetration [[54]15]. The prolonged feeding of
   the mite on the phloem sap of tender buds and leaves results in stunted
   plant growth, leaf curling and leaf drop. According to the reports, the
   mite outbreaks have serious drawbacks, reduced yield, and damaged tea
   quality traits in the subtropical and tropical tea growing regions,
   where yield losses are typically between 10 and 20% [[55]16, [56]17].
   Mite controlin tea is challenging because the use of synthetic
   pesticides leads to new pest outbreaks and resistance development.. In
   addition, the small size and and tendency of mites facilitate
   widespread breeding, and they don't appear until significant damage has
   been done to the target leaf. Because of these disturbing
   characteristics,alternative control methods and the adoption of
   integrated pest management are becoming essential to control the pink
   tea mite in the world without potential threats to the environment.

   Host-plant resistance or plant defense against insect pest herbivores,
   is regulated by a diverse range of phenotypic plasticity depending on
   the type of damage. The initial defense responses to herbivore attack
   consists of leaf surface waxes, morphological traits such as trichomes,
   and various secondary metabolites [[57]18, [58]19]. However, tea plants
   respond to herbivores by producing and emitting volatile chemicals that
   appear to be effective in further reducing the induction of pest
   feeding [[59]20, [60]21]. Recent evidence has revealed a complex
   defense mechanism that significantly ameliorates responses and is
   mediated by several interconnected phytohormone signaling pathways. In
   particular, jasmonic acid, salicylic acid, abscisic acid, gibberellin,
   and brassinosteroid mediate tea plant defense against leaf herbivores
   [[61]22, [62]23]. The contribution of these hormones to defense
   pathways depends on feeding mode as well as pest species. In the
   indirect mechanism, the various species of flavonoid compounds,
   glucosinolates, phenolics, alkaloids, and terpenoids not only reduced
   the stress caused by insect pest invasion but also triggered a cascade
   of signalling molecule that is essential to for reducing leaf damage in
   tea plants [[63]24]. The integrated approach with metabolomics and
   transcriptomic data has recently been widely applied to reveal novel
   aspects of plant–insect interactions in many agronomic crops
   [[64]25–[65]27]. For example, recent studies in tea revealed potential
   genes and biochemical components associated with resistance toEctropis
   oblique [[66]28] and green leafhopper attack [[67]24]. However, limited
   molecular research has been conducted on pink tea mite infestation.
   Characterizating the key transcriptional changes involved in the
   biosynthesis of defensive metabolites has practical significance in
   dissecting the interactions between the pink mite and the tea plant.
   Therefore, this study was conducted to elucidate the resistance and
   susceptibility of tea to the pink tea mite. The deviating effects of
   mite on tea production were determined in the field. Furthermore, the
   control and infested samples of resistant and susceptibility tea
   verities were used for metabolomics and transcriptomic analyses. This
   study identifies the network of potential genes and metabolites
   associated with defense responses against the pink tea mite. These
   datasets can be used to limitthe damage caused by the pink tea mite.

Results

Phenotypic variations of resistant and susceptible tea varieties against
Acaphylla theae Watt

   Two early-maturing tea varieties selected from Wenzhou Huangyezao were
   used in this study to investigate the prevalence and damage caused by
   the pink tea mite. A comparison of field tests showed that Guilu 1 and
   Pingyang Tezao had significantly different responses to the pink tea
   mite. In Pingyang Tezao, the mite population invasion was significantly
   lower (Fig. [68]1A). Notably, the percentage of mite populations per
   bud and damage level was significantly (p < 0.001) higher in Guilu 1
   than Pingyang Tezao (Fig. [69]1B-C). To further determine the
   difference in resistance level, pink tea mite feeding induction traits
   were carried out, and evaluation showed that pink tea mites had more
   adverse effects on tea leaves and buds in Guilu 1 than Pingyang Tezao
   (Fig. [70]2A). In particular, the fresh weight of one hundred buds
   (Fig. [71]2B), bud length (Fig. [72]2C), and leaf area (Fig. [73]2D) of
   Guilu 1 were significantly (p < 0.001) reduced after mite infestation
   compared with the control, while the infested plants had a higher ratio
   of dry to wet biomass than the control (Fig. [74]2E). In contrast, both
   control and infested plants of Pingyang Tezao showed no significant
   difference in fresh weight of one hundred buds (Fig. [75]2B) as well as
   leaf area (Fig. [76]2D). It showed a significant difference (p < 0.001)
   for traits such as bud length and the ratio of dry to wet biomass.
   These results confirm that Guilu 1 has a higher infestation and
   extremely low resistance to the pink tea mite, whereas Pingyang Tezao
   shows a lower infestation and is more resistant to the pink mite. These
   two tea varieties are suitable for studying the potential resistance
   mechanisms associated with pink tea mites.

Fig. 1.

   [77]Fig. 1
   [78]Open in a new tab

   Comparison of field surveillance of pink tea mite invasion in
   susceptible (Guilu No. 1; GL) and resistant (Pingyang Tezao; PYTZ) tea
   varieties (A) Representative images of pink tea mite population on a
   single plant of both varieties were examined using a Nikon SM2800
   stereomicroscope (B) Total number of mite occurrences per bud. Error
   bars represent (mean ± SEM) and *** represents statistical significance
   at p < 0.001 as determined by Student's t-test (C) The level of damage
   (mean ± SEM) in both varieties as determined from 100 plants and
   independent experiments and detailed information for this damage
   grading can be found in methods

Fig. 2.

   [79]Fig. 2
   [80]Open in a new tab

   The detrimental effect of pink tea mite infestation on one bud and two
   leaves in susceptible (Guilu No. 1; GL) and resistant (Pingyang Tezao;
   PYTZ) tea varieties (A) Representative phenotypic images of control and
   infested plant samples (B) Fresh weight of 100 buds (g) (C) Bud length
   (D) Leaf area (E) Fresh and dry weight ratio. Different letters
   indicate statistical significance at p < 0.001 as determined by the
   least significant difference test

Metabolite profiling and quantification of potential defense metabolites

   Plants respond to insect damage by inducing certain defense-related
   metabolites. To elucidate the role of metabolites in defense systems
   against pink tea mite, metabolomics analysis was performed on control
   and mite-infested plants of resistant and susceptible tea varieties.
   Principal component analysis based on metabolic profiling revealed
   observable variance in both tea varieties before and after mite
   infestation (Supplementary Figure S1A). In addition, a weak correlation
   was observed between both cultivars (Supplementary Figure S1B). A total
   of 2180 metabolites were identified belonging to different types of
   alkaloids, phenolic acids, flavonoids, lipids, organic acids, amino
   acids and derivatives, nucleotides and derivatives, lignans and
   coumarins, terpenoids, tannins, and quinone. However, flavonoids
   (24.1%), phenolic acids (16.5%), and alkaloids (8.8%), followed by
   lipids (7.8%), were the major metabolites (Supplementary Figure S1C).
   Notably, most of the identified metabolites showed qualitative and
   quantitative differences before and after the infestation of the pink
   tea mite (Supplementary Figure S1D). Apparently, most of the detected
   metabolites associated with flavonoids showed a relatively higher
   abundance level in response to mite infestation in resistant varieties
   compared with susceptible varieties, while the most of the metabolites
   associated with phenolic acids showed a relatively lower abundance
   level (Supplementary Figure S1D).

   The comparisons of all detected metabolites were performed to identify
   potential differences among key metabolites likely to be involved in
   the resistance mechanism against the pink tea mite. Many metabolites
   showed significant differences between resistant and susceptible tea
   varieties during mite feeding and control periods. For example,
   metabolite differences before and after mite infestation were 472 in
   susceptible varieties. Among these, most had a lower abundance level
   before mite infestation (Supplementary Figure S2A), and clovanmagnolol,
   liquidambaric lactone, and lindleyin had higher differences
   (Supplementary Figure S2B). There were 401 metabolites with differences
   before and after mite infestation in the resistant variety. Most of
   these had a higher abundance level after mite infestation
   (Supplementary Figure S2C), and (E)-resveratroloside,
   salicyloylsalicin, and 5-geranyloxy-1,3-dihydroxyxanthone had higher
   differences (Supplementary Figure S2D). A total of 680 metabolites
   showed a significant a significant difference in abundance between the
   infested plants of both tea varieties (Fig. [81]3A). The 350
   metabolites showed a higher abundance, while the 330 had a lower
   abundance level in the resistant than in the susceptible tea variety.
   In addition, there was a relatively higher number of specific
   differential metabolites rather than overlap in each combination (Fig.
   [82]3B). In particular, there were 115 metabolites that specifically
   showed differential abundance in response to mite infestation in
   resistant varieties compared to susceptible varieties. These suggest
   that both varieties synthesized different biochemical components before
   and after pink tea mite infestation. In this direction, kmeans
   clustering divided 1118 metabolites into six groups according to their
   differential level in each variety (Fig. [83]3C). The 274 metabolites
   of group 3 had a higher abundance level in infested samples of
   susceptible tea varieties, while the 201 metabolites of group 4 had a
   higher abundance level in infested samples of resistant tea varieties.
   In addition, metabolites from groups 1 and 5 had a relatively similar
   abundance levels among infested plants of both varieties, but group 5
   had a lower abundance while group 1 had a higher abundance of
   metabolites before mite infestation. The other two groups had similar
   abundance levels before and after infestation in each variety. However,
   their accumulation was different between resistant and susceptible tea
   varieties. The putative key metabolites that are most likely mediate
   resistance against pink tea mite include secondary metabolites related
   to flavonoid biosynthesis, flavone and flavonol biosynthesis,
   isoflavonoid biosynthesis, monoterpenoid biosynthesis, and
   phenylpropanoid biosynthesis-based phenolic acids. Among the key
   flavonoid-related compounds with the highest differential level, most
   species of quercetin-3-O related compounds, kaempferol-3-O related
   compounds, and myricetin biosynthesize in greater abundance level
   differences in resistance than susceptible tea varieties (Fig. [84]3D).
   This predicts their pivotal role in minimizing the deleterious effects
   of pink mites in tea as well as in resistance to pink tea mites. In
   addition, many other complex chain flavonoidmetabolites species of
   apigenin-6/7/8-C-glucosides, chrysoeriol-6-C-glucosides,
   eriodictyol-7-O-glucosides, hispidulin-8/6-C-glucosides,
   luteolin-7-O-glucosides, and vitexin-2/7-O-galactosides had higher
   accumulation during mite invasion (Supplementary Figure S3). In
   addition, several phenolic acids, including 5-O-caffeoylshikimic acid,
   4-O-cinnamoyl-5-O-caffeoylshikimic acid,
   2-O-salicyl-6-O-galloyl-D-glucose, 1-O-salicyloyl-β-D-glucose,
   syringing, and hydroquinone, were the major phenolic components
   detected in higher abundance. In contrast, epicatechin
   3',4,5-digallates related flavonoids along with phenolic acids such as
   2'-hydroxydaidzein, 3-O-digalloyl quinic acid,
   7,3'-di-O-gallyoltricetiflavan, 1,2,3,6-tetra-O-galloyl-β-D-glucose,
   and 5-galloylshikimic acid were identified at lower levels during pink
   tea mite feeding damage (Supplementary Figure S3). In summary,
   metabolic profiling revealed that the composition and abundance of
   various flavonoids and phenolic compounds are key regulators of defense
   against pink tea mite.

Fig. 3.

   [85]Fig. 3
   [86]Open in a new tab

   Characterization of differentially accumulated metabolites in
   comparison of susceptible (Guilu No. 1; GL) and resistant (Pingyang
   Tezao; PYTZ) tea varieties before and after pink mite infestation (A)
   Statistics of differential metabolites among different comparisons (B)
   Distribution of differential metabolites (C) Cluster analysis of
   differential metabolites (D) Fold change values of the key
   differentially accumulated metabolites identified among comparisons of
   susceptible (Guilu No. 1; GL) and resistant (Pingyang Tezao; PYTZ) tea
   varieties during pink mite infestation

Transcriptome sequencing and dynamic gene expression differences among
resistant and susceptible tea varieties

   In order to identify the potential genes likely to confer resistance to
   pink tea mite, comparative transcriptome analysis was performed on
   control and infested plants of both resistant and susceptible tea
   varieties. Approximately 8 GB of clean reads were obtained from each
   sequenced library with a very low error rate of 0.02% and a Q30%
   of > 93%. A total of 87% of clean reads were mapped on the tea
   reference genome. Of these, 73% were uniquely mapped, 13% were
   multi-mapped, and almost 36% were mapped to each strand of DNA
   (Supplementary Table S1). Further quantification confirms that
   comparatively both varieties showed different responses to pink tea
   mite infestation, but infected and control samples of each cultivars
   had less variance. The PC1 and PC2 together exhibited a 48.14% variance
   in the transcript datasets (Supplementary Figure S4A), as well as a
   different level of relationship among transcriptome data for both tea
   verities (Supplementary Figure S4B). There were large-scale transcript
   modifications during prolonged periods of mite infestation in each
   variety. The pairwise comparative analysis identified less than 2300
   differentially expressed genes (DEGs) among comparisons of control vs.
   infested plants in susceptible tea varieties (Fig. [87]4A). The
   majority of these DEGs had a lower abundance level before mite
   infestation (Supplementary Figure S4C), and genes such as novel.12103,
   novel.6545, CSS0045784, CSS0037009, and CSS0022909 exhibited higher
   fold change differences (Supplementary Figure S4D). However, 1591 DEGs
   were identified in control vs. infested plants of the resistant tea
   variety. Among these, most DEGs had a lower abundance level after mite
   infestation. (Supplementary Figure S4E). For example, CSS0042297 and
   CSS0008361 had higher fold change differences after mite infestation
   (Supplementary Figure S4F). In addition, the control plants of both
   varieties had 8895 DEGs, of which, 4286 genes showed down regulation
   and 4609 genes showed up regulation in the resistant variety. Notably,
   the infested plants of resistant and susceptible tea varieties had a
   relatively higher number of DEGs. There was a total of 9402 DEGs in
   this comparison, of which 4731 were downregulated and 4671 were
   upregulated in resistant plants. This means that the prolonged invasion
   of the pink tea mite causes dynamic transcript changes in tea.

Fig. 4.

   [88]Fig. 4
   [89]Open in a new tab

   Characterization of differentially expressed genes (DEGs) in
   comparisons of susceptible (Guilu No. 1; GL) and resistant (Pingyang
   Tezao; PYTZ) tea varieties before and after pink mite infestation (A)
   Statistics of DEGs among different comparisons (B) Distribution of DEGs
   (C) Cluster analysis of DEGs (D) The expression heatmap potential
   defense-related DEGs in susceptible (Guilu No. 1; GL) and resistant
   (Pingyang Tezao; PYTZ) tea varieties before and after pink mite
   infestation

   Further analysis of specific and overlapped DEGs showed a limited
   number of overlapped DEGs in comparisons of infested plants of both tea
   varieties. For example, 2318 DEGs were specific in this comparison,
   while the comparisons of control and infested plants had 351 DEGs
   specific to resistant and 682 DEGs specific to susceptible tea
   varieties (Fig. [90]4B). These DEG distribution landscapes indicate
   that post pink tea mite infestation effects are not similar in
   resistant and susceptible tea varieties. K-means clustering divided the
   genes into three distinct groups based on the expression patterns
   before and after mite infestation in each variety (Fig. [91]4C). In
   group 1, the expression profiles of 4724 genes had relatively similar
   expression dynamics before and after infestation in each variety but
   were significantly higher in susceptible than resistant groups. In
   contrast, in group 2, 6179 DEGs had relatively stable expression
   dynamics before and after infestation in each variety, but
   significantly lower expression in susceptible than resistant varieties.
   The 2389 DEGs of group 3 showed similar expression profiles in control,
   but their expression was highest in infested plants of susceptible
   rather than resistant varieties. Expression profiles of DEGs further
   confirmed that the attack of pink tea mite infestation produced higher
   quantitative gene expression differences in susceptible tea variety
   infested plants than control, whereas resistant variety showed less
   quantitative gene expression differences among control and infested
   plants (Fig. [92]4D). It was found that an array of DEGs with
   differential regulation among infested plants of both verities had
   functional enrichment with phenlypropanoid metabolic processes as well
   as response to oxygen levels (Fig. [93]5A). In addition, rudimentary
   pathways such as plant pathogen interaction, biosynthesis of various
   secondary metabolites, and biosynthesis of cofactors showed a dynamic
   response during mite attacks (Fig. [94]5B). Interesting, most disease
   resistance protein RPM1 and RSP2 annotated genes enriched in plant
   pathogen interaction pathways exhibited significant differentials
   during the attack of mites. Specifically, RPM1 genes such as
   CSS0012875, CSS0040702, CSS0024074, and CSS0050158 had shown many folds
   higher regulation, while RSP2 genes CSS0004521, novel.5458, CSS0039663,
   novel.5460, novel.2858, CSS0040253, CSS0015515, and novel.2861 were
   down-regulated in infested plants of resistant variety compared with
   susceptible tea variety (Fig. [95]4D). Among the expression profiles of
   biosynthesis of various secondary metabolites pathway genes, several
   kaempferol 3-O-beta-D-galactosyltransferas-related genes (novel.11321,
   novel.9442, novel.11449, CSS0021747, and novel.6409) together with
   cinnamyl-alcohol dehydrogenase (novel.9219), CSS0016418, CSS0045693,
   CSS0038085, and CSS0012945 were identified as having downregulation in
   resistant tea varieties (Fig. [96]4D). The majority of cinnamoyl-CoA
   reductase and 5-O-(4-coumaroyl)-D-quinate 3'-monooxygenase-annotated
   genes had higher expression in plants; those had higher mite damage.
   Besides, transcripts involved in various flavonoids precursors such as
   CSS0046455, novel.8441, CSS0029726, CSS0022064, CSS0041417, CSS0015915,
   CSS0001900, and CSS0006772, as well as the shikimate
   O-hydroxycinnamoyltransferase genes CSS0031453, novel.2630, CSS0036048,
   CSS0043094, CSS0042802, and CSS0012876, had upregulation in resistant
   varieties. While brassinosteroid-6-oxidase 1 (CSS0009174, CSS0036278,
   and CSS0004050) annotated genes and gibberellin-2 beta-dioxygenase
   (CSS0001566, CSS0031596, novel.5250, novel.11228, and novel.4965)
   encoded genes from the functional pathway of secondary metabolites,
   they were identified with differential regulation (Fig. [97]4D),
   especially to mitigate the attack of mites in tea. The differential
   regulation of various secondary metabolic genes together with disease
   resistance protein genes most likely disrupts the plant's defense
   mechanism against the pink tea mite.

Fig. 5.

   [98]Fig. 5
   [99]Open in a new tab

   Functional enrichment analysis of DEGs in comparison of susceptible
   (Guilu No. 1; GL) and resistant (Pingyang Tezao; PYTZ) tea varieties
   before and after pink mite infestation (A) GO enrichment analysis (B)
   Pathway enrichment analysis

Gene regulatory network construction related to defense systems against the
pink tea mite

   A weighted gene co-expression network based on the FPKM values of the
   identified 26,747 DEGs was performed to explore the relationship
   between various modules and mite-infested samples. Finally, 14 modules
   were identified with distinct characteristics as well as different
   levels of relationship with gene dendrograms (Fig. [100]6A). Further,
   there were several hundred genes with a higher connectivity score than
   other genes. The black module had a positive correlation with infested
   plants of susceptible varieties, while the lightcyan module had a
   positive correlation with infested plants of resistant varieties (Fig.
   [101]6B). However, an array of genes in both modules had undulating
   expression in each replicate of mite infestation in each variety. Our
   results further identified blue and turquoise modules in particular
   showed a relatively contrasting correlation among infested plants of
   susceptible and resistant varieties. The blue module showed a positive
   correlation with infested samples of the resistant variety and a
   negative correlation with infested samples of the susceptible variety
   (Fig. [102]6B). In contrast, the turquoise module had a positive
   correlation with infested samples of the susceptible variety and a
   negative correlation with infested samples of the resistant variety. It
   was noteworthy that an array of genes in the blue module were
   upregulated in the resistant variety following mite infestation (Fig.
   [103]6C), but genes in the turquoise module were upregulated in the
   susceptible variety following mite infestation (Fig. [104]6D). Our
   final gene regulatory network construction from blue and turquoise
   further identified many gene sets with strong network linkage and
   higher connectivity. The disease-resistant genes from the plant
   pathogen interaction pathway as well as gibberellin hormone-annotated
   genes showed an independent network (Supplementary Figure S5). However,
   flavonoid metabolite-encoding genes and brassinosteroid-annotated genes
   showed a strong network of co-expression genes. These genes can
   therefore be considered key genes, which are probably involved in plant
   defense responses following a pink tea mite infestation. Our
   co-expression gene regulatory network analysis suggests that
   transcripts involved in plant pathogen interaction and secondary
   metabolism are of great importance in the elaboration of pest
   resistance mechanisms in tea.

Fig. 6.

   [105]Fig. 6
   [106]Open in a new tab

   Weighted gene co-expression network analysis to identify pink mite
   defense-related modules in tea (A) Heatmap plot of gene network: light
   and dark red colors represent lower and higher overlap of gene pairs,
   respectively. B Module relationship with control and pink mite-infested
   traits in susceptible (Guilu No. 1; GL) and resistant (Pingyang Tezao;
   PYTZ) tea varieties (C) Expression heatmap of blue module genes (D)
   Expression heatmap of turquoise module genes

Exploring the association among key transcripts and metabolites for pink tea
mite defense in tea

   To further extract the genes and metabolites associated with plant
   defense, integration between transcripts and metabolites was performed
   after mite infestation. A differential degree of association was
   observed between a large number of differentially accumulated
   metabolites and differentially regulated transcripts). Based on our
   above analysis results, only the key genes from the blue and turquoise
   modules as well as potential metabolites were utilized to display the
   transcript in the metabolite final network (Fig. [107]7A). In this way,
   it was observed that complex interactions of genes and metabolites from
   various metabolic pathways induce defense responses against prolonged
   invasion by the pink tea mite. The functional network of various
   flavonoid pathway-enriched metabolites and genes, however, has higher
   significance than others. Especially the strong network linkage of
   up-regulated genes such as shikimate O-hydroxycinnamoyltransferase
   CSS0000116, CSS0019758, CSS0031453, CSS0036048, and CSS0043094, along
   with the flavanone 3-hydroxylase-like gene CSS0006772 and
   trans-cinnamate 4-monooxygenase gene CSS0005999 (Fig. [108]7B),
   modulate the higher accumulation of metabolites such as
   epigallocatechin, chlorogenic acid, coniferyl alcohol, syringin,
   apigenin-8-C-glucoside, luteolin, 3-O-acetylpinobanksin, neohesperidin,
   fustin, myricetin, and epiafzelechin during pint tea mite severe
   infestation (Fig. [109]7C). While their interaction with other
   transcripts may contribute to a lower abundance of caffeic acid,
   cinnamic acid, p-coumaryl alcohol, naringenin, naringin, and phloretin.
   In this way, their synchronization probably helps to fend off further
   damage through the activation of various other defense-related
   biochemical and molecular mechanisms. In short, these results provide
   valuable insights into the defense regulatory network against pink mite
   attack and highlight the application of genetic engineering research to
   improve the systemic resistance of tea to mites.

Fig. 7.

   [110]Fig. 7
   [111]Open in a new tab

   Overview of network association between potential pathways, genes, and
   metabolites related to defense against pink mite infestation in tea (A)
   Network association between pathways, genes, and metabolites. The red
   large nodes represent potential disease-resistance-related pathways,
   the green large nodes represent potential disease-resistance-related
   genes, and the purple large nodes represent potential
   disease-resistance-related metabolites identified in this study. The
   blue nodes represent genes and metabolites with differential regulation
   before and after mite infestation in tea (B) The expression profiles of
   candidate genes in susceptible (Guilu No. 1; GL) and resistant
   (Pingyang Tezao; PYTZ) tea varieties before and after pink mite
   infestation (C) The abundance levels of key metabolites

Validation of key DEGs involved in pink tea mite defense in tea by qRT-PCR

   To confirm the relative expression profiles of the key genes
   (Supplementary Table S2) predicted to influence the defense responses
   of tea against the pink tea mite, 14 DEGs were selected in total. These
   genes were related to flavonoids and phenolics that enriched
   phenylpropanoid biosynthesis. Further, these showed strong linkage with
   their target metabolites and had higher expression differences in
   transcript data analysis of resistant and susceptible varieties. The
   results of the qRT-PCR analysis confirmed that the expression trends of
   selected genes showed significant differences among resistant and
   susceptible varieties (Fig. [112]8). In addition, most genes had
   relatively higher expression trends in resistant than susceptible
   varieties. Further, the expression profiles were similar to those
   observed in transcriptome datasets (Fig. [113]7B). This confirms the
   accuracy of the RNA-seq results in this study.

Fig. 8.

   [114]Fig. 8
   [115]Open in a new tab

   qRT-PCR validation of key DEGs predicted to be involved in defense
   against pink mite infestation in susceptible (Guilu No. 1; GL) and
   resistant (Pingyang Tezao; PYTZ) tea varieties before and after pink
   mite infestation. The means followed by the same letters were not
   significantly different (P ≤ 0.05) as determined by the least
   significant difference test

Discussion

   Pink tea mite is one of the most devastating sucking insect pests in
   various tea growing regions of the world. Growing evidence suggests
   that feeding mode of mite causes mechanical injury, reduced net
   productivity, and degradation of tea quality-related parameters
   [[116]29, [117]30]. Considering this significant economic loss, this
   study quantified the transcript and metabolite changes during pink tea
   mite feeding damage and further discussed the possible regulatory
   mechanism of tea defense against this mite.

The abundance of flavonoids and phenolic encoding metabolites reduces pink
tea mite infestation

   A comparison of field monitoring showed that Guilu 1 and Pingyang Tezao
   tea varieties had contrasting performance during pink tea mite
   infestation. Further experiments in controlled laboratory conditions
   showed that Pingyang Tezao (a resistant tea variety) is less
   susceptible to pink mite outbreaks than Guilu 1 (a susceptible tea
   variety). As expected, the composition and concentration of various
   biochemical components in both verities showed a significant difference
   during pink tea mite attack. These changes suggest feeding mechanism of
   pink tea mites is probably limited by the biosynthesis, accumulation,
   and abundance of various metabolites. Recently, a large number of
   studies have clearly demonstrated that the biosynthesis as well as
   metabolism of various specialized secondary metabolites contribute to
   plant defense responses against invading leaf herbivores
   [[118]31–[119]33]. These biochemicals sometimes cause volatile
   emissions, act as pest repellents, and ultimately disrupt the pest
   defenses [[120]34, [121]35]. Among different species of secondary
   metabolites, flavonoids, e.g., flavones, flavonols, flavanones, and
   chalcones, are indispensable for shutting down the stress responses
   induced by various herbivore insect species. These induce a cascade of
   plant defense signaling molecules that promote resistance in host
   plants before further insect pest damages occur [[122]36, [123]37].
   During the pink tea mite attack, our study observed that a large number
   of flavonoids were up-regulated in resistant tea varieties. The higher
   abundance of kaempferol-annotated compounds and myricetin probably
   contribute to the reduction of mite damage. These results suggest that
   suppression of these flavonoids may cause more detrimental damage from
   pink mite attack in tea. In particular, the higher the concentration of
   these defense metabolites, the more they trigger the defense signal
   transduction cascade in plants, which ultimately not only disrupts
   further pink mite invasion, but also reduces the efficiency of their
   mating, oviposition, or reproduction. It was previously reported that
   the differential abundance of kaempferol, delphinidin, and quercetin
   activates the response to the tea geometrid attack [[124]28]. In
   another study on tea green leafhopper, non-targeted metabolomics
   analysis identified that various flavonoid compounds act as effective
   defenses [[125]24]. Many recent studies in other crops have shown that
   flavonoid compounds, together with other biochemicals such as
   alkaloids, phenolics, glucosinolates, and terpenoids, mediate
   synergetic or antagonist effects during the invasion of leaf herbivores
   [[126]38–[127]40]. Similarly, the higher accumulation of
   2-O-salicyl-6-O-galloyl-D-glucose, 1-O-salicyloyl-β-D-glucose,
   syringing, and hydroquinone in tea may play an important indirect role
   in detecting early feeding damage and in modulating the defense
   responses of plant components. A previous study in tomato reported
   reduction of total phenolics increased leaf damage by Tuta absoluta
   [[128]41]. However, different species show different defense responses
   to leaf insect pest attacks. The type of insect and chemical cues
   released by the feeding induction of leaf herbivores vary in plants.
   Thus, the precise role of flavonoids and phenolic metabolites that
   mediate the defense mechanism against the pink tea mite needs to be
   investigation in detail, with special emphasis on efficient
   identification of insect chemical secretions that permit the host plant
   to mediate defense signals. These studies will contribute to the
   development of sustainable and environmental friendly control measures
   for the pink tea mite..

Potential transcripts reprogramming enriched in plant pathogen interaction
and flavonoids pathways influence defense responses against the pink tea mite

   Our targeted RNA sequencing identified significant transcript
   reprogramming in response to of pink tea mite feeding between
   susceptible and resistant tea varieties. These differential transcripts
   further identify disruption in genes annotated to plant pathogen
   interaction in response to pink tea mite. According to our findings,
   the upregulation of disease resistance proteins RPM1 and RSP2
   functionally related genes is effective in reducing pink mite damage in
   tea. The differential regulation of these transcripts due to host-plant
   interaction may evolve an appropriate defense system. It is well known
   that disease resistance proteins are the largest family of defense
   genes, encoding nucleotide-binding site leucine-rich proteins, and
   elicit the adaptive mechanism against various insect pests, viruses,
   fungi, and nematodes [[129]42, [130]43]. The activation of these
   specific disease resistance proteins, as well as their co-effectors, is
   linked with the molecular profiles of salivary effectors of invading
   herbivore insect pests [[131]44, [132]45]. The characterization of pink
   tea mite conserved elicitors that induce defense should be explored to
   identify key disease resistance proteins encoding genes. This
   transcript study further recognizes functional genes annotated to the
   biosynthesis of various secondary compounds that are crucial to
   combatthe attack of pink mites in tea. In particular, differential
   regulation of genes that phenylpropanoid-derived various flavonoids and
   phenolics had shown significant disruption in response to mite damage.
   These gene expression changes are most likely correlated with their
   target metabolite components to fend off further exchanges of pink tea
   mite. In addition, these secondary metabolites linked transcript
   variation together with brassinosteroid-6-oxidase-1 annotated genes and
   gibberellin-2-beta-dioxygenase, may provide stimulus to phytohormone
   signaling networks. These results may unbalance the interactions of
   brassinosteroid and gibberellin with other plant hormones such as
   jasmonic acid, salicylic acid, and abscisic acid. Finally, the combined
   use of these different transcripts for reprogramming significantly
   helps tea plants survive pink tea mite outbreak. Considering the
   paramount importance of polyphenols and phytohormones for multiple
   stress scavengers in plants [[133]46, [134]47] including other
   herbivorous insect pests of tea [[135]24, [136]28], the functional
   characterization of these identified genes can help to understand the
   genetic aspects of pink mite interaction with tea plants.

Insights into the defense regulatory network that fends off the attack of
pink mites in tea

   The regulatory mechanism of resistance to herbivorous insect pests is
   complex in tea and mainly depends on a variety of signaling and
   developmental processes from host plant insect interactions [[137]22,
   [138]48, [139]49]. Our key gene co-expression network analysis reveals
   functional network interconnection of multiple genes from various
   pathways mediating defense responses during the invasion of pink mites
   in tea. As expected, several genes from the plant interaction pathway,
   together with the secondary metabolite biosynthesis and their
   interaction genes, play a vital role in boosting tea plant immunity by
   reprograming various metabolic and gene expression changes. At earlier
   stages of the defense response, the transcriptional regulation of
   phenylpropanoid pathway-annotated genes synchronized differential
   accumulation levels of different types of flavonoids and phenolic
   acids. This may be the main reason that our metabolic analysis
   quantified a significant change in different groups of flavonoids and
   phenolic compounds during surveillance of the pink tea mite attack.
   Furthermore, we confirmed a correlation between these pathway genes and
   metabolites. In particular, the regulatory network of up regulated
   candidate genes (PAL, F3H, HCT, and C4H) along with their associated
   potential metabolites (epigallocatechin, chlorogenic acid, syringin,
   apigenin-8-C-glucoside, luteolin, 3-O-acetylpinobanksin, neohesperidin,
   fustin, myricetin, and epiafzelechin shikimic acid, caffeic acid,
   chlorogenic acid, cinnamic acid, p-coumaryl alcohol, and coniferyl
   alcohol) are most likely essential to prevent long-term damage from
   pink mites in tea. In plants, flavonoids, as the largest group of
   secondary metabolites, have already been identified as obligatory to
   avert the problematic aspects of various stresses [[140]50, [141]51].
   Many previous studies, especially on the resistance genetic mechanism
   of herbivorous insect pests, have revealed a possible defense mechanism
   by this group of secondary metabolites. In cotton, the genes encoding
   flavonoid metabolites contribute to the defense system against the
   feeding of whiteflies and aphids [[142]39, [143]52]. The various
   studies in tea reported that a plethora of flavonoid
   metabolite-encoding genes along with various hormones reduce damage
   caused by geometrid and leafhopper in tea [[144]24, [145]28, [146]48,
   [147]53]. Their combined effect causes activation of certain oxidase
   enzymes for detoxification and produces defense proteins to improve
   resistance in tea against leaf herbivory. Considering our results
   together with previous knowledge on leaf herbivory, our study revealed
   that defense against the pink tea mite is mediated by a complex network
   of transcripts and metabolites linked to precursors of various
   secondary metabolites and disease-resistant proteins. Theoretically,
   synchronization of flavonoids genes encoding target metabolites may
   modulate the dynamic regulatory network of various defense responses to
   mitigate the further attack of pink mites in tea. However, the
   mechanism of constitutive or systemic resistance to pink mite invasion
   requires further genetic transformation research studies.

Conclusion

   Our phenotypic evaluation shows that Pingyang Tezao is a useful tea
   variety with significantly lower pink tea mite infestation. It has the
   potential to be used in the breeding of tea varieties to improve mite
   resistance. The qualitative and quantitative differences in the
   metabolites of flavonoids and phenolic acids play important role in
   controlling the long term damage to tea production. In addition, In
   addition, a number of functional genes involved in plant pathogen
   interactions and a dynamic network of regulatory genes associated with
   flavonoids and phenolic compounds contribute to defense responses. The
   combined effects flavonoid-related genes and the abundance of their
   target metabolites are likely to act as dominant defense signals during
   severe pink tea mite infestations. The potential metabolites and genes
   identified in this study will be important for the development of
   sustainable integrated strategies to combat pink mite damage in tea.

Methods

Plant materials and analysis of metabolic substances

   Two early maturing tea varieties selected from Wenzhou Huangyezao,
   named Guilu 1 (susceptible) and Pingyang Tezao (resistant), were used
   as plant materials in this study. Guilv No.1 and Pingyang Tezao tea
   plants were grown for 6 years in the natural environment of the tea
   plant germplasm resource nursery of Zhejiang, Lishui Institute of
   Agricultural and Forestry Sciences, Lishui City, Zhejiang Province,
   China (28°35 N, 119°23 E, 135 m sea level). Both varieties were
   infested with 40 adult pink tea mites per one bud and two leaves in
   controlled conditions. The subjected plants were protected with a
   protective insect net of 50 mesh (100 cm × 100 cm × 100 cm) and treated
   with light and dark cycles of 12 and 12 h at 25 C with 65% RH,
   culturing in an artificial climate room. The level of damage done by
   the pink tea mite was measured on more than 100 plants before and after
   feeding induction. The number of tea mites on the second tender leaf
   under each shoot was examined under a microscope and the degree of mite
   damage was recorded. The grading standards are as follows: 0 mites at
   level 0, 1–10 mites at level 1, 11–50 mites at level 3, 51–100 mites at
   level 5, 101–200 mites at level 7, and 201 or more mites at level 9 as
   previously reported [[148]54]. At the same time, 20 buds were randomly
   selected after 10 days of infestation to measure physiological indexes
   such as fresh eight, bud length, and dry weight in 5 replications. The
   control and infested leaf samples for subsequent analyses were
   harvested in three biological replicates (each biological repeat had 20
   buds collected from different plants), placed immediately in liquid
   nitrogen, and stored in an ultra-low temperature refrigerator at -80 °C
   before target analysis. One gram of biological material was ground into
   powder by using liquid nitrogen in order to examine the metabolite
   profiles. Next, 1.2 ml of 70% aqueous methanol was used to dissolve 100
   mg of powder, vortexed for 30 min, and finally kept in the freezer at 4
   °C for the night. The samples were then centrifuged at 12,000 rpm for
   10 min, filtered with a pore size of 0.22 to get an upper-clear liquid
   for metabolic analysis studies with a combination of tandem mass
   spectrometry (MS/MS) (Applied Biosystems 4500 QTRAP) and
   ultra-performance liquid chromatography (UPLC) (Shimadzu Nexera X2).
   The standards for UPLC-MS/MS analysis were used as previously reported
   for tea plants [[149]28]. In brief, the metabolites were identified
   from different local metabolic databases: MzCloud, Massbank, Metlin,
   and HMDB by comparing the accurate precursor ion (Q1) and production
   (Q3) values, retention time, and fragmentation pattern. After
   normalizing the original peak area information with the total peak
   area, the qualitative and quantitative aspects of metabolites were
   followed by searching the internal database and public databases. The
   statistics base package 3.5.1 within R was applied to do unsupervised
   principal component analysis, whereas hierarchical cluster analysis and
   Pearson correlation coefficients were acquired with the R package
   ComplexHeatmap 2.8.0 [[150]55]. Variable importance in projection (VIP)
   value ≥ 1 and fold change (FC) ≤ 0.5 or ≥ 0.5 were utilized to identify
   substantially different metabolites between target groups by OPLS
   statistical analyses of R MetaboAnalystR 1.0.1 [[151]56]. In OPLS-DA
   modeling, the permutation tests were performed multiple times to verify
   the reliability of the model and cross-validate each compound. The
   metabolites that differed more than 2 times or less than 0.5 were
   considered significant high or low differences between the targeted
   experimental groups. Through the use of the OmicsPLS package in R
   [[152]57], the target metabolomics and transcriptome data were combined
   in this study. Analysis of variance was performed using Statistix 8.1
   software. The significant differences between means were performed with
   the least significant difference test at P-value < 0.001 or the
   Student’s t-test. Each graphical bar in each figure represents the
   results of multiple independent experiments, and values are presented
   as means ± standard deviation (SD).

RNA sequencing and target downstream analysis of differential transcripts

   The control and infested samples from the resistant and susceptible
   varieties were used to perform deep RNA sequencing. The samples were
   taken in three biological replicates. Total RNA was extracted from each
   sample using the CTAB standard protocol [[153]58]. Then, agarose gel
   electrophoresis, a nanophotometer spectrophotometer, a Qubit 4.0
   fluorometer, and a QSEP400 bioanalyzer were utilized to ensure the
   higher purity and integrity of the extracted RNA. The protocols and
   parameters for final library construction and machine sequencing on the
   Illumina platform were followed as previously reported for tea plants
   [[154]24]. In our study, the paired-end sequencing (300 ± 50 bp) was
   generated on an Illumina HiseqTM4000 with the default parameters of the
   vendor’s recommended protocol. Our study used Fastp 0.23.2 to filter
   raw sequenced data [[155]59], Cutadapt 1.10 with the default parameters
   for quality assurance, and clean readings were matched to the tea plant
   reference genome using HISAT 2.2.1 [[156]60]. Genome mapped reads were
   assembled with StringTie 2.1.6 [[157]61] and this software was further
   applied for transcript quantification with the default parameters. Gene
   expression levels were determined as reads per kilobase of transcript
   sequence per million base pairs sequenced (FPKM) with both DESeq 1.22.1
   and EdgeR 3.24.3 [[158]62, [159]63]. Genes having an absolute value of
   log2 fold change ≥ 1 and a P value < 0.05 were classified as
   significant differentially expressed genes (DEGs) among target
   compressions. The Gene Ontology (GO) and Kyoto Encyclopedia of Genes
   and Genomes (KEGG) enrichment analyses of specific groups of DEGs were
   carried out in an R package named ClusterProfiler v4.6.0, and the
   minimum significant criteria was at a P value < 0.05. The weighted gene
   co-expression network (WGCNA) analysis of targeted DGEs was performed
   by WGCNA 1.71 with mergeCutHeight = 0.25 as described previously
   [[160]64]. In brief, the FPKM values were filtered out from the
   susceptible and resistant varieties. The genes were clustered into
   modules based on their topological overlaps that characterize the
   degree of common connections between any two genes and further
   determine scaled correlations for all genes. Using the dynamic
   tree-cutting procedure, a value of 0.85 was used to distinguish between
   the dendrogram's branches, resulting in a network with 14 modules. The
   expression of module genes was shown using a heat map for each module
   in order to identify the relationship between module eigengenes and
   gene expression. In addition, the modules were randomly color-labeled
   based on their Pearson correlation coefficient with control and
   infested samples of each variety. Finally, several thresholds were
   chosen for exporting the gene networks with igraph packages in R
   [[161]65].

qRT-PCR validation of gene expression

   Three biological replicates were used in the qRT-PCR study to confirm
   the validity of the candidate transcriptome genes. Primer 5.0 was used
   to create gene-specific primers (Supplementary Table S2), commercially
   available from Genscript Bioscience in Nanjing, China. SuperScript III
   Reverse Transcriptase (Invitrogen) was used to create first-strand cDNA
   from lg of total RNA, which was then reverse transcribed into cDNA. The
   20-volume reaction for qRT-PCR was prepared with a stand-available kit
   in the lab with default recommendations. The Applied Biosystem 7500
   real-time PCR system (Applied Biosystem, Foster City, CA) was used for
   qRT-PCR with three biological replicates and three technical
   replicates. The protocol was as follows: an initial cycle of 5 min at
   95 °C was used as the first denaturation step, followed by 30 s of
   denaturation at 95 °C, 30 s of annealing at 60 °C, 30 s of extension at
   72 °C for 35 cycles, and a final step of 10 min at 72 °C. As an
   internal reference, relative quantification of gene expression was
   derived and standardized using GhUBQ7. The dissociation curve was used
   to validate the specificity of the primers during qRT-PCR. The relative
   gene expression level was calculated utilizing the comparative 2^−ΔΔCt
   technique [[162]66]. The statistical significance analyses of gene
   expression levels were performed using analysis of variance. The
   significant differences between means were performed with the least
   significant difference test at P-value < 0.05.

Supplementary Information

   [163]Supplementary Material 1.^ (1.9MB, pdf)
   [164]Supplementary Material 2.^ (4MB, pdf)
   [165]Supplementary Material 3.^ (19.4MB, jpg)
   [166]Supplementary Material 4.^ (4.8MB, pdf)
   [167]Supplementary Material 5.^ (2.7MB, png)
   [168]Supplementary Material 6.^ (11.8KB, xlsx)

Acknowledgements