Roles of the APETALA3–3 ortholog in the petal identity specification and morphological differentiation in Delphinium anthriscifolium flowers

Abstract The genus Delphinium (Ranunculaceae) with its unique and highly complex floral structure is an ideal system to address some key questions in terms of morphological and evolutionary studies in flowers. In D. anthriscifolium, for example, the original eight petal primordia differentiate into three types at maturity (i.e., two dorsal spurred, two lateral flat, and four ventral reduced petals). The mechanisms underlying their identity determination and morphological differentiation remain unclear. Here, through a comprehensive approach combining digital gene expression (DGE) profiles, in situ hybridization, and virus-induced gene silencing (VIGS), we explore the role of the APETALLATA3–3 (AP3–3) ortholog in D. anthriscifolium. Our findings reveal that the DeanAP3–3 not only functions as a traditionally known petal identity gene but also plays a critical role in petal morphological differentiation. The DeanAP3–3 gene is expressed in all the petal primordia before their morphological differentiation at earlier stages, but shows a gradient expression level difference along the dorsventral floral axis, with higher expression level in the dorsal spurred petals, intermediate level in the lateral flat petals and lower level in the ventral reduced petals. VIGS experiments revealed that flowers with strong phenotypic changes showed a complete transformation of all the three types of petals into non-spurred sepals. However, in the flowers with moderate phenotypic changes, the transformation of spurred petals into flat petals is associated with moderate silencing of the DeanAP3–3 gene, suggesting a significant impact of expression level on petal morphological differentiation. This research also shed some insights into the role of changes in gene expression levels on morphological differentiation in plants.


Introduction
Flowers, unique to angiosperms, exhibit a wide diversity in size, morphology, structure, and color, with a particularly rich abundance in the basic structural variation.The basic structure of a f lower, represented at the types, number, arrangement, fusion, characteristics, and symmetry of its f loral components (e.g., sepals, petals, stamens, and carpels), plays an integral role in its overall form [1]. Any change in one or more of these aspects can have a profound effect on the f lower's structure [1].This, in turn, allows f lowers to be categorized as simple or complex based on their basic structure.Simple f lowers have a small and fixed number of simple f loral organs or a larger number while maintaining radial symmetry, such as Arabidopsis (Brassicaceae), Papaver (Papaveraceae), and the Chloranthaceae [2][3][4].Conversely, complex f lowers, manifest bilateral or asymmetric symmetry or possess specialized f loral organs, as typified by taxa such as the Fabaceae, Lamiaceae, Ranunculaceae, Orchidaceae, and Zingiberaceae [3][4][5].Complex f lowers, widespread among angiosperms, hold substantial adaptive significance and make these plants essential resources for breeding ornamental plants [3,5].Therefore, an in-depth exploration into the molecular mechanisms that underpin establishment and diversification of f loral basic structures is imperative.
One of the most foundational research achievements in the establishment of f lower basic structure is the proposal of the ABC model of f lower development: A function genes specifies sepals; A + B function genes, petals; B + C function genes, stamens; and C function genes, carpels [6][7][8].The ABC model is generally conserved, with most f loral organ identity genes responsible for A, B, and C functions predominantly belonging to the MADSbox gene family, such as APETALA1 (AP1) and AGAMOUS-LIKE6 (AGL6), APETALA3 (AP3) and PISTILLATA (PI), and AGAMOUS (AG) subfamily members [9][10][11].The ABC model also provides a theoretical foundation and research clues for understanding diversification of f lower basic structure.In numerous taxa, alterations in f loral organ types and compositions are closely associated with the modification of f loral organ identity genes.For instance, the specialized orchid labellum (Orchidaceae) results from the duplication and neofunctionalization of a B-function gene, AP3, coupled with a reduction in its expression domain [12].Similarly, the formation of petal-like tepals in lilies (Liliaceae) plants, such as Lilium brownii and tulips (Tulipa gesneriana), is caused by heterotopic expression of B-functional genes in the outer two whorls of f loral organs [13,14].In the Ranunculaceae, the independent loss of petals is often associated with the 'inactivation' of B functional genes, resulting in the transformation of petals into sepals, or the expansion of C functional genes into the petal whorl (leading to the transformation of petals into stamens) [15,16].While these studies have provided a crucial foundation for comprehending the molecular mechanisms behind establishment and diversification of f lowers with relatively simple structures, they fall short of explaining how complex f lowers form and evolve.The Delphinieae tribe (Ranunculaceae) are an ideal group for addressing these questions, given the extraordinary complexity of their f loral structures.Characterized by a spiral arrangement of f loral organs and a nearly bilateral symmetry, the f lowers in this tribe are often referred as almost monosymmetric spiral f lowers exemplifying a form of 'irregular' f lowers [17] (Fig. 1A-C).In particular, a significant factor contributing to the heightened complexity of the f lower basic structure in this tribe is the presence of different petal types and their combinations [4,18,19].
For instance, in the Delphinium genus, most plants have f lowers with two dorsal spurred petals (SP), two lateral f lat petals (FP), and four ventral reduced petals (RP), forming a '2 + 2 + 4' pattern (Fig. 1B and C); however, Delphinium consolida f lowers consist of two fused dorsal spurred petals and six reduced petals, forming a '(1 + 1) + 6' pattern; the Aconitum and Gymnaconitum genera showcase a '2 + 6' petal arrangement with two dorsal spurred petals and six reduced petals.The basal genus Staphisagria also follows a '2 + 2 + 4' petal combination, akin to Delphinium [18,20] (Fig. 1A).Therefore, investigating the mechanisms governing the petal morphological differentiation in this tribe is crucial for comprehending the diversity of fundamental plant structures and the formation of complex f lowers.
Previous studies on the Ranunculaceae have significantly advanced our understanding of the development of complex f loral structures.Studies have demonstrated that the AP3 class genes in this family underwent two successive gene duplications, resulting in the emergence of three distinct copies, namely AP3-1, AP3-2, and AP3-3 [16,[20][21][22].It has been established that the AP3-3-lineage genes specifically determine the petal identity due to subfunctionalization, exhibiting highly specific expression in petals of the Ranunculaceous f lowers.Mutations or silencing of this gene can lead to the transformation of petals into sepals [4,11,[21][22][23][24].A recent study in a Delphinium species, Delphinium ajacis, indicates that the DeajAP3-3 gene functions specifically as a spurred petal identity gene, while has no roles in reduced petals, which was associated with its differential expression in levels along the dorsiventral axis of f lower [4].However, the '(1 + 1) + 6' combination of D. ajacis petals is still relatively simpler compared with other lineages of the Delphinieae tribe.In Delphinium anthriscifolium f lowers, for example, their petals display a '2 + 2 + 4' combination (Fig. 1C), representing a more complex f loral structure, but the mechanisms underlying their morphological differentiation remain unclear.
To gain a deeper understanding of the molecular mechanisms underlying morphological differentiation of petals in complex f lowers of the Delphinieae, we are using the representative plant D. anthriscifolium as our research material.Through morphological, comparative transcriptome analysis, and molecular biology approaches, we aim to reveal the distinct morphological characteristics and developmental trajectories of three petal types, identify key genes and molecular processes controlling the morphological differentiation of different petal types, and validate their functional characteristics.These research findings are expected to lay the foundation for understanding the molecular mechanisms of complex f lower formation and diversification in the Delphinieae and, more broadly, in angiosperms.Additionally, they will provide crucial theoretical insights for ornamental horticulture breeding.

Morphological differentiation of petals
To understand the morphological and developmental bases of petal differentiation, we chose D. anthriscifolium as a representative for morphological and anatomical studies.According to Zhang et al. (2022), the development of D. anthriscifolium flowers, involves 16 stages, from the initiation of f loral meristems to maturity [25].Prior to stage 9 (S9), all petal primordia initiate as spherical shape, then rapidly elongate and expand into shapes resembling f lattened rectangles, exhibiting no obvious morphological differentiation.At S9, the pair of dorsal petals begin to show basal depressions, while that of lateral petals develop bifurcated lobes, and the development of ventral petals ceases, indicating the onset of dorsal-ventral differentiation of the f lowers (Fig. 1D).By S10, the depressions of the dorsal petals grows into spurs, and the lateral petals continue to enlarge, while the ventral petal primordia cease to grow (Fig. 1D).At S11, both the dorsal spurred petals and lateral f lattened petals undergo rapid development, with progressive spur elongation specifically observed in the dorsal spurred petals (Fig. 1D).In S12, the spurs of the dorsal petals have doubled in length compared to S11, and the lateral petals start asymmetric folding (Fig. 1D).By S14, the petals form their final structure, and pigmentation gradually begins (Fig. 1D).

Floral organ identity genes associated with petal morphological differentiation
To understand the molecular mechanisms underlying the morphological differentiation of petals in D. anthriscifolium flowers, we cloned putative A-and B-function genes implicated in specifying the petal identity.As a result, three, three, and two AGL6, AP3, PI lineage genes were obtained, respectively.Phylogenetic analysis revealed that the three lineage genes all underwent duplication events (Fig. 2A).Specifically, two successive Delphinieae-specific duplication events led to the generation of the DeanAGL6-1a, DeanAGL6-1b, and DeanAGL6-2.For DeanAP3 lineage, DeanAP3-3, together with its two closest paralogs, DeanAP3-1 and DeanAP3-2, were generated through two gene duplication events predating the divergence of the Ranunculaceae.In the case of DeanPI, independent duplication events occurred subsequent to the last common ancestor of all analysed species, generating DeanPI1 and DeanPI2 genes.
We next performed digital gene expression (DGE) analysis for the three lineage genes (Fig. 2B; Fig. S1, see online supplementary material).The results revealed that all three DeanAGL6-like genes exhibited higher expression levels in both sepals and petals relative to carpels and stamens across four developmental stages (S9, S10, S12, and S16) of f lowers.Notably, the DeanAGL6-1a gene exhibited significantly higher expression levels compared to the other two paralogs.For the DeanPI-like genes, the DeanPI1 was mainly expressed in sepals, petals, and stamens, while the DeanPI2 was exclusively expressed in petals and stamens.Moreover, the two genes demonstrated increased expression levels in petals through development.For the DeanAP3-like genes, the expression patterns of the three genes varied distinctly.DeanAP3-1 exhibited higher expression levels in both petals and stamens, particularly in stamens, while DeanAP3-2 showed its highest expression in stamens.DeanAP3-3, a key petal identity gene, was predominantly expressed in petals as expected [16,21,26].However, this gene drew our special attention due to its higher expression in spurred petals than in lateral petals during stages S9-S16, coinciding with distinct morphological changes (Fig. 2B; Fig. S1, see online supplementary material).Therefore, we consider that the expression level differences of DeanAP3-3 may be instrumental in the morphological differentiation of petals in D. anthriscifolium.

Effects of DeanAP3-3 on the morphological differentiation of petals
To verify the above hypothesis on DeanAP3-3 gene's role in petal differentiation in D. anthriscifolium, we firstly conducted in situ hybridization experiments to gain precise insights into its spatiotemporal expression.At S4 of f loral development, its expression was specifically detected in all petal primordia, with no obvious difference among different types of petals (Fig. 3A).This uniform expression pattern persisted until S6, prior to the eight petal primordia differentiated (Fig. 3B-C).By S8, with the differentiation of petal primordia along the dorsoventral axis, its expression remained high in the two dorsal and two lateral petals, whereas the ventral reduced petals showed only faint expression (Fig. 3D).At S9, coinciding with the onset of petal morphological differentiation, DeanAP3-3 still maintained high expression levels in the dorsal and lateral petals, but no expression signal was detected in the degenerated petals (Fig. 3E-F).During S10 and S11, the DeanAP3-3 gene continued to exhibit strong expression signals in the dorsal spurred petals but showed weakened expression signals in lateral f lat petals (Fig. 3G-I).By S12, the expression signal of the DeanAP3-3 gene in dorsal spurred petals was stronger than that in lateral f lat petals (Fig. 3J), aligning with the morphological differentiation of petals.
To further investigate the role of DeanAP3-3 in D. anthriscifolium, we utilized the virus-induced gene silencing (VIGS) technique with the Tobacco rattle virus (TRV) construct, TRV2-DeanPDS-DeanAP3-3, with DeanPDS as a silencing efficacy marker (Fig. 4   Plants with strong phenotypic changes showed a complete transformation of all the three types of petals-i.e., spurred, f lat, and reduced petals-into non-spurred sepals, with other f loral organs minimally affected (Fig. 4B; Fig. S3, see online supplementary material).Interestingly, f lowers with moderate phenotypic changes could be categorized into two groups.In the first group, the dorsal spurred petals were completely transformed into flat petals (Fig. 4C; Fig. S3, see online supplementary material).
In the second group, the dorsal spurred petals manifested as intermediates between sepals and f lat petals, and some ventral reduced petals converted into sepals (Fig. 4D; Fig. S3, see online supplementary material).In plants with weak phenotypic changes, only dorsal spurred petals showed a transition towards f lat petal characteristics, while the lateral f lat petals are minimally affected (Fig. 4E and F; Fig. S3, see online supplementary material).
Corresponding to these morphological shifts, significantly reduced AP3-3 expression levels were observed in petals with strong phenotypic changes compared to WT and mock (Fig. 4G), indicative of DeanAP3-3's pivotal role in identity determination of all the petals.However, it is interestingly to find that in the group of the weak phenotypic changes, the expression of DeanAP3-3 in dorsal petals that transformed from the spurred to f lat petals showed an intermediate level between those in the true dorsal spurred and lateral f lat petals (Fig. 4H).
To verify that dorsal spurred petals in plants with weak phenotypic changes indeed transitioned towards f lat petal characteristics, we performed SEM analysis and compared the epidermal cell types on the ventral side of the dorsal spurred petal and f lat petal in mock plants, as well as the dorsal petal in plants exhibiting weak phenotypic changes.In the proximal region (domain 2), all three petal types were characterized by elongated epidermal cells (Fig. 5A-C).In the distal region (domain 1), the epidermal cells of spurred petals in mock plants were elongated, whereas the f lat petals displayed conical cells (Fig. 5A and B).Interestingly, the dorsal petals of plants with weak phenotypic changes also had conical epidermal cells in the distal region, aligning with the cell type observed in f lat petals of mock plants (Fig. 5A-C).This suggests that in plants with weak silencing of the AP3-3 gene, the dorsal spurred petals undergo a transformation into f lat petals.

Discussion
In angiosperms, AP3-like genes are pivotal in the determination of petal and stamen identities.Within the Ranunculaceae family, the AP3 genes have expanded via two duplication events, yielding three distinct copies, with AP3-3 being crucial for specifying petal identity [4, 11, 16, 20-24, 26, 27].Unlike the broader expression of AP3-1 and AP3-2, AP3-3 is uniquely expressed in petals, often at significantly higher levels [4,23,24].In D. ajacis, the DeajAP3-3 is expressed in all the petal primordia before their morphological differentiation at earlier stages, but is exclusively expressed in the dorsal spurred petals at later stages [4,28].Our study in D. anthriscifolium largely corroborates these findings, revealing a preferential and persistent expression of AP3-3 in the petals, and most notably in the dorsal spurred and lateral f lat petals throughout the developmental stages (Figs 2B and 3).
Despite the conservativeness of the AP3-3 expression in Delphinium, their functions showed not only the similar but also distinct roles.For instance, in D. ajacis, knockdown of DeajAP3-3 resulted in the transformation of dorsal petals into doublespurred sepals [4].Similarly, in our study, the pivotal role of AP3-3 in petal identity is further supported by VIGS experiments, where silencing of DeanAP3-3 induced a complete morphological reversion of petals to non-spurred sepals (Fig. 4B; Fig. S3, see 1.8 Relative expression Relative expression  online supplementary material).These results suggest that the AP3-3 orthologs have conserved roles in the determination of the dorsal petal identity.Despite the functional conservation of AP3-3, it exhibits some non-conservative feature across different species.For instance, in D. ajacis, knockdown of DeajAP3-3 resulted in the transformation of the fused single-spurred dorsal petal into a fused double-spurred sepal [4], while in DeanAP3-3 silenced D. anthriscifolium f lowers, the spurred petals transformed into nonspurred sepals, suggesting a potential association with CYC-like genes that determine f loral symmetry.In addition, the D. ajacis In plants, alterations in gene expression patterns in time, space and amount play a pivotal role in the morphological diversification of both vegetative organs (such as roots, stems, and leaves) and reproductive organs (such as f lowers in angiosperms).For instance, the expression of Class 1 KNOTTED1-LIKE HOMEOBOX (KNOX1) genes after leaf primordia initiation determines the formation of simple or compound leaves [29,30].Similarly, in various species within the Liliaceae, Orchidaceae, and Ranunculaceae families, the expansion of expression domains of AP3 and PI genes can lead to the development of petaloid structures in the outer whorls of f loral organs [13,31,32].Among the various changes in gene expression patterns, changes in expression levels may be the most significant, but least understood [33].In Arabidopsis and its close relatives, differences in the expression levels of the SHOOT MERISTEMLESS (STM) gene explain variations in leaf serration or absence among different species [34].In the Ranunculaceae, the presence or absence of petals is closely associated to the expression levels of the AP3-3 gene, which exhibits specific and high expression in species with petals, while in species without petals, it is either not expressed or expressed at low levels [16,24,26].
Interestingly, as a petal identity gene, the changes in the AP3-3 gene expression levels also significantly inf luence the morphological differentiation of petals.For instance, in Nigelleae species, a close relative of Delphinieae, the AP3-3 gene is consistently expressed at a similar level in all eight non-spurred petals, indicating its role in maintaining a consistent petal phenotype [11].Complete AP3-3 silencing results in all the petals transforming into sepals, while partial silencing produces a range of petal phenotypes, from an intermediate to sepal-like or petal-like forms, suggesting a connection between morphological changes in Nigella petals and AP3-3 expression levels [11].Within Delphinieae, AP3-3 expression gradients are thought to contribute to the petal differentiation along the dorsoventral f loral axis, with species like Delphinium exaltatum and Staphisagria picta exhibiting higher AP3-3 expression in dorsal spurred petals [4,24,28].In this study, we also confirmed that the expression level of the DeanAP3-3 gene is higher in dorsal spurred petals compared to lateral f lat petals.From the S10-S12, a critical phase for morphological differentiation between dorsal and lateral petals, the DeanAP3-3 gene consistently maintained higher expression levels in spurred petals than f lat petals.No expression signal was detected in degenerated petals during this period (Fig. 3E-F).In addition, when the DeanAP3-3 gene is not completely silenced, dorsal spurred petals transform into f lat petals, or take an intermediate form between sepals and f lat petals, or resemble f lat petals, while ventral reduced petals transform into sepals (Figs 4C-F and 5A-C; Fig. S3, see online supplementary material).Therefore, we can conclude that the morphological differentiation of D. anthrisfolium petals is inf luenced by changes in AP3-3 gene expression levels.

Plant materials and growth conditions
Seeds of D. anthriscifolium were collected from Taibai Mountain, Shaanxi, China, and sowed in a soil mixture (vermiculite: nutrient soil = 1:1).The plants were cultivated under controlled conditions, with a photoperiod of 14 hours of light at 20000 lux and 10 hours of darkness, at a temperature of 24 • C and a relative humidity of 50-60%.

Morphological and developmental observations
Typical mature f lowers were photo-documented using an EOS 760D digital camera (Canon, Tokyo, Japan) paired with a Canon EF-S 60 mm f/2.8 lens.Floral organs were dissected under an Olympus SZX16 stereomicroscope (Olympus, Tokyo, Japan) and imaged with an OLYMPUS SC180 digital camera.Micromorphological observations were performed as previously described [4].
For scanning electron microscopy (SEM) analysis, petals were dissected under a stereomicroscope, followed by dehydration through a graduated series of water-ethanol solutions.Subsequently, they were dried using a CO 2 critical-point dryer.Once prepared, the samples were mounted on aluminum stubs, subjected to gold sputter coating, and observed under a S-3400 N scanning electron microscope (Hitachi, Tokyo, Japan).

Gene isolation and phylogenetic analysis
Total RNA was extracted from f loral buds using the SV Total RNA Isolation System (Promega) and reverse-transcribed into cDNA using the thermo scientific RevertAid First Strand CDNA Synthesis Kit.The CDS sequence of DeanAP3, DeanAGL6, and DeanPI lineage genes was then amplified with specific primers (Table S1, see online supplementary material).The PCR products of the expected lengths were inserted into the pEASY ® -Blunt E1 cloning vector (TransGen) and confirmed by sequencing.
Candidate AP3-like, AGL6-like, and PI-like genes of other species were obtained from the publicly available databases through BLAST searches (File S1, see online supplementary material).Nucleotide sequences for each gene lineage were aligned and manually adjusted using MEGA X [35].The alignable nucleotide matrix was utilized for phylogenetic reconstruction in MEGA X, employing the maximum-likelihood method with the best-fit model of JTT-G.Analysis was carried out using 1000 bootstrap replicates.Trees were rooted with genes from early diverging angiosperms.

Gene expression studies
The reference transcriptome and digital gene expression (DGE) profiles were obtained from published data (GSE249849) [36].The DGE profiles included 28 samples with three biological replicates each.These samples encompassed seven organ types (dorsal spurred petals, lateral non-spurred petals, dorsal spurred sepals, lateral sepals, ventral non-spurred petals, stamens, carpels), each collected at four developmental stages (S9, S10, S12, S16).Three biological replicates of each were collected.The clean reads of DGE profiles were separately mapped to the reference transcriptome we previously assembled by TopHat2 and fragments per kilobase per million mapped reads (FPKM) values were calculated using RSEM [37,38].
The expression patterns of DeanAP3-3 were revealed through mRNA in situ hybridization.Floral buds at various developmental stages of D. anthriscifolium were fixed in fresh 4% paraformaldehyde and embedded in Paraplast Plus (Sigma) [11].The antisense probe for DeanAP3-3 was synthesized via RNA transcription using the DeanAP3-3-R primer containing the T7 promoter in conjunction with the DeanAP3-3-F primer (Table S1, see online supplementary material).

Figure 1 .
Figure 1.Floral structures of the Delphinieae and the petal differentiation in Delphinium anthriscifolium.(A-C) Front views (first column) and side views (second column) of mature representative Delphinieae species f lowers, along with disassembled f loral organs: sepals (third column), petals (fourth column), and stamens with carpels (fifth column).Perianth diagrams in the last column illustrate f lower structure.(D) Petal differentiation in Delphinium anthriscifolium f loral buds from stages S9 to S14, showing side views of f loral buds (top), and front views of dorsal spurred petals (DPe), lateral petals (LPe), and reduced petals (RPe, indicated by white five-pointed stars).Scale bars: 1.0 mm.

Figure 2 .
Figure 2. Phylogenetic analysis and spatiotemporal expression patterns of putative A and B function genes in Delphinium anthriscifolium.(A) Phylogenetic trees for AGL6, AP3, and PI gene lineages.Genes from Delphinium anthriscifolium are highlighted.Branch lengths represent the number of nucleotide substitutions per site.(B) Spatiotemporal expression patterns of AGL6, AP3, and PI lineage genes as determined by DGE analysis.The bottom right inset depicts sepals, petals, stamens, and carpels across developmental stages S9-S16, with color intensity indicating the stage.Ca, Carpel; DPe, dorsal spurred petal; Dse, dorsal spurred sepal; LPe, lateral petal; Lse, lateral sepal; St, Stamen; Vse, ventral sepal.

Figure 3 .
Figure 3. Spatiotemporal expression of DeanAP3-3 in Delphinium anthriscifolium f lowers.Serial sections of the same f loral bud or petal are labeled as serial numbers.The images F, G, I, and J shows the transverse section of the f loral bud, while the remaining images display longitudinal sections.Expression models of DeanAP3-3 gene are showed in the upper right corner of each set of pictures, with the color illustrating the expression domain.The final images (K) show the results using sense probes.DPe, dorsal spurred petal; LPe, lateral petal; RPe, reduced petal.Scale bars: 200 μm.

Figure 4 .
Figure 4. Function of DeanAP3-3 as revealed by virus-induced gene silencing.(A-F) Flowers in the mock (A), TRV2-DeanPDS-DeanAP3-3 treated plants with strong (B), moderate (C and D), and weak (E and F) phenotypic changes.Each panel, from left to right, presents the front and side views of the whole f lower, followed by dissections revealing the anatomical structure of sepals, petals (with reduced petals marked by a white five-pointed star), and stamens and carpels.The f loral diagram in the last column illustrates the altered structure of f lowers.(G) Relative expression levels of DeanAP3-3 in S12 f loral buds of WT, mock and TRV2-DeanPDS-DeanAP3-3 treated plants.AP3-3 1-3 are f lowers with strong phenotypic changes, and AP3-3 4 are f lowers with moderate phenotypic changes.(H) Relative expression levels of DeanAP3-3 in dorsal (DPe) and lateral petals (LPe) of mock plants, and in lateral petals (D-LPe, dorsal petals transformed) of TRV2-DeanPDS-DeanAP3-3 treated plants.Scale bars: 1000 μm.

Figure 5 .
Figure 5. Morphological characteristics in Delphinium anthriscifolium petals.(A) Dorsal petal of a mock-treated plant with close-ups of the epidermal cells in the distal (A1) and proximal (A2) regions, as indicated by square regions 1 and 2, respectively.(B) Lateral petal of a mock-treated plant with close-ups of the epidermal cells in the distal (B1) and proximal (B2) regions, as indicated by square regions 1 and 2, respectively.(C) Dorsal petal of a plant exhibiting weak phenotypic changes, with detailed views of the distal (C1) and proximal (C2) epidermal cells, as indicated by square regions 1 and 2, respectively.Scale bars: 100 μm