ABSTRACT

   Gall‐inducing insects manipulate host plant development, redirecting
   cellular fate and physiological processes to form novel structures.
   This phenomenon is even more intriguing when the host itself is a
   holoparasitic plant with minimal photosynthetic capacity. In the stem
   of Cuscuta campestris , the weevil Smicronyx madaranus forms galls that
   unexpectedly activate photosynthesis, in contrast to the typical
   suppression of photosynthetic activity observed in leaf‐derived galls.
   This reversal of the usual source‐to‐sink transition highlights a
   unique form of insect‐induced organogenesis. To elucidate the
   underlying mechanisms, we performed transcriptomic, histological, and
   physiological analyses of these galls. RNA‐seq across four
   developmental stages identified differentially expressed genes and
   associated gene ontology terms. Consistent with histological
   observations, genes related to cell division and the cell cycle were
   upregulated in early stage but decreased as the gall matured. Similar
   to leaf‐derived galls, we found high expression of PLETHORA and
   meristem‐related homeobox genes in early gall development, suggesting
   that induction of cell division is involved in various gall types.
   Interestingly, the expression of genes related to floral organ
   development increased through gall development. However, their
   expression patterns showed a marked temporal shift: Floral organ
   identity genes were highly expressed at the initial gall stage, whereas
   floral transition genes were activated later. This suggests that the
   weevil triggers ectopic activation of the flowering pathway in
   non‐floral tissues, potentially redirecting the typical flowering
   cascade to drive gall formation. Consistent with previous findings,
   photosynthesis‐related genes were highly expressed in later stage of
   galls, despite the host being a holoparasitic plant. Shading
   experiments confirmed that photosynthesis is crucial for both gall and
   the weevil growth. This study highlights how gall‐inducers can co‐opt
   host resources and genetic pathways, offering new insights into the
   complexity of plant–insect interactions.

   Keywords: C. campestris , cell cycle, flowering, gall, photosynthesis,
   Smicronyx madaranus

1. Introduction

   Galls are abnormal plant outgrowths induced by various organisms,
   including bacteria, fungi, mites, and insects. Insects are particularly
   noteworthy for their diversity and prevalence in gall formation, with
   approximately 211,000 species capable of inducing galls under various
   environmental and host conditions (Espírito‐Santo and
   Fernandes [31]2007; Desnitskiy et al. [32]2023). These structures can
   develop on different plant organs, including leaves, stems, roots, and
   flowers, reflecting the complex relationship between insects and host
   plants (de Oliveira et al. [33]2014). Galls provide a unique niche that
   offers shelter and nutrition to the gall‐inducing insects. They create
   a controlled microenvironment, protecting the larvae from external
   environmental stresses and predators (Harris and Pitzschke [34]2020).

   The formation of galls is a highly organized process, typically
   involving four developmental stages: initiation, growth, maturation,
   and senescence (Rohfritsch and Shorthouse [35]1982). It is intriguing
   how insects manipulate each stage along with responding to plant
   developmental processes, resulting in complex gall structures. Galls
   typically consist of three tissue types: meristematic tissues, vascular
   tissues, and protective tissues (Stone and Schönrogge [36]2003; Giron
   et al. [37]2016). During gall formation, meristematic cells are
   controlled in their proliferation and differentiation to generate the
   specific tissues that compose the gall. For instance, phylloxera
   utilizes undifferentiated cells associated with leaf veins to develop
   galls (Schultz et al. [38]2019). Meristematic marker genes such as
   class‐1 KNOTTED‐like homeobox (KNOX) transcription factors are often
   activated in the early stages of gall development, as seen in Rhus
   javanica galls induced by the aphid Schlechtendalia chinensis (Hirano
   et al. [39]2020). Interestingly, flowering‐related genes are
   upregulated in the initial stage of galls across distant plant species
   (Takeda et al. [40]2021). Auxin, cytokinin, and other
   hormone‐responsive genes likely collaborate with these transcription
   factors to drive complex morphological features in galls induced by
   insects (Bartlett and Connor [41]2014; Body et al. [42]2019). However,
   gall development in natural field conditions presents significant
   challenges for analyzing developmental stages or gene expression
   patterns, complicating the control of variables that influence gall
   formation. Thus, model systems for gall research are necessary to
   explore the detailed molecular mechanisms.

   In a previous study, we established a new model system consisting of
   the parasitic plant Cuscuta campestris and the stem‐galling weevil
   Smicronyx madaranus (Murakami et al. [43]2021). This system enables
   year‐round stable cultivation under laboratory conditions, allowing for
   controlled experimental studies. Gall induction occurs within a few
   days, with galls enlarging and fully developing into fruit‐like
   structures within 2 weeks. Additionally, RNA interference (RNAi) has
   proven effective for gene function analysis in S. madaranus (Ushima
   et al. [44]2024), enabling detailed molecular analysis of gall
   formation. Unlike typical gall‐inducing systems, where the host plant
   is autotrophic, this model involves a parasitic plant as the gall host,
   presenting an exceptional system to study insect parasitism nested
   within plant parasitism. This multilayered interaction among insect,
   parasitic plant, and host plant allows for the exploration of how
   cross‐kingdom relationships influence physiological and developmental
   processes. Notably, observation of the development of C. campestris
   galls indicated that the photosynthetic activity increases within the
   galls, leading to the accumulation of starch, which serves as food for
   the larvae (Murakami et al. [45]2021). This contrasts with previous
   findings in leaf‐derived galls, where photosynthesis‐related genes are
   typically suppressed, converting leaf tissues from nutrient sources to
   sinks (Takeda et al. [46]2019; Hirano et al. [47]2020). The reverse
   process observed in the C. campestris stem‐galling system offers a
   unique opportunity to investigate the diversity of gall formation
   mechanisms.

   In this study, we aimed to investigate the molecular mechanisms
   underlying gall formation in C. campestris , focusing on gene
   expression and physiological processes, particularly the involvement of
   the flowering pathway. The controlled laboratory conditions of our
   system facilitated time‐course histological analyses and experimental
   manipulations. Unlike most previous studies, which primarily focused on
   initial gall development, our system allowed us to examine the entire
   developmental trajectory, including middle‐ and late‐stage gall
   maturation. This approach uncovered developmental parallels between
   gall formation and fruit development at the cellular level.
   Additionally, we assessed the impact of photosynthetic activity within
   the gall on both gall growth and larval development.

2. Results

2.1. 3D Structure of Different Stages of C. campestris Gall

   The lifecycle of S. madaranus was indicated in Figure [48]1A. Mother
   insects lay eggs at the young node of C. campestris stem, causing the
   galls to swell over time (Figure [49]1B–E). For convenience, we refer
   to galls at different developmental stages as Sgall (small gall), Mgall
   (medium gall), and Lgall (large gall) (see Section [50]4 for detailed
   definitions). Depending on environmental conditions, S. madaranus
   spends its entire lifecycle within the gall from egg to adult, or it
   emerges from the gall at the final larval stage to pupate in the soil
   based on our observation. Notably, larval emergence from the gall
   occurs only after the Lgall stage. During the Sgall stage, three
   elongated spaces were found in the center of the gall by x‐ray
   micro‐computed tomography (micro‐CT) observation (Figure [51]1F). Near
   the eggs, an object with x‐ray absorption similar to that of the egg
   was always located on the epidermal side, implying that this lid‐like
   structure may have been created by the mother and functioned as a cover
   for the oviposition hole (Figure [52]1G). The oviposition hole was
   observed at the Sgall stage (Figure [53]S1A), but it was difficult to
   observe at the Mgall stage due to repair or cell enlargement as the
   gall grows (Figure [54]S1B). One to several larvae were observed per
   gall, and larvae were present inside the larval chamber
   (Figure [55]1H). Several egg‐ or larval‐free holes were observed at all
   gall stages. The size of these holes increased as the stage progressed.
   This indicated that rapid tissue enlargement may create spaces where
   the supply of cells cannot keep up (Figure [56]1I). In the Lgall stage,
   larvae within the chambers had significantly developed (Figure [57]1J).
   Additionally, the number of vascular bundles originally present in the
   stem was six, and these vascular bundles were branching and positioned
   close to the larval chambers (Figure [58]1K,L). Pseudo‐colored 3D
   surface model images revealed striking differences in vascular
   structure between the stem before and during gall formation. Before
   gall formation (Movie [59]S1), the central vascular bundle runs through
   the node and branches toward the axillary bud. However, during gall
   formation (Movies [60]S2 and [61]S3), the vascular bundle diverges from
   the center and surrounds the larval chamber, demonstrating the
   structural reorganization.

FIGURE 1.

   FIGURE 1
   [62]Open in a new tab

   Morphological characteristics of each stage of galls in Cuscuta
   campestris . (A) Life cycle of Smicronyx madaranus . (B–E)
   Representative image of young node before oviposition (B), Sgall (C),
   Mgall (D), and Lgall (E). White triangle in (B) indicate node position
   of C. campestris stem. Scale bars indicate 5 mm. (F–J) Representative
   3D renderings from x‐ray micro‐CT scans of Sgall (F), magnified part of
   Sgall (G), Mgall (H), Lgall (I), and magnified part of Lgall (J). A
   “heatmap” color scale represents x‐ray absorption. Same color
   represents similar x‐ray absorption, indicating similar tissue
   characteristics. Black triangle in (F) indicate the location stem
   connected. White square indicates magnified region shown in (G). (K,L)
   Vascular bundles detected in Lgall. Transverse view of the bottom
   center (K) and branched vascular bundle (L). The numbers in (K)
   indicate the number of vascular bundles at the base of Lgall. Arrow
   indicates branching point. Ho, hole; la, larva; lc, larval chamber; vb,
   vascular bundle.

2.2. Cellular Differentiation Depends on the Gall Growth Stages

   To examine cellular differentiation in C. campestris galls at different
   stages, we analyzed transverse histological sections (Figures [63]2A–D
   and [64]S2A–D). An oviposition hole was observed in the upper part of
   Sgall, as indicated in Figure [65]1F (white square) (Figure [66]S2B).
   Schizogenous ducts were conspicuous around vascular bundles and newly
   formed vascular bundles observed around the larval chamber
   (Figure [67]S2E,F). Compared to the node, the variation of cell number
   was high at the Sgall but moderate at the Mgall and Lgall stages. The
   large variation in cell number at the Sgall stage may reflect that some
   individuals undergo active cell division at this stage (Figure [68]2E).
   The distribution of cells in transverse sections showed smaller cells
   densely packed in the central region, with larger cells toward the
   peripheral gall region, a pattern consistent across all gall stages
   (Figure [69]S2B–D). However, the size of the parenchyma cells on the
   peripheral region increases as the stages progressed (Figure [70]2B–D).
   The median cell size per transverse section of Sgall and Mgall was 100
   and 240 μm^2, respectively, indicating an increase in cell size,
   especially at the peripheral region from Sgall to Mgall
   (Figure [71]2F).

FIGURE 2.

   FIGURE 2
   [72]Open in a new tab

   Cell proliferation depends on gall development. (A–D) Representative
   images of outer layer of transverse paraffin section of node (A), Sgall
   (B), Mgall (C), and Lgall (D). Scale bars indicate 100 μm. (E) Cell
   numbers in node, Sgall, and Mgall. (F) Histogram of the cell size of
   Sgall and Mgall. Dashed lines indicate median value. n = 5221 (Sgall),
   n = 5967 (Mgall). (G) Clustering analysis with normalized CPM data by
   clust. Other clusters were shown in Figure [73]S2. (H) GO terms of
   biological process enriched in Cluster 1. Color indicates ‐log[10]FDR
   and circle size indicates number of genes. The categories marked in
   purple are related to cell division or the cell cycle. (I) Gene
   expression pattern of cyclin and cyclin dependent kinase (CDK). logFC
   is calculated compared with node.

   To understand the molecular mechanisms of C. campestris gall
   development, we conducted transcriptome analysis using three gall
   stages and unoviposited nodal regions as control. Principal component
   analysis (PCA) revealed distinct transcriptome profiles for each gall
   stage and unoviposited nodal regions (Figure [74]S3A). Differentially
   expressed genes (DEGs) were categorized by clustering based on their
   expression patterns across the developmental stages (Figures [75]2G and
   [76]S3B). Gene ontology (GO) enrichment analysis of DEGs in each
   cluster was performed. In Cluster 1, which showed increased expression
   at the Sgall stage and decreased expression as the gall developed
   (Figure [77]2G), many GO terms related to cell division and the cell
   cycle were enriched (Figure [78]2H). Examining the expression patterns
   of cyclins and cyclin‐dependent kinases (CDKs) revealed consistent
   decreases in expression from the Sgall to Lgall stages (Figure [79]2I).
   This is consistent with histological findings that cell division is
   highly active during the early gall stages, leading to an increase in
   the number of smaller cells. As the gall matures, cell division
   decreases, giving way to cell enlargement.

   Genes showing expression changes were categorized into four clusters
   besides Cluster 1 (Figure [80]S3B). In Cluster 2, which consistently
   showed decreased expression from node to Lgall, GO terms such as DNA
   replication and ribosome were enriched (Figure [81]S3C). Clusters 3 and
   4, which showed consistent expression increases from node to Lgall, or
   sequential increases from node to Mgall with no significant difference
   between Mgall and Lgall periods, significantly enriched GO terms
   related to photosynthesis (Figure [82]S3D,E). Other GO terms enriched
   in Clusters 3 and 4 included defense response genes, namely, autophagy,
   MAPK signaling pathway, plant–pathogen interaction, and biosynthesis of
   secondary metabolites that are involved in defense responses
   (Figure [83]S3D,E).

2.3. Expression of Flowering Related Genes in C. campestris Galls

   During the initial stages of gall formation, a shift from normal stem
   development to gall development was anticipated, resulting in a change
   in cell fate. We identified 624 upregulated DEGs between Sgall and
   control (false discovery rate [FDR] < 0.01 and expression‐level
   log[2]FC > 2). GO enrichment analysis of upregulated genes in Sgall
   revealed significant enrichment of genes involved in floral organ
   development (Figure [84]3A). Additionally, terms such as response to
   abscisic acid (ABA) and hormone‐mediated signaling pathways were also
   enriched. To understand the expression patterns of flowering‐related
   genes, we particularly focused on the expression changes of class ABCE
   genes and floral transition regulatory genes, represented in a heatmap.
   A heatmap of the DEGs involved in floral organ development showed that
   the expression of many genes increased during the Sgall stage, followed
   by a decrease (Figure [85]3B). Especially, AGAMOUS (AG) and SEPALLATA
   (SEP) genes were highly expressed (log[2]FC > 3) at the Sgall stage but
   decreased in later stages. On the other hand, the expression of
   Flowering locus T (FT), a positive regulator of flowering, remained
   unchanged from node to Sgall but increased during the Mgall stage
   (Figures [86]3C and [87]S4). Interestingly, the expression of Terminal
   Flower 1 (TFL1), a negative regulator of flowering, showed an increase
   in later stages. Other genes related to floral transition such as FD
   and SOC1 were slightly increased in these expressions (Figures [88]3C
   and [89]S4).

FIGURE 3.

   FIGURE 3
   [90]Open in a new tab

   Time‐course transcriptome analysis using several stage of Cuscuta
   campestris galls. (A) GO terms of biological process enriched in
   upregulated genes of Sgall compared with node (logFC > 2). Color
   indicates −log[10]FDR and circle size indicates number of genes. (B,C)
   Heatmap showing the expression changes of C. campestris orthologs of
   floral organ identity genes (B) and floral transition genes (C) depends
   on gall development based on the RNA‐seq data. The heatmaps present the
   logFC values, relative to the previous stage of tissue (logFC_S means
   Sgall vs. node, logFC_M means Mgall vs. Sgall, logFC_L means Lgall vs.
   Mgall). Asterisks indicate genes that examined the relative expression
   levels by qRT‐PCR in (D). (D) Relative expression levels of flowering
   related genes measured by qRT‐PCR in several tissues of C. campestris .
   Expression levels with three to seven biological replicates were
   normalized with respect to that of CcACT8. Statistically significant
   differences are indicated by different letters above the boxes (ANOVA
   with Tukey's HSD test, p < 0.05).

   The expression levels of these key flowering‐related genes were
   verified by quantitative reverse transcription PCR (qRT‐PCR). In
   qRT‐PCR, several tissues such as internode, haustoria, and flower buds
   were used as samples in addition to the galls. The results showed that
   AG and SEP2 were highly expressed in flower buds (Figure [91]3D) where
   both genes functioned. An increase in the expression of these genes
   during the Sgall stage was observed compared to the node or other gall
   stages. Consistent with previous findings (Mäckelmann et al. [92]2024),
   FT1 and FT2 were highly expressed in the haustoria. Additionally,
   increased expression of these genes was observed during the Mgall stage
   (Figure [93]3D). Interestingly, TFL1 was highly expressed in the
   haustoria, co‐localizing with high FT1 and FT2 expression. The
   expression of FD, a gene known to interact with both FT and TFL1, was
   elevated in haustoria but was also ubiquitously expressed across other
   tissues as previously reported (Mäckelmann et al. [94]2024). These
   results indicated that floral organ identity genes, like AG, were
   predominantly expressed at the initial gall developmental stage. In
   contrast, floral transition genes, like FT, were more actively
   expressed during the later stage of gall maturation. The differential
   expression patterns of these genes appear to correlate with the
   different stimuli arising from the adult and larval stages of
   S. madaranus (Figure [95]S5).

2.4. Spatial Expression Patterns of Flowering‐Related and Meristematic Cell
Niche Marker Genes in C. campestris Galls

   To understand the spatial expression patterns of genes involved in the
   flowering pathway and their contribution to gall development, we
   conducted in situ hybridization analysis using Sgall and Mgall. AG was
   predominantly expressed in parenchyma cells at the Sgall stage, but
   this expression pattern was restricted to the areas surrounding
   vascular tissues at the Mgall stage (Figure [96]4A–D). Conversely, FT2
   was faintly expressed at the Sgall stage but showed strong expression
   in the central areas, especially around the larval chambers in Mgall
   tissues (Figure [97]4E–H). Gall development, a form of de novo
   organogenesis, is thought to require meristematic cells as a
   foundational element (Schultz et al. [98]2019; Hirano et al. [99]2020).
   We found that the meristematic cell niche marker gene, PLETHORA (PLT)
   was expressed specifically at Sgall according to the transcriptome
   analysis (logFC = 6.61 at Sgall compared to node), and it was validated
   by qRT‐PCR (Figure [100]S6A,B). PLT signals were observed throughout
   the Sgall tissue (Figure [101]4I,J), indicating active cell division. A
   strong signal of PLT was also observed in the central area and
   surrounding vascular tissues in Mgall (Figure [102]4K,L). These results
   indicated that both the majority of the Sgall tissue and the central
   area of Mgall exhibited meristematic cell‐like characteristics.

FIGURE 4.

   FIGURE 4
   [103]Open in a new tab

   Expression patterns of flowering‐related and meristematic cell niche
   marker genes in galls. (A–I) In situ hybridization using CcAG (A–D),
   CcFT2 (E–H), and CcPLT (I–L) probes on Sgall and Mgall samples. Scale
   bars indicate 100 μm. lc, larval chamber; oh, oviposition hole. AG,
   AGAMOUS; FT2, Flowering locus T2; PLT, PLETHORA.

2.5. Upregulation of Homeobox Genes in Early Gall Stage

   To identify the meristematic marker genes comprehensively other than
   PLT, GO enrichment analysis was performed. The GO term of meristem
   development (GO:0048507) was enriched in the upregulated DEGs in Sgall
   involving 23 genes (Table [104]S7), including homeobox genes,
   WUSCHEL‐related homeobox 2 (WOX2) and ATHB‐14. As homeobox genes like
   SHOOT MERISTEMLESS (STM) and WUSCHEL (WUS) are known to induce ectopic
   meristems (Gallois et al. [105]2002), we analyzed the expression
   pattern of identified homeobox genes across the developmental stages.
   The results indicated that WOX2 and 24 other homeobox genes were
   significantly upregulated in the Sgall stage (logFC > 1.0), with the
   highest being [106]Cc006977 at logFC of 5.16 (Table [107]S9).
   Conversely, homeobox genes upregulated in the Mgall and Lgall stages
   were predominantly from different classes than those of Sgall, and
   their numbers decreased from 25 in the Sgall stage to eight in Mgall
   and five in Lgall.

2.6. Expression Patterns of Photosynthesis‐Related Genes Identified as Highly
Expressed in Tomato Fruit in C. campestris Galls

   As mentioned above, we observed that the expression of
   photosynthesis‐related genes increased in the later stages of gall
   development. In addition, genes involved in floral organ formation were
   also upregulated during gall development. In developing tomato fruits,
   internal photosynthetic activity contributes to organ growth and
   metabolism (Dong et al. [108]2024). To explore potential
   transcriptional similarities, we compared the expression patterns of
   photosynthesis‐related genes in C. campestris galls to those in tomato
   fruit, which belong to the same order Solanales. We identified
   orthologs for seven out of ten target genes in the C. campestris genome
   (Figure [109]S7). Two orthologous copies were detected for each gene,
   likely attributable to a past whole‐genome duplication event (Vogel
   et al. [110]2018). Analysis of their expression levels in the node and
   across different gall developmental stages revealed that all genes
   exhibited increased expression, either from the Sgall stage or
   progressively at later stages of gall development (Figure [111]S7).

2.7. Impact of Shading Treatment on Gall Development

   To assess the impact of photosynthesis on gall development, we
   conducted a shading experiment (Figure [112]5). In this experiment, we
   measured the size of galls, the amount of chlorophyll and starch, and
   the developmental stages of the insects inside galls under both shaded
   and unshaded conditions. To ensure equal nutrient and water absorption,
   we adjusted the number of C. campestris stems and the number of
   haustoria on the host to be the same. The shaded galls show a
   significant difference in color and size compared to the control
   (Figure [113]5A,B). Regardless of whether there were one or two larvae
   inside the galls, the gall volume under shaded conditions was
   consistently smaller compared to the control (Figure [114]S8A).
   Additionally, there were significant differences in the amounts of
   chlorophyll a and b but not in carotenoids (Figure [115]5C). This
   result supported the color change after shading treatment.
   Unexpectedly, the starch amounts in the galls between the shaded group
   and the control were not significantly different (Figure [116]S8B,C).
   To evaluate the effect of shading treatment on insect development, we
   checked the stage of S. madaranus inside the galls. In shaded galls,
   the proportion of larvae was significantly higher (71.4%), whereas the
   proportion of pupae was lower (28.6%) compared to the control
   (χ^2 = 5.73, p < 0.05; Figure [117]5D). The results showed that shading
   reduced chlorophyll levels and affected gall size and larval growth.

FIGURE 5.

   FIGURE 5
   [118]Open in a new tab

   Effects of shading in developing galls. (A) Galls with or without
   shading treatment. (B,C) Effect on gall development by shading.
   Calculated gall volume (B) and chlorophyll contents (C). **p < 0.01,
   ***p < 0.001, NS, not significant difference; Wilcoxon rank sum test
   with Bonferroni correction. (D) Developmental stage of Smicronyx
   madaranus in the galls of both treatments. Significance of the
   proportion difference between the treatments was assessed using a
   chi‐square test.

3. Discussion

   Gall development involves the complex interplay between insect‐derived
   molecules and plant developmental pathways. Previous studies have
   highlighted the activation of flowering‐related genes during gall
   formation. For instance, in R. javanica galls induced by the aphid
   S. chinensis , genes like APETALA1 (AP1, Class A), SEP (Class E), and
   LEAFY (LFY) are upregulated (Hirano et al. [119]2020). Similarly, in
   Artemisia montana galls, genes such as AP1 (Class A), PI (Class B), and
   SEPs (Class E) are expressed in the initial stage of galls (Takeda
   et al. [120]2019). Schultz et al. ([121]2019) conducted RNA‐seq
   analysis across four developmental stages of Phylloxera‐induced galls
   on grapevine leaves (Schultz et al. [122]2019). They found that TFL1
   expression increased in early galls, whereas AG and SEP (Classes C and
   E) showed higher expression in later stages. In this study, we showed
   that the genes in the flowering pathways were also activated during
   stem‐derived gall formation in C. campestris . However, although there
   were some similarities, the activated genes were somewhat different
   from those identified in the previous studies. Our time‐course RNA‐seq
   revealed increased AG and SEPs expression in the Sgall stage, similar
   to those in R. javanica and A. montana leaf‐derived galls, but no
   increase in LFY expression. The combination of AG and SEP is required
   to produce a carpel (Ferrándiz et al. [123]2010), suggesting that the
   transition from normal stem to carpel‐like characteristics might cause
   gall initiation in C. campestris gall. Unlike grapevine leaf galls,
   TFL1 expression increased in later gall stages, alongside FT expression
   in C. campestris galls. Genes controlling FT expression in Arabidopsis
   thaliana , such as CONSTANS (CO) and GIGANTEA (GI), showed no
   expression changes, similar to other plant galls. Such variation in
   plant responses may depend, at least in part, on differences among host
   plant species. In addition, differences in gall‐inducing insect
   species, together with variations in their biological characteristics,
   such as the location of gall formation (leaf vs. stem), reproductive
   modes (asexual reproduction inside of the gall vs. sexual
   reproduction), feeding styles (sap‐sucking vs. biting), and gall
   induction manner (feeding vs. feeding and oviposition), may further
   contribute to the variation.

   Our study supports the idea that gall induction and growth are
   influenced by specific insect developmental stages; oviposition by
   adult females triggers early gall development, whereas larval feeding
   drives later‐stage gall enlargement (Ushima et al. [124]2024). In this
   study, we demonstrate that the expression of floral organ identity
   genes increases during the early gall stage, which is responding to the
   stimuli by adult females and/or eggs (Figure [125]S5). We also show
   that the induction of floral transition genes occurs during the later
   stage of gall formation induced by larvae. This suggests that the
   insect might activate the flowering pathway in non‐flowering tissues,
   potentially rewiring the typical flowering cascade through effector
   molecules to influence gall development. However, it remains unclear
   whether this phenomenon is a direct or indirect response to
   S. madaranus . Several reports indicate that ABA accelerates flowering
   by promoting the expression of flowering genes (Domagalska
   et al. [126]2010; Riboni et al. [127]2013; Riboni et al. [128]2014).
   The elevation of ABA response genes in Sgall might explain the
   upregulation of flowering genes. Moreover, biotic stress including
   herbivory often triggers changes in gene expression, including
   flowering genes (Takeno [129]2016), which could relate to FT expression
   in Mgall and Lgall stages. To further explore this, investigating the
   role of ABA and other stress hormones could provide valuable insights
   into how insect‐induced gall formation affects flowering pathways.
   Notably, RNAi has been effective for gene function analysis in
   S. madaranus (Ushima et al. [130]2024), paving the way for more
   detailed molecular analysis of the insect's role in gall formation.

   Our findings on the cellular dynamics of gall development suggest
   parallels to fruit growth. In many fruits, rapid cell division in the
   early stages is followed by cell enlargement (Gillaspy
   et al. [131]1993; Bertin et al. [132]2003). A similar pattern is
   observed in C. campestris galls. In the Sgall stage, rapid cell
   division occurs, accompanied by increased expression of PLT and many
   meristem‐related homeobox genes. Earlier studies using leaf‐derived
   galls have similarly shown that key regulators of meristematic cell
   development, such as homeobox genes, play crucial roles in gall
   formation (Schultz et al. [133]2019; Hirano et al. [134]2020; Takeda
   et al. [135]2021). In this study, the upregulation of PLT and
   meristem‐related homeobox genes during the Sgall stage indicates that
   gall formation involves both the induction and maintenance of
   meristematic cells, highlighting a common mechanism shared between
   leaf‐derived and stem‐derived galls. However, as the gall develops to
   the Mgall and Lgall stages, cell enlargement might become more
   prominent, suggesting a shift from active cell division to cell
   expansion, much like the developmental trajectory seen in fruit growth.
   The formation of meristematic tissue around the larval chamber acts as
   a continuous source of cells for nutritive tissue and adjacent regions.
   We observed some homeobox genes highly expressed in the late stage of
   C. campestris galls and verified that PLT continued expressing in the
   central area of Mgall. Schultz et al. ([136]2019) also reported
   increased expression of meristem maintenance genes in later stages,
   which suggests that different meristem‐related genes may be upregulated
   depending on the developmental stage of the insect inside, thereby
   driving plant cell proliferation.

   Previous studies showed a decrease in photosynthesis‐related gene
   expression in leaf‐derived galls compared to leaves (Nabity
   et al. [137]2013; Takeda et al. [138]2019; Hirano et al. [139]2020). In
   contrast, our RNA‐seq analysis on stem‐derived galls of parasitic plant
   C. campestris , which typically show little photosynthetic activity
   (Pattee et al. [140]1965; Sherman et al. [141]1999), revealed an
   increase in photosynthesis‐related gene expression as the galls
   developed. This was associated with the findings that Cuscuta galls
   induced by Smicronyx exhibit higher chlorophyll and chloroplast
   concentrations, along with increased photosynthetic activity compared
   to stems (Anikin et al. [142]2017; Zagorchev et al. [143]2018; Murakami
   et al. [144]2021; Zagorchev et al. [145]2021). But why does chlorophyll
   accumulation occur in C. campestris galls? McNeal et al. ([146]2007)
   reported chlorophyll accumulation during ovule development in Cuscuta
   genus, contributing to the synthesis of lipids serving as energy
   reserves for seedlings. Due to limited lipid availability from the host
   and high lipid demand during fruiting, this efficient synthesis pathway
   likely explains the preservation of the photosynthetic apparatus in
   Cuscuta (McNeal et al. [147]2007). Additionally, the enrichment of the
   GO term “response to lipid (GO:0033993)” at all stages of gall
   development suggests that lipids are accumulating in the galls of
   C. campestris (Table [148]S7). Under shaded condition, chlorophyll
   content decreased in C. campestris gall as well as in other plants
   (Pružinská et al. [149]2005; Song et al. [150]2014), and the
   development of S. madaranus within the galls slowed. However, there was
   little difference in starch content between the control and shaded
   galls. This suggests that the primary role of photosynthesis in the
   galls is not only to supply starch to the larvae but potentially to
   produce other products. Another study has also shown that the growth of
   the larvae inside the shaded galls was substantially reduced (Haiden
   et al. [151]2012). The authors suggested that photosynthesis at the
   gall may contribute to oxygen provision in addition to photosynthates.
   In the case of C. campestris gall, chlorophyll may contribute to supply
   lipids and/or potentially oxygen to the larvae within the closed galls
   for promoting normal development. These aspects will be further
   explored in future research. Additionally, research on tomatoes, which
   belong to the same order Solanales as C. campestris , has shown that
   photosynthesis‐related genes are upregulated in the locule, which is
   located at the center of the tomato fruit (Lemaire‐Chamley
   et al. [152]2005; Lytovchenko et al. [153]2011). Chlorophyll
   accumulation and photosynthesis in tomato fruits play an important role
   in fruit metabolism and development (Dong et al. [154]2024). Our
   analysis confirmed that the orthologous genes with tomato were highly
   expressed in later stage of C. campestris galls (Figure [155]S7). These
   findings suggest that the activation of photosynthesis‐related genes in
   C. campestris galls may involve transcriptional programs similar to
   those operating during fruit development in Solanales, supporting the
   idea that galls may have partially co‐opted fruit‐like developmental
   pathways.

   This study clarifies that the developmental pathway of C. campestris
   galls at the gene expression, cellular, and physiological levels
   partially follows the fruit developmental pathway. This provides a
   fascinating example of how gall‐inducers can co‐opt host resources and
   rewire genetic pathways to support their growth. These findings
   underscore the complexity of plant–insect interactions and offer new
   insights into the adaptive strategies of gall‐inducing insects.

4. Materials and Methods

4.1. Plant and Insect Materials and Growth Conditions

   A laboratory strain of S. madaranus was collected from Neagari Nomi
   City, Ishikawa, Japan. It was maintained on C. campestris parasitizing
   Nicotiana benthamiana in the cargo box and placed in the growth room at
   25°C in a long day condition (14 h light/10 h dark) as described
   (Murakami et al. [156]2021). In this paper, we refer to the “initial
   gall” (Murakami et al. [157]2021) as Sgall, which appears approximately
   2 days after the insect introduction. Mgall and Lgall were defined as
   stages 6 and 10 days after Sgall formation (Figure [158]1A). Eggs are
   always present, but no larvae are found in Sgall; however, larvae are
   present in all Mgall and Lgall. All gall samples used in this study
   were confirmed to contain larvae at the time of collection; no samples
   included post‐emergence galls. The pictures of each stage of gall and
   C. campestris node were taken by Leica MZ10F (Wetzlar, Germany).

4.2. Micro‐CT Imaging

   Several stages of C. campestris galls and stems were collected under
   lab conditions on December 23, 2022, for micro‐CT scanning. Samples
   were fixed in 4% paraformaldehyde (PFA) solution in 1X
   phosphate‐buffered saline (PBS), and after subsequent immersion in 70%
   (v/v) and 35% (v/v) ethanol, distilled water, and 25% (v/v) Lugol's
   solution as described previously (Maeno and Tsuda [159]2018). The
   samples were scanned using ScanXmate‐L090T (Com Scan Techno, Japan)
   with a tube voltage of 25 kV and tube current of 170 μA. Visualization
   and measurement of the micro‐CT images were conducted using Volume
   Viewer of Fiji ver. 2.15.1 (Schindelin et al. [160]2012). Fiji can
   apply color mapping to the dataset to highlight the different regions
   as shown in Figure [161]1F–L. Reconstruct 3D models and videos from
   sectional image sequences by x‐ray CT were created using Molcer Plus
   v1.8.5 (White Rabbit, Japan).

4.3. Histological Analysis Using Paraffin Sections

   About 3‐week‐old tobacco ( N. benthamiana ) plants were attached with a
   5‐cm‐long C. campestris vine. Adult S. madaranus weevils were infested
   onto C. campestris for inducing galls. Two days after inoculation of
   the weevils, galls were collected and fixed in the fixative solution
   consisting of 4% PFA in PBS under vacuum conditions until the samples
   sank to the bottom of the tube. After fixation, the samples were
   dehydrated through a graded ethanol series (50%, 70%, 80%, 90%, 95%,
   and 100% ethanol) for 45 min each and then followed by Lemosol (Wako,
   Japan) in the ethanol series (2:1, 1:1, 1:2 and 100% lemosol) for
   45 min each. The tissues were then infiltrated with Paraplast Plus
   (Sigma‐Aldrich, USA) at 60°C with overnight incubation. After the
   tissue was embedded in paraffin, it was mounted on a wooden block,
   which was then clamped onto a microtome for section cutting. The
   embedded samples were cut to 10 μm thickness to form ribbons by MICROM
   HM325 rotary microtome, flattened fully on adhesive glass slides
   (Matsunami, Japan) placed on a hot plate at 42°C, and then dried. The
   sections were deparaffinized in 100% Lemosol solution and rehydrated
   through a graded ethanol series. The sections were stained with 1%
   Safranine or 0.1% Toluidine blue O for histological observation.

4.4. Histological Analysis Using Technovit Resin

   Tissues were fixed in 500 μL of 4% PFA solution for 2 h. Following
   fixation, the samples were dehydrated sequentially in graded ethanol
   solutions (70%, 80%, 90%, 95%, and 100%). The dehydrated samples were
   then placed in a tube containing 500 μL of 100% ethanol and agitated on
   a rotator at approximately six rotations per minute, with 100 μL of
   Technovit 1 solution (Technovit 7100 + hardener I) (Nisshin EM Co.
   Ltd., Japan) added every 10 min until a total volume of 1000 μL was
   reached. The Technovit/ethanol mixture was then removed, and 500 μL of
   fresh Technovit 1 solution was added to the sample, followed by a 1‐h
   incubation. Post‐incubation, the samples were transferred to embedding
   dishes, and 1 mL of Technovit II solution (Technovit 7100 + hardener
   II) (Nisshin EM Co. Ltd., Japan) was added. The samples were left
   overnight at 4°C, covered with parafilm, to allow the resin to harden.
   Once hardened, the resin blocks were removed from the embedding dishes,
   trimmed, and attached to support blocks using instant adhesive. The
   samples were then clamped onto a microtome for sectioning. The embedded
   samples were sectioned at a thickness of 10 μm using a MICROM HM325
   rotary microtome. The sections were then placed on adhesive glass
   slides (Matsunami, Japan) and dried on a hot plate at 42°C. The
   sections were subsequently stained with 1% safranin for 15 s, washed
   under running water, dried, and observed under a microscope for further
   analysis.

4.5. Cell Segmentation

   Cell number and cell size were quantified from safranin‐stained
   transverse sections using Fiji ver. 2.15.1 (Schindelin
   et al. [162]2012). At each developmental stage, one transverse section
   was obtained from the center of each of different individual galls. The
   number of galls analyzed was three for node, seven for Sgall, five for
   Mgall, and three for Lgall. Cell identification was performed using the
   Find Maxima function under the Process menu. Settings included checking
   the Preview point selection option, setting the Output type to
   Segmented particles, and enabling the exclusion of cells touching the
   edge. An automated segmented cell image was then generated and saved
   for analysis. For analysis, measurement settings were configured by
   navigating to Analyze and selecting Set Measurements. Necessary
   measurement options such as area, shape descriptors, display label,
   mean gray value, and perimeter were chosen. For particle analysis, the
   Analyze Particles function under Analyze was used. Settings were
   adjusted to display outlines and results. This process resulted in an
   image with detailed information on each cell's size and shape metrics,
   ready for analysis.

4.6. Transcriptome Analysis

   Gall samples were collected under lab conditions on August 3, 2020, and
   immediately frozen in liquid nitrogen. All samples used for RNA‐seq
   analysis contained larvae at the time of collection, ensuring
   consistency in insect presence across developmental stages.
   S. madaranus larvae in Mgall and Lgall were carefully removed prior to
   sampling, ensuring that our RNA‐seq data do not contain insect‐derived
   reads. Eggs in Sgall were not removed before RNA extraction, however,
   we filtered out all reads that did not map to the C. campestris genome
   during data processing, effectively removing potential insect‐derived
   sequences. Total RNA was extracted from different developmental gall
   stages (unoviposited nodal region including 1 cm of stems at the node,
   Sgall, Mgall, and Lgall as described above) by Maxwell RSC Plant RNA
   kit (Promega, USA) according to the manufacturer's instructions. Total
   RNA integrity was quantified using an Agilent 2100 Bioanalyzer (Agilent
   Technologies, USA). RNA‐seq libraries for Illumina sequencing were
   constructed with TruSeq kit (Illumina, USA), and the qualities of
   libraries were checked on an Agilent 2100 Bioanalyzer. Libraries with
   three biological replicates per stage were sequenced on a HiSeq2000
   Illumina Sequencer at Macrogen Co. (Tokyo, Japan). A total of
   621,387,924 raw sequencing reads were generated (10–17 million read
   pairs per sample; Table [163]S1) and used for gene expression
   profiling. Raw reads were mapped to the published genome of
   C. campestris (Vogel et al. [164]2018) with Hisat2, and subsequently
   read count table was created by using StringTie and prepDE.py (Pertea
   et al. [165]2016).

4.7. Functional Annotation and GO Enrichment Analysis of RNA‐Seq Data

   For C. campestris genes, because many genes are not functionally
   annotated in the current published C. campestris genome (Vogel
   et al. [166]2018), we did blastp to the A. thaliana genes for annotated
   C. campestris transcriptome with e‐value < 1e‐04. We also used
   eggNOG‐mapper (Cantalapiedra et al. [167]2021) for identifying
   hypothetical proteins. TAIR ID hits were used for GO enrichment
   analysis on [168]http://geneontology.org/ for gene clusters and
   modules. ShinyGO ver. 0.80 was used for visualization of GO analysis
   results (Ge et al. [169]2019). Detection of the DEGs was performed by
   edgeR (Robinson et al. [170]2010) with FDR < 0.01 and log[2]FC > |2|.
   Clustering analysis was done by using clust (Abu‐Jamous and
   Kelly [171]2018) with default option and clustered genes based on read
   count. The genes included in each cluster are listed in Tables [172]S2
   – [173]S6.

4.8. Detection of Orthologous Genes With Tomato, Solanum lycopersicum

   Ten photosynthesis‐related genes upregulated in the tomato fruit
   (Lemaire‐Chamley et al. [174]2005) were selected for the analysis. The
   gene sequence of tomato ( S. lycopersicum ) was retrieved from
   Phytozome v12 ([175]https://phytozome‐next.jgi.doe.gov/). Orthologous
   genes between C. campestris and S. lycopersicum were identified using
   OrthoFinder (Emms and Kelly [176]2019), a sequence similarity‐based
   ortholog group prediction tool, with default parameters.

4.9. Expression Analysis by qRT‐PCR

   Several stages of gall samples as indicated Sgall, Mgall, and Lgall
   were collected from lab‐grown samples. Nodes, including the stem axis,
   were collected 1 week after parasitism induction on N. benthamiana .
   Flower buds, internodes, and haustoria samples were collected 1 month
   after parasitism induction. Tissues were frozen in liquid nitrogen
   immediately after collection and homogenized by Tissue Lyser II
   (QIAGEN, Netherlands). Total RNA was isolated using a Maxwell RSC Plant
   RNA Kit (Promega, Madison, WI). RNA concentration and purity were
   assessed using a NanoDrop ND‐2000 spectrophotometer (Thermo Scientific,
   USA). For synthesizing first strand cDNA, we used 1 μg of total RNA as
   a template using the oligo‐dT primer supplied with the SuperScript IV
   kit (Thermo Scientific, USA). Quantitative real‐time (qRT) PCR of the
   target genes was performed using SYBR Green qPCR Master Mix (Thermo
   Scientific, USA) and LightCycler 480 Instrument (Roche, Switzerland)
   with specific primer sets, as shown in Table [177]S10. The PCR
   temperature profile was 95°C for 1 min, followed by 40 cycles of 95°C
   for 15 s and 60°C for 1 min as two‐step PCR. The dissociation stage was
   performed at 95°C for 15 s and 60°C for 1 min, followed by a slow ramp
   to 95°C. CcACT8 was used as an internal control for normalization.
   Quantitative PCR and dissociation curve analyses were performed on
   three to seven biological replicates per tissue using a standard curve
   method.

4.10. In Situ Hybridization

   RNA probes for in situ hybridization were produced using CcAG, CcFT2,
   and CcPLT partial sequences cloned in pENTR‐D‐Topo vector using primers
   listed in Table [178]S10 followed as the manufacturer's protocol. The
   labeled RNA probes were synthesized using in vitro transcription in the
   presence of Digoxigenin‐label by RNA polymerases T7 or SP6 provided
   with the DIG RNA labeling kit (Roche Diagnostics, Switzerland). The
   tissue sections of 10 μm thickness were sectioned by MICROM HM325
   rotary microtome followed by deparaffinization with 100% lemosol for 10
   min and hydrated in a graded ethanol series. Further hybridization
   procedure was performed according to a previously published protocol
   (Shimizu et al. [179]2018). Nitroblue tetrazolium chloride (NBT) (Roche
   Diagnostics, Switzerland) and 5‐bromo‐4‐chloro‐3‐indolyl phosphate
   (BCIP) (Roche Diagnostics, Switzerland) were used as substrates and
   incubated for 10 min or until positive signals on sense/antisense
   probes were seen. After detecting the signals, Immuno In Situ Mounting
   Medium (Funakoshi, Tokyo, Japan) was added for mounting. Images were
   obtained using a Leica microscope. The captured images were processed
   using Leica LAS X.

4.11. Shading Experiment

   Galls finding after 2 days of Sgall were used for the shading
   experiment. The stem of 5 mm above and below the galls was covered with
   foil as a control, whereas the entire galls, including the stems above
   and below the galls, were covered with foil as the shading treatment.
   After 12 days of foil covering, the galls' pictures were taken after
   removing the foils, then the length and width of the galls were
   measured using ImageJ (ver. 1.51n). Subsequently, the galls were
   bisected with a razor blade, revealing either one or two larvae within
   each gall. The developmental stages of the larvae were then recorded.
   One‐half of the bisected galls were used for chlorophyll
   quantification, and the other half for starch quantification. The
   calculation of gall volume and the quantification of chlorophyll and
   starch followed the methods in the previous study (Murakami
   et al. [180]2021). The estimation of total carotenoid was done as
   described in the previous study (Wellburn [181]1994).

5. Quantification and Statistical Analysis

   Statistical analyses and graphical illustrations were all performed
   using the statistical software R v.4.3.1 with the interface RStudio
   v2023.06.1 and the R packages, tidyr, ggpubr, ggplot2, purrr, and
   multcompView. For differences in relative gene expression level, we
   performed one‐way ANOVA followed by a post hoc analysis with Tukey's
   honest significance difference test. For differences in gall size,
   carotenoid, chlorophyll, and starch content between control and shading
   treatment, we performed the Wilcoxon rank sum test with Bonferroni
   correction. The chi‐square test was used to examine the difference in
   the proportion of larvae and pupae in the galls between control and
   shading treatment.

Author Contributions

   N.J.U., T.T., and K.B.U. designed this study and wrote the manuscript.
   R.U., T.T., and K.B.U. performed RNA‐seq analysis. Y.Y. conducted
   qRT‐PCR. N.J.U. and K.B.U. conducted all other experiments and created
   figures. All authors have reviewed and commented on the manuscript.

Conflicts of Interest

   The authors declare no conflicts of interest.

Supporting information

   Data S1: Peer review.
   [182]PLD3-9-e70099-s009.pdf^ (139.8KB, pdf)

   Figure S1: Morphological change of Cuscuta campestris gall. (A)
   Oviposition hole at Sgall. Red arrowhead and the magnified picture
   surrounded by red indicate the oviposition hole. (B) Oviposition hole
   was hardly observed at Mgall stage.

   Figure S2: Complete transverse sections at different stage of Cuscuta
   campestris galls. Node before oviposition (A), Sgall (B), Mgall (C),
   and Lgall (D). (E,F) Transverse sections of Mgall (E) and Lgall (F).
   lc, larval chamber; oh, oviposition hole; pa, parenchyma cells; ph,
   phloem; vb, vascular bundle; xy, xylem. Asterisks indicate schizogenous
   ducts. Arrows in (F) indicate newly formed vascular bundles. Scale bars
   are 100 μm.

   Figure S3: Summary of RNA‐seq data using several stage of Cuscuta
   campestris galls. (A) PCA analysis on all the samples. (B) Clustering
   analysis with normalized CPM data by clust. (C–E) KEGG pathway
   enrichment analysis using Cluster 2 (C), Cluster 3 (D), and Cluster 4
   (E). No GO term was enriched in cluster 5. FDR < 0.05; Top 10 pathways
   are represented.

   Figure S4: Flowering related genes expression levels in node, Sgall,
   Mgall, and Lgall quantified as normalized counts per million (CPM).
   CPMs are calculated by normalizing the read counts by the total counts
   per sample.

   Figure S5: Flowering related signal pathway and interaction with
   Smicronyx madaranus . The flowering genes' relationship was referred
   from Corbesier and Coupland ([183]2006). Arrows and clasps indicate
   positive regulation and negative regulation of downstream genes
   respectively. Solid and dashed line indicate direct and indirect (or
   unknown) regulation.

   Figure S6: CcPLT‐like gene expression pattern. (A) Expression levels in
   node before oviposition, Sgall, Mgall, and Lgall quantified as
   normalized CPM. (B) Expression levels in several tissues including
   galls quantified by qRT‐PCR. Statistically significant differences are
   indicated by different letters above the boxes (ANOVA with Tukey's HSD
   test, p < 0.05)

   Figure S7: Expression pattern of photosynthesis related genes which
   were the orthologous to tomato genes. The IDs starting with “Solyc”
   above each box plot represent the tomato gene IDs, indicating that
   Cuscuta campestris has two corresponding orthologous genes for each.
   The expression levels in node, Sgall, Mgall, and Lgall quantified as
   normalized counts per million (CPM). CPMs are calculated by normalizing
   the read counts by the total counts per sample.

   Figure S8: Influence of shading experiments on starch content. (A)
   Relationship between the gall size and number of larvae in the gall.
   (B) Starch in the galls stained blueblack with Lugol's iodine solution.
   Scale bars, 5 mm. (B) Starch content quantification by using starch
   assay kit. **P < 0.01, NS, not significant difference; Wilcoxon rank
   sum test with Bonferroni correction.
   [184]PLD3-9-e70099-s011.pdf^ (13.4MB, pdf)

   Table S1: Sequence information of samples used in the present study.
   [185]PLD3-9-e70099-s008.xlsx^ (12KB, xlsx)

   Table S2: Genes listed in Cluster 1 (n = 2256).
   [186]PLD3-9-e70099-s007.xlsx^ (255.8KB, xlsx)

   Table S3: Genes listed in Cluster 2 (n = 6635).
   [187]PLD3-9-e70099-s013.xlsx^ (629.8KB, xlsx)

   Table S4: Genes listed in Cluster 3 (n = 7189).
   [188]PLD3-9-e70099-s015.xlsx^ (583KB, xlsx)

   Table S5: Genes listed in Cluster 4 (n = 2142).
   [189]PLD3-9-e70099-s006.xlsx^ (239.1KB, xlsx)

   Table S6: Genes listed in Cluster 5 (n = 361).
   [190]PLD3-9-e70099-s004.xlsx^ (52.6KB, xlsx)

   Table S7: Top50 GO terms enriched by GO enrichment analysis using
   up‐regulated genes (logFC > 1) in gall of Cuscuta campestris compared
   to node.
   [191]PLD3-9-e70099-s001.xlsx^ (21KB, xlsx)

   Table S8: Top50 GO terms enriched by GO enrichment analysis using
   down‐regulated genes (logFC < −1) in gall of Cuscuta campestris
   compared to node.
   [192]PLD3-9-e70099-s010.xlsx^ (20.7KB, xlsx)

   Table S9: Differential expression levels of transcription factors
   containing homeobox domain in several stage of Cuscuta campestris
   galls.
   [193]PLD3-9-e70099-s003.xlsx^ (26.4KB, xlsx)

   Table S10: List of primers used in this study.
   [194]PLD3-9-e70099-s012.xlsx^ (10.5KB, xlsx)

   Movie S1: Micro‐CT movies of Cuscuta campestris node. Blue color
   indicates vascular bundle inside the stem. White color indicates
   axillary bud attached the node.
   [195]Download video file^ (14.2MB, mp4)

   Movie S2: Micro‐CT movies of Cuscuta campestris Mgall. Blue color
   indicates vascular bundle inside the stem. Pink color indicates S.
   madaranus larvae.
   [196]Download video file^ (15.1MB, mp4)

   Movie S3: Micro‐CT movies of Cuscuta campestris Mgall with
   cross‐sectional view. Blue color indicates vascular bundle inside the
   stem. Pink color indicates S. madaranus larvae.
   [197]Download video file^ (1.2MB, mp4)

Acknowledgments