Working smarter, not harder: silencing LAZY1 in Prunus domestica causes outward, wandering branch orientations with commercial and ornamental applications
Andrea R Kohler, Courtney A Hollender, Doug Raines, Mark Demuth, Lisa Tang, Macarena Farcuh, Chris Dardick
TL;DR
Silencing the LAZY1 gene in plum trees changes branch orientation, making them easier to manage in orchards and potentially improving fruit production.
Contribution
Silencing LAZY1 in Prunus domestica produces trees with outward, wandering branches suitable for horticultural applications.
Findings
LAZY1 silencing significantly increased branch and petiole angles in transgenic plum trees.
LAZY1-antisense trees displayed 'wandering' or weeping branch trajectories without reduced branch strength.
These trees were easier to train in planar orchard systems like super slender axe and espalier.
Abstract
Controlling branch orientation is a central challenge in tree fruit production, as it impacts light interception, pesticide use, fruit quality, yield, and labor costs. To modify branch orientation, growers use many different management practices, including tying branches to wires or applying growth regulator sprays. However, these practices are often costly and ineffective. In contrast, altering the expression of genes that control branch angles and orientations would permanently optimize tree architecture and reduce management requirements. One gene implicated in branch angle control, LAZY1, has potential for such applications as it is a key modulator of upward branch orientations in response to gravity. Here, we describe the phenotypes of transgenic plum (Prunus domestica) trees containing an antisense vector to silence LAZY1. We found that LAZY1 silencing significantly increased…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7- —Michigan State University, The Michigan State Horticultural Society, The United States Department of Agriculture National Institute of Food and Agriculture HATCH project 1013242 (C.A.H.), MSU AgBioRes
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsPlant Reproductive Biology · Plant Molecular Biology Research · Plant Physiology and Cultivation Studies
Introduction
Controlling the orientation of lateral organs is crucial for plant survival in response to changing light environments and competition with other plants. In a horticultural or agricultural context, lateral organ angles impact essential traits such as light interception (through positioning of branches and leaves), drought tolerance (through positioning of lateral roots), and harvest method (through positioning of fruit). Wide branch angles are considered desirable in tree fruit production; they generally reduce bark inclusion, leading to stronger branch unions and an increased ability to support fruit load [1]. Highly branched canopies with wide angles are particularly desirable in some types of high-density, commercial training systems where the focus is on producing a narrow and uniform canopy that can support a high crop load.
Due to the impacts of branch angle on production, growers often expend considerable labor trying to control the position of branches through pruning, tying, and applying growth regulators [2, 3]. However, as the angle of lateral organs is largely determined by genetics, these attempts can be futile. Some species, such as Prunus domestica (European plum), exhibit highly vigorous, upright growth habits, making it particularly difficult to train them in high-density planar systems [2]. In contrast to cultural practices, using selective breeding or gene editing to manipulate expression of genes that control lateral organ orientation can provide a permanent solution to branch angle control throughout the lifetime of the plant [2].
One of the primary gene families known to control lateral organ angle is the IGT family, whose members have a conserved (GϕL(A/T)IGT) amino acid motif [4, 5]. The family includes TAC1 homologs, which promote outward lateral organ orientation in shoots, and LAZY1 homologs, which promote narrower lateral organ angles [5–18]. Most species have multiple paralogs of LAZY, which can be further divided into three subclades: LAZY1-like, DEEPER-ROOTING1-like (DRO1-like), and LAZY5-like [5]. These paralogs are partially redundant in controlling shoot and root angles but show distinct expression patterns. LAZY1 homologs are the most influential in shoots, while the DRO1-like genes predominantly regulate lateral root angles.
LAZY homolog mutants across plant clades have wider lateral organ angles [5–15]. They also exhibit slowed or absent gravitropic response, sometimes to the point that shoots grow prostrate on the ground—as in Oryza sativa (rice) [15], Zea mays (maize) [10], and the Arabidopsis thaliana (Arabidopsis) lazy1,2,4 mutant [12]. In some species, lazy mutants also have roots that are negatively gravitropic, as in Medicago truncatula [11], Lotus japonicus [13], and the Arabidopsis lazy2,3,4 multiple mutant [11]. This suggests that LAZY proteins promote narrowed lateral shoot and root angles in response to gravity. Consistent with this, overexpression of some LAZY homologs led to narrower branch and root angles and enhanced gravitropism [16–18]. LAZY homologs also are involved in responses to light quantity and quality, likely functioning to integrate light and gravity signaling during the establishment of the default lateral organ angle or ‘set-point angle’ [14, 19, 20].
The function of LAZY proteins in the gravitropic pathway has been extensively explored during the last two decades. Upon gravistimulation, LAZY proteins are phosphorylated by the kinases MKK5 and MPK3 [21]. This promotes the binding of LAZY proteins to TRANSLOCON AT THE OUTER CHLOROPLAST ENVELOPE (TOC) proteins on the outside of amyloplasts [21]. As a result, LAZY proteins become enriched at the plasma membrane at the lower side of gravity-sensing cells (statocytes) as the amyloplasts sediment [21, 22]. Once polarized to the lower side of the cell, LAZY proteins recruit PIN3 auxin efflux proteins, which control polar auxin transport during gravitropism and development [23]. This recruitment is likely through interactions with RCC1-like domain (RLD) proteins [23]. The PIN3 proteins transport auxin laterally in stems and roots to establish gravitropic auxin gradients with greater auxin concentrations on the lower side of gravistimulated roots and shoots. In gravistimulated Arabidopsis lazy1,2,4 triple mutants, PIN3 mislocalizes to the upper side of statocytes, resulting in an inverted auxin gradient [24].
Here, we present results from a long-term phenotypic study of P. domestica (European plum) trees that were transformed with a LAZY1 silencing construct. This work was initiated >12 years ago for horticultural purposes after identifying TAC1 as a Prunus persica (peach) branch orientation regulator and grouping it with LAZY1 in the IGT family [4]. Our goal was to test if reduced LAZY1 expression would result in trees with wider branch angles and more outward orientations. The silencing construct used sequence from peach LAZY1 (Prupe.1G222800) because an annotated genome for P. domestica, a hexaploid with a complex lineage, did not exist at that time. A later analysis supports the specificity of our silencing sequence to plum LAZY1 alleles and not other LAZY/DRO plum genes.
Our data includes qualitative and quantitative greenhouse and field data for LAZY1-antisense plums on their own roots and on commercial rootstocks. We report on how silencing LAZY1 influenced plum branch angles, growth trajectories, and branch orientations. In connection, we present the results of a tree-training study that tested the utility of LAZY1-antisense trees for planar (2D) commercial orchard systems. Further, we report novel phenotypes related to LAZY1 silencing impacting wood development and photosynthesis that have important implications for use of such mutants in agricultural applications.
Prunus domestica (plum) trees transformed with a LAZY1-antisense silencing construct exhibited reduced expression of LAZY1 along with wider petiole and branch angles. (A) LAZY1 gene expression in control and LAZY1-antisense lines. Expression was determined by qPCR on at least three biological replicates (trees) per line, each with three technical replicates. Values are in relation to a standard curve of known RNA from control plants. (B) Average branch and petiole angles for LAZY1-antisense and control lines. Both petiole and branch angles represent averages from 3 to 10 trees per line. Diagram in the upper right indicates that angles reported are those between the branch or petiole and the apex of the shoot from which it emerged. Bars represent standard deviation and * indicates a significant difference between the control plants and a LAZY1-antisense line (P < 0.05) according to a Student’s t-test. For A and B, control trees were plum seedlings from same cultivar that did not contain the LAZY1-antisense vector.
Results
A silencing sequence from peach is specific to PdoLAZY1 alleles
European plum, like peach, has two independent LAZY genes (LAZY1 and LAZY2), and two DRO genes (DRO1 and DRO2) (Supplementary Table S1, Supplementary Fig. S1). European plum is a hexaploid species with unknown origin but is thought to be an allopolyploid [25]. Consequently, European plum could have up to six alleles of each LAZY homolog. Between four and six alleles were identified for each gene (Supplementary Table S1) [26]. Variation in number of alleles could be due to the collapse of homologous regions in the genome assembly. For LAZY1, we identified six distinct alleles, which we named PdoLAZY1A through PdoLAZY1F (Supplementary Table S1, Supplementary Fig. S1). The amino acid sequences for the first four LAZY1 alleles (PdoLAZY1A, PdoLAZY1B, PdoLAZY1C, and PdoLAZY1D) were identical to the peach (PpeLAZY1) sequence (Supplementary Fig. S2). The amino acid sequence for the fifth allele (PdoLAZY1E) had a single substitution (T48N). The sixth LAZY1 allele (PdoLAZY1F) has a 278-amino acid truncation, as shown in Supplementary Fig. S2. This truncation may be a natural mutation or possibly an assembly error.
We then analyzed the plum LAZY and DRO sequences to check the specificity of the antisense construct we derived from the PpeLAZY1 sequence (Supplementary Fig. S3). The 306-nucleotide silencing sequence had >98% identity with each of the five full-length LAZY1 plum alleles (Supplementary Table S2 and Supplementary Fig. S4). In contrast, there was minimal overlap between the silencing sequence and LAZY2, DRO1, and DRO2 alleles. Percent identity between the alleles of all three genes and the silencing sequence ranged from ~14% to 56% (Supplementary Table S2). Additionally, there were no stretches of ≥20 nucleotides that aligned to the LAZY1 silencing sequence (Supplementary Fig. S4; Supplementary Table S2). Thus, it is likely that our silencing vector is highly specific for LAZY1.
LAZY1-antisense trees have wider branch angles and undirected lateral branch growth.
Seven independent plum lines containing the LAZY1-antisense vector were generated using agrobacterium-mediated hypocotyl transformation (see methods). PdoLAZY1 expression was quantified by quantitative polymerase chain reaction (qPCR) using primers with 100% sequence identity to the full-length PdoLAZY1 alleles, but several mismatches with the nontarget LAZY/DRO alleles (Supplementary Fig. S5). In six out of the seven lines (Lines 1, 3, 4, 5, 6, and 7) LAZY1 expression was significantly reduced (Fig. 1A), and trees from those lines had wider petiole and primary branch (crotch) angles (Fig. 1B). Those lines also had dramatically more horizontal (outward) branch orientations in both greenhouse and field environments (Fig. 2). Line 2 trees did not have reduced LAZY1 expression, and their branch orientations were upward similar to controls (Figs 1 and 2). As LAZY1-antisense trees aged, their branches displayed a mix of wandering, arching, weeping, and outward branch phenotypes (Fig. 2B and C).
Shoots from grafted LAZY1-antisense tree buds did not exhibit upward growth
In preparation for our tree-training field studies, dormant vegetative buds from a representative LAZY1-antisense line (Line 4) and a ‘Stanley’ Open Pollinated (OP) plum tree (serving as a control for seedling variation among standard trees) were grafted onto ‘Myrobalan’ plum seedlings in the greenhouse. In some cases, rootstock interactions can influence scion architecture. After bud break, shoots emerging from the LAZY1-antisense buds grafted on ‘Myrobalan’ rootstock grew horizontally, as shown in Fig. 3 (black arrows). This indicated that mobile signals from the rootstock could not rescue the branch phenotype. Furthermore, new vegetative shoots growing from the rootstock stem above the graft union were upward orientated. Thus, silencing LAZY1 in the branch below did not induce a branch phenotype in shoots growing above the graft union (Fig. 3). The inability of LAZY1-antisense shoots to exhibit normal upward growth (negative gravitropism) when grafted to a wild-type rootstock indicates that these trees can be used on conventional rootstocks without significantly altering their growth habits.
LAZY1-antisense lines show unique characteristics for planar training and revealed an impaired shoot gravitropism phenotype
To test performance of LAZY1-antisense plum trees in planar training systems, the LAZY1-antisense Line 4 trees on ‘Myrobalan’ rootstock were trained in the horizontal palmette espalier system used in fruit trees for hundreds of years [27], as well as the super slender axe (SSA) system designed for sweet cherry (Figs 4 and 5) [28]. Trees were trained and observed over the course of 5 years. Tree trunks (which originated as a shoot from a grafted bud) were forced to maintain a vertical orientation by tying them to support stakes and trellis wire at the time of planting and regularly throughout each season.
The trunks from the ‘Stanley’ OP controls consistently grew straight up (Fig. 4). These trees generally sent up a single leader with limited lateral branches, which rapidly grew above the trellis (Fig. 4). Heading this leader provoked a strong flush of growth at the top of the tree, shading the lower tree and draining vigor away from where it is desired lower in the tree (Fig. 4). In contrast, the trunks of the LAZY1-antisense trees wandered side-to-side and/or arched downward at sites where they were tied to trellis wire or their supportive stake (Fig. 4). Because LAZY1-antisense trunks had to be staked upright for vertical growth, they did not grow significantly taller than the trellis (Fig. 4). Further, heading the main trunk did not provoke a flush of vegetative growth at the top of the tree (Fig. 4).
The actively growing branch tips of lateral branches from control trees trained to maintain a 90-degree orientation in the espalier study were consistently oriented upward, exhibiting classic negative gravitropism (Fig. 5A). Secondary shoots growing from lateral buds on control tree branches also exhibited upward growth (Fig. 5A). However, LAZY1-antisense tree branch shoot tips and secondary shoots did not show any clear gravitropic response or consistent directional growth (Fig. 5B). In addition to growing in every direction, secondary shoots from the horizontal LAZY1-antisense branches emerged from buds on every side of the tied branches, and not just buds on the upper side, as they did in control trees (Fig. 5B).
LAZY1-antisense plums exhibit altered leaf and branch orientations. (A) Representative control and LAZY1-antisense trees (1–2 years old) from each transgenic line growing in the greenhouse. Each line had a minimum of five trees and a maximum of 13. LAZY1 expression was not significantly reduced in Line 2. (B) A mature LAZY1-antisense Line 6 tree growing in the field. (C) A LAZY1-antisense Line 6 tree growing the greenhouse, demonstrating the wandering growth trajectory.
Shoots from grafted LAZY1-antisense buds grew horizontally from standard rootstock. Representative plum tree with a Line 4 LAZY1-antisense branch (black arrow) growing from a bud that was grafted onto a standard ‘Myrobalan’ plum rootstock. Vertical shoots emerging from rootstock buds are indicated by the white arrow.
Representative ‘Stanley’ OP and LAZY1-antisense Line 4 on ‘Myrobalan’ rootstock trained into planar systems. (A) Espalier training. (B) SSA training. Note that ‘Stanley’ OP trees trained as Espalier and SSA grew past the top (fourth) trellis wire by August 2021 while the LAZY1-antisense trees grew along it and downwards. ‘Myro’ indicates ‘Myrobalan’ rootstock.
Lateral branch phenotypes of LAZY1-antisense trees. (A) When ‘Stanley’ OP branches were tied horizontally for espalier training, they continued trying to reorient upwards and buds broke from the top of the branch. (B) LAZY1-antisense branch tips did not reorient upwards, and buds broke from all sides of the branch.
Branches from LAZY1-antisense trees do not have reduced stiffness or strength, but their wood has altered material properties.
Xylem differentiation and development are associated with polar auxin transport [29], and LAZY proteins influence auxin transport in response to gravity [8]. Therefore, we investigated the biomechanical properties of wood from our LAZY1-antisense plum trees. To do this, we performed materials testing experiments on comparable new growth and 1-year-old branches from the ‘Stanley’ OP control and LAZY1-antisense Line 4 trees on the ‘Myrobalan’ rootstock (Fig. 6A). We found that compared to controls, the LAZY1-antisense tree branches did not have any alterations in their ability to resist bending (flexural stiffness; EI) (Fig. 6A and B). Nor was there a change in the maximum force that their branches could resist (F_max_) (Fig. 6A and C). Thus, the wandering phenotype of the branches is not due to floppiness, or an inability to hold themselves upright. Despite branch stiffness (EI) and strength (F_max_) remaining constant, the material properties of the wood from LAZY1-antisense branches were different than wood from ‘Stanley’ OP branches. The 1-year-old LAZY1-antisense wood had a lower modulus of elasticity (MOE) indicating that it was more flexible (Fig. 5D). It also had a lower modulus of rupture (MOR), which indicates weakness, or a greater likelihood that it would break from a given force (Fig. 6E). However, the lower MOE and MOR did not reduce the overall stiffness or strength of the LAZY1-antisense branches because they had increased area moment of inertia (I), which is a measure of cross-sectional area, accounting for how the area is distributed relative to the force; Fig. 6F).
Biomechanical properties of current year growth and 1-year-old branches from LAZY1-antisense trees. (A) Diagram illustrating how biomechanical properties relate to bending force and failure. (B–F) Biomechanical properties of LAZY1-antisense Line 4 branches compared to ‘Stanley’ OP control branches: (B) flexural stiffness (EI); (C) maximum force (Fmax); (D) modulus of elasticity (MOE); (E) modulus of rupture (MOR); and (F) area moment of inertia. Bars represent standard error. Branches taken from four trees per genotype. n = 30 per genotype for new growth, n = 13 for ‘Stanley’ OP first year, and n = 14 for LAZY1-antisense first year. Comparisons between genotypes done with pairwise t-tests. * indicates significantly different at α = 0.10, ** indicates significant at α = 0.05, *** indicates significant at α = 0.01, n.s. indicates not significant at α = 0.10.
LAZY1-antisense trees exhibit leaf chlorosis and reduced net photosynthesis.
Toward the end of each growing season, we observed leaf chlorosis in field-grown LAZY1-antisense lines growing in both Kearneysville, WV, and Clarksville, MI (Fig. 7A–D). The phenotype was somewhat variable. The extent and timing of the chlorosis depended on the line, year, and location, suggesting a Genotype-by-Environment interaction. For example, chlorosis was regularly observed in Lines 1, 5, 6, and 7, but generally not in Line 4 trees (Fig. 7D). To quantify the chlorosis, relative chlorophyll content was measured for all lines in 2021 (Fig. 7A). We also investigated if the chlorotic tendency impacted photosynthesis. In both spring and fall 2022, net photosynthesis for Line 4 and Line 6 trees growing in Michigan was measured. Consistent with the chlorophyll measurements, net photosynthesis was reduced in Line 6, but not in Line 4 (Fig. 7E). Interestingly, Line 6 trees had decreased photosynthesis at both time points, despite their leaves not showing chlorosis at the spring time point.
LAZY1-antisense trees had reductions in chlorophyll content and net photosynthesis (A) Chlorophyll content of LAZY1-antisense lines in Kearneysville, WV. The ˜ serves as a reminder that Line 2 did not have significant reduction in LAZY1 expression. * Indicates P < 0.05. (B) A ‘Stanley’ OP control and a LAZY1-antisense Line 6 tree growing in Clarksville, MI. (C) Leaf chlorosis comparison for Stanley and LAZY1-antisense Line 6 in Kearneysville, WV. (D) Leaf photos taken fall 2021 in Clarksville, MI. (E) Net photosynthesis in spring versus fall for 2022 growing season in Clarksville, MI. Means within the same time point with the same letter are not significantly different at α = 0.05. Thirty or more measurements were taken for each GenotypeTimepoint combination.*
LAZY1-antisense plums produced normal fruit
After several seasons in the field, the LAZY1-antisense plum trees flowered and fruited in both West Virginia and Michigan. Due to the change in canopy shape and potential influence of the observed chlorosis, we investigated potential impacts on fruit quality. All lines produced flowers and fruit that were visually comparable to control trees (Supplementary Fig. S6). A preliminary fruit quality assessment was done for fruit from five lines growing in West Virginia in 2024 by determining their soluble solid content (SSC). The LAZY1-antisense lines exhibited a slight decrease in SSC (Supplementary Fig. S6), however the values were still within the normal range for plum seedlings [30].
Discussion
This long-term study began with plum tree transformations initiated in 2012, included two field studies, one planted in West Virginia starting in 2015, and another planted in Michigan in 2018, and ended with preliminary fruit evaluations in 2024. Our results demonstrated that reducing LAZY1 expression can effectively alter branch angle in a tree fruit crop. We observed LAZY1-antisense plums at all stages of development and under diverse conditions, bringing the observations closer to real-world application. LAZY1-antisense lines exhibited increased branch angles and orientations from the seedling stage through maturity. This phenotype was stable when grafted onto standard rootstock, grown in locations with different climates, and after various pruning regimes.
These LAZY1-antisense lines displayed multiple traits that are beneficial for commercial production. The wider branch angles resulted in more open canopies that could increase light penetration, improving flower bud development and fruit quality [31]. An unexpected benefit of LAZY1-antisense trees came from their more bush-like growth habits (Fig. 2B). The main trunk did not grow upright, there was no clear leader, and the trees grew more horizontally than vertically. This unexpectedly solved one of the central problems in planar systems, which is keeping the trees at or below the height of the trellis without stimulating excessive vegetative vigor by heading the tree [27]. The agravitropic growth noticeably facilitated planar training by limiting tree height to the highest point at which it was tied, decreasing bursts of localized vegetative vigor, and increasing uniformity of the canopy. Control of vigor is particularly important in European plum, as extremely vigorous shoots generally do not produce floral buds [3]. Crucially, LAZY1-antisense lines did not have reduced branch strength or stiffness, indicating they maintain the ability to support heavy crop loads.
These LAZY1-antisense lines also exhibited phenotypes that may limit their applicability for commercial production but could prove useful for ornamental applications. The undirected growth of LAZY1-antisense branches presents training challenges. For espalier with control trees, horizontal branches can be trained between wires via counteracting negative gravitropism (upward growth) by tying the branches to the wire below. As LAZY1-antisense trees do not have strong gravitropic responses, it was difficult to train branches between wires; they needed to be tied to the wires below and above to maintain their trajectory (Figs 4 and 5). From a training perspective, the problem can be solved by placing a trellis wire at each interval where horizontal growth is desired and tying the branch directly to it. The wandering habit also causes idiosyncratic changes in direction of the main trunk, which made it difficult to create a uniform canopy.
When allowed to develop freely, the wandering branch phenotype produces very attractive trees for ornamental use (Fig. 2B). Wandering branches, as in corkscrew willows and Young’s weeping birch, are prized for adding winter interest to landscaping. Interestingly, this phenotype has not previously been reported in LAZY1 knockdowns, even in woody species such as apple [14]. However, this phenotype may not be limited to plum. A correlative study suggested that the weeping trait in Young’s weeping birch (Betula pendula ‘Youngii’) is the result of a premature stop codon in a LAZY1 homolog [32]. While this study does not describe a wandering phenotype, it is clearly visible in photographs of this cultivar available from arboretums and nurseries. This twisting phenotype may be related to circumnutation, especially as previous studies have shown that rice lazy1 mutants have decreased circumnutation [33]. However, it is unclear why these windings are ‘frozen in time’ and remain visible in the woody branches.
The role of auxin transport in wood development and the changes we observed in the biomechanical properties of first-year branches from LAZY1-antisense trees may point to a role for LAZY1 in wood development. Alternatively, because the changes in biomechanical properties and diameter are consistent with what is observed in flexure wood or tension wood, these changes may be a secondary effect of the unique LAZY1-antisense branch orientation, with branches forming reaction wood and/or flexure wood in response to their horizontal trajectory.
Multiple LAZY1-antisense lines unexpectedly showed reduced chlorophyll content (Fig. 7) and leaf chlorosis. Chlorosis in at least one of these lines was associated with a decrease in net photosynthesis. While undesirable for production systems, their unique color may be a desirable trait for ornamental use. The variance of the phenotype among lines and seasons suggests that the trait may not have 100% penetrance, which may assist in selecting the desired phenotype. Leaf chlorosis has not been previously reported for lazy1 mutants. This may be because most studies on lazy1 mutants were short-term experiments using annual species grown in growth chambers. In plum, the chlorotic phenotype only became apparent midway through the growing season in the high light intensity of the greenhouse and field. Given recent revelations about the ability of LAZY1 protein to associate with the TOC1 complex found in amyloplasts and other plastid membranes, as well as its role in light response/signaling [21, 22], the leaf chlorosis may indicate a role for LAZY1 in the response of photosynthesis to light-related signaling, the absence of which results in chlorosis. This hypothesis is strengthened by the observation that photosynthesis is reduced in the spring in chlorotic Line 6, prior to any observable leaf chlorosis.
In the ongoing quest to produce tree fruit more efficiently, advances that improve tree structure and fruit yield while decreasing labor inputs will be crucial to success. Due to the substantial amounts of labor required to physically constrain and control tree canopies over the lifespan of the orchard, modifying the genetics controlling canopy structure will be a key part of long-term solutions. Here we have demonstrated that manipulation of LAZY1 expression can permanently alter canopy structure, providing traits beneficial to planar training systems. While some of the pleiotropic phenotypes we observed were deleterious to commercial production, this work provides a strong foundation for further investigation of genetic manipulation of branch angle in tree fruit.
Materials and methods
LAZY gene identification and phylogeny
LAZY genes in P. domestica were identified using BLASTp of the previously identified peach LAZY proteins [5] against predicted proteins from the P. domestica draft genome v1.0 (Fig. 1 and Supplementary Table S1) [26]. Due to a highly unusual exon–intron structure, LAZY genes are often misannotated [34]. This structure is conserved across species, with each gene containing five exons, with the first exon including just two codons (often an ‘ATGAAG’) and the last exon containing only ~20 bases [34]. To identify and correct errors in annotation, the DNA sequences of plum genes identified as LAZY homologs were aligned to the homologous peach and Arabidopsis sequences. Using that alignment, the plum and peach genes were reannotated with particular attention to identifying a gene model that fit the conserved gene structure. Following reannotation, the protein sequences were predicted, and the resulting protein sequences were used for subsequent alignments.
Alignments and phylogenetic trees were produced in CLC Genomics Workbench v22.0. Alignments were performed using a gap extension cost of 1.0, and a gap open cost of 10.0, the alignment set to ‘Very accurate’. The protein phylogenetic tree was generated using the Neighbor Joining method with 100 bootstrap replicates and Jukes–Cantor as the protein distance measure. For Supplementary Table S2, alignments were performed using Clustal Omega.
Cloning
To silence LAZY1 in P. domestica (European plum), hypocotyls extracted from ‘Stanley’ plum seeds were transformed with a silencing construct containing a 306-nucleotide sequence from the peach LAZY1 gene (PpeLAZY1; peach genome version 1.0 ID ppa007017, now named Prupe.1G222800 in genome version 2) in the reverse orientation behind a 35S promoter. A peach sequence was used because an annotated plum genome was unavailable at the time. After the publication of the European plum genome [26], LAZY and DEEPER ROOTING (DRO) homologs were identified by protein BLAST searches using peach LAZY and DRO sequences (see Supplementary Table S1; Supplementary Figs S1 and S2). The gene fragment was amplified using primers PpLazy-1 1F (5’AAGCCAAACTGTGGCACAAAGC) and PpLazy-1 2R (5’AGCTGCCAGGACTTTCTCCAA T), cloned into pENTR-D TOPO (Invitrogen, Carlsbad, CA), and from there into the pHellsgate 8.0 vector [35] using LR Clonase (Invitrogen, Carlsbad, CA). While the intended construct would have had two copies of the 306-bp fragment as a hairpin, the construct used for transformation was later discovered to only have a single insert in the reverse orientation behind the 35S promoter within pHellsgate 8.0, creating an antisense vector rather than an RNAi construct.
Plum transformation
The pHellsgate 8.0 vector with the PpeLAZY1 fragment was transformed into Agrobacterium tumefaciens GV3101 and then engineered into European plum (P. domestica L) according to Petri et al. [36]. Briefly, cold (4°C) stored seeds from OP ‘Stanley’ plum trees were cracked to remove their endocarp, surface sterilized with 15% bleach for 15 min and washed with sterile water three times. Hypocotyl slices were excised under a laminar flow hood using a stereomicroscope. After a 20-min incubation with the Agrobacterium suspension, the transformed hypocotyl sections were cultured in cocultivation medium for 3 days. The hypocotyl sections were then transferred to antibiotic (80 mg/l kanamycin) selection medium. Resulting kanamycin-resistant (transgenic) shoots were multiplied in plum shoot multiplication medium, rooted on rooting media, and then acclimated in a growth chamber before being transplanted into 15- to 23-cm pots in a greenhouse. Each line contains trees that originated from the same cluster of callus on a hypocotyl slice, and between 5 and 13 trees were generated per line.
Plant material
Greenhouse-grown LAZY1-antisense and control lines were used for quantifying gene expression and for branch and petiole angle analysis. Trees were 1–2 years old when gene expression was quantified and petiole angles were measured, and 2–3 years when branch angles were measured. The controls for LAZY1 expression were transformed plum seedlings from OP plum cultivar ‘Stanley’ (‘Stanley’ OP) that did not contain the LAZY1-antisense vector. These plums instead contained the pSUC/PSUL vector, which does not impact branch angle [37].
LAZY1-antisense plum trees and ‘Stanley’ OP controls were planted at the USDA-ARS Appalachian Fruit Research Station (AFRS) research field in Kearneysville, WV, and at the Michigan State University Clarksville Research Center (CRC) near Clarksville, MI. The trees at AFRS were planted at ~2.4-m spacing in October 2014. The LAZY1-antisense plums were planted in two blocks of three trees. Control plums were planted between groupings of LAZY1-antisense trees. The trees for training studies at CRC were generated in 2017 and 2018 from budwood from LAZY1-antisense Line 4 and ‘Stanley’ OP control taken from AFRS and grafted onto ‘Myrobalan’ rootstock. After a year and a half of greenhouse growth, the grafted trees (0.3–0.5 m in height) were planted in September 2018. For horizontal palmette espalier training, trees were planted at ~2.4-m spacing, and for SSA training they were spaced at ~0.9 m. These trees were trained to trellis with ~46-cm spacing between wires. The CRC planting also contains the LAZY1-antisense Line 6 trees on their own roots.
RNA extraction and gene expression analysis
The E.Z.N.A SQ Total RNA Kit (Omega Bio-tek, Inc., USA) was used to extract total RNA from frozen leaf tissue samples for LAZY1 expression analysis. RNA was treated with DNase I to remove contaminating genomic DNA. To determine gene expression levels, qPCR reactions were carried out using the gene-specific primers PpLAZY-qPCR-5F (5’ ATGCTTTATGCTTCTTCTCG) and PpLAZY-qPCR-5R (5’ TTGCTCAGCAGATGAGGT; Supplementary Fig. S3) and the Super Script III Platinum SYBR Green One-Step qRT-PCR Kit with ROX (Invitrogen Corp., USA). The qPCR was run using an ABI 7900DNA Sequence detector (Applied Biosystems) using the following protocol: cDNA synthesis step was performed at 50°C for 5 min, followed by a PCR starting with 95°C for 5 min, then 40 cycles of 95°C for 15 s and 60°C for 30 s, followed by 40°C for 1 min. The qPCR was performed on RNA from three to four independent biological replicates (trees) for each transgenic line, as well as on five control trees. Each biological replicate had three technical replicates for the reactions. A standard curve generated from serial dilutions of RNA from a tree that did not have the LAZY1 vector was run at the same time to determine relative expression values.
Petiole and branch angle measurements
Petiole angles for up to 10 leaves growing from the trunk of young (<1-year-old) trees that had not yet initiated lateral shoots were measured in 2013 using a protractor. For the control and Lines 3 through 7, five trees were measured. Ten trees were measured for Line 1, and three trees for Line 2. Crotch (branch) angles for six branches from 3 to 10 trees per genotype growing at AFRS were measured in 2021 using a protractor. Angles represent the angle between the branch and the shoot from which it emerged. When the branch angles were measured, the trees were ~9 years old and had been growing in the field for 6 years.
For both petiole and branch angle statistical analyses, the average angle per tree was considered a biological replicate, and standard deviations are based on those averages. A P-value <0.05 from Student’s t-test was used to determine significant differences between the LAZY1-antisense lines and control trees.
Fruit-soluble solids measurements
Between 15 August and 9 September 2024, 6–15 fruit were harvested from six Stanley trees and three to six trees for each LAZY1-antisense line (except Line 1, for which only one tree with fruit was available). A digital handheld refractometer (PAL-1, Atago, Tokyo, Japan) was used to measure Soluble Solid Content, and firmness was measured using a Lutron FR-5120 digital penetrometer with an 8-mm tip. The average brix for each line was modeled in R 4.4.1 using the ‘lme4’ package with the equation Brix = μ + Line + Tree + Firmness + e, where μ is the grand mean, ‘Line’ was the effect of the line, ‘Tree’ was the effect of the individual tree, ‘Firmness’ was a linear variable, and ‘e’ was the residual error. This model was gated using analysis of variance (ANOVA) and marginal means for lines were estimated using the ‘emmeans’ package. t-tests were performed against Stanley using the ‘multcomp’ package.
Biomechanics measurements
Clonal trees were generated from budwood from a single OP ‘Stanley’ (‘Stanley’ OP) tree and a single LAZY1-antisense Line 4 tree grafted onto ‘Myrobalan’ rootstock, grown at CRC and trained as horizontal palmette espalier. Horizontally oriented new growth and first-year wood (branches initiated the previous season) were collected 9 July 2023. Four to five branches per tree were collected from four trees per genotype for new growth, for a total of 20 ‘Stanley’ OP and 19 LAZY1-antisense samples. Three to five branches per tree were collected from three trees per genotype for first-year samples, for a total of 13 ‘Stanley’ OP and 14 LAZY1-antisense samples.
An Instron Universal Testing Machine (Model 4202, Instron Corporation, Canton, MA) with a 500 N load cell was used to conduct four-point bending tests. The span of the outer supports was set to 90 mm, while the two posts of the actuator were set to 30 mm. Leaves and lateral branches were removed from a section of wood ~10 cm long, and the section was oriented so that the upper side of the branch was positioned down on the Instron, so that force was exerted in the same direction as gravitropic force acted on the branch in its original position. The force versus displacement curve was measured until a maximum force was reached.
The equation EI = (F/V)(a^2^/12)(3L–4a) was used to quantify Flexural stiffness (EI), with (F/V) representing the slope of the linear section of the force/displacement curve, ‘a’ representing the post-to-load distance, and L represented the total length of the span. Flexural stiffness was then used to calculate the MOE, using the equation MOE = EI/I, where I is the area moment of inertia. For a branch with an elliptical cross section, I = (π/4) (R_P_)(R_L_^3^), where R_L_ is the radius in the direction of loading and R_P_ is the radius perpendicular to loading. The MOR was determined using the equation MOR = (1/2 Fmax)(a)(R_L_)/I, where Fmax was the maximum force withstood by the branch. A Student’s t-test was performed to test for statistically significant differences between the genotypes using the T.TEST function in Excel, assuming a homoscedastic two-tailed distribution.
Chlorophyll and photosynthesis measurements
Chlorophyll measurements were taken from field-grown trees at the AFRS with a SPAD 502 Chlorophyll Meter (Konica Minolta, Inc., Tokyo, Japan). For 2021, six leaves were measured per tree for 3–10 trees per genotype, on 25 August.
Photosynthetic parameters were measured in late spring (9 and 23 June 2022) and early fall (5 October 2022) using a CI-340 infrared gas analyzer (CID Bio-Science, Camas, WA). The measurements were taken with an open system in differential mode, with 6.25 cm^2^ of leaf area in the chamber, and a flow rate of 0.3 l/min. Measurements were taken from trees at CRC, including control trees (‘Stanley’ OP on ‘Myrobalan’ rootstock and on own root), LAZY1-antisense Line 4 on ‘Myrobalan’, and LAZY1-antisense Line 6 trees on their own roots. At each time point, three to five measurements were taken per leaf, on three to five leaves per tree, with two to six trees per genotype.
Statistical analysis was performed in R v4.3.1. To control for the environmental variability of the field conditions, net photosynthesis (Pn) for all genotypes was modeled as a response to photosynthetically active radiation (PAR), the concentration of carbon dioxide at intake (CO2in), air temperature (Tair), air pressure (Pressure), and water vapor pressure (H2Oin) at intake. Externally studentized residuals were used to identify influential outliers, and observations with a studentized residual >|2| were dropped (10 in total). Studentized residuals and the criterion Cooks Distance for point>3(mean Cooks distance) was used to identify outliers. Outliers were then manually checked. Outliers for which there were two or more outliers per leaf were retained as reflecting true biological variability, while single outliers were omitted. The model was then rechecked without outliers. Variance inflation factor (VIF) was used to check for collinearity of the environmental variables. Added variable plots were used to check for linearity and the contribution of each additional variable. Variables with the least contribution or the highest VIF were sequentially dropped until VIF < 3 for all variables, and all variables had a clear contribution to the model. Finally, variables for genotype and for accounting for the repeated measures were added to the model.
The resultant model for photosynthesis was Pn = μ + Genotype + Timepoint + GenotypeTimepoint + (1|Genotype:Tree) + (1|Genotype:Tree:Leaf) + PAR+ CO2 + Tair + Pressure + error where μ indicates the grand mean, ‘Genotype’ is a variable for the control or RNAi line, ‘Timepoint’ accounts for the effects spring versus fall, ‘Genotype^^Timepoint’ accounts for the interaction between genotype and time point, (1|Genotype:Tree) controls for the effects of tree, which is a random variable nested within genotype, (1|Genotype:Tree:Leaf) controls for leaf-to-leaf variation, and is nested in tree, and ‘PAR’, ‘CO2in’, and ‘Tair’ are continuous variables controlling for their respective environmental factor. This model was gated with ANOVA, and normality and equal variances of the residuals were checked. Since interaction between genotype and time point was significant, all pairwise comparisons were performed for genotype slicing by time point using t-tests at α = 0.05.
Supplementary Material
Web_Material_uhaf106
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Warner J . Rootstock affects primary scaffold branch crotch angle of apple trees. Hort Science. 1991;26:1266–7
- 2Quinlan JD, Tobutt KR. Manipulating fruit tree structure chemically and genetically for improved performance. Hort Science. 1990;25:60–4
- 3Cvetković M, Đurić G, Micic N. Canopy management practices in modern plum (Prunus domestica, L.) production on vigorous rootstocks. Scientific Papers Series B Horticulture. 2017;61:117–122
- 4Dardick C, Callahan A, Horn R. et al. Ppe TAC 1 promotes the horizontal growth of branches in peach trees and is a member of a functionally conserved gene family found in diverse plants species. Plant J. 2013;75:618–3023663106 10.1111/tpj.12234 · doi ↗ · pubmed ↗
- 5Waite JM, Dardick C. The roles of the IGT gene family in plant architecture: past, present, and future. Curr Opin Plant Biol. 2021;59:10198333422965 10.1016/j.pbi.2020.101983 · doi ↗ · pubmed ↗
- 6Hollender CA, Waite JM, Tabb A. et al. Alteration of TAC 1 expression in Prunus species leads to pleiotropic shoot phenotypes. Horticulture Research. 2018;5:2629736251 10.1038/s 41438-018-0034-1PMC 5928093 · doi ↗ · pubmed ↗
- 7Nakamura M, Nishimura T, Morita MT. Bridging the gap between amyloplasts and directional auxin transport in plant gravitropism. Curr Opin Plant Biol. 2019;52:54–6031446250 10.1016/j.pbi.2019.07.005 · doi ↗ · pubmed ↗
- 8Li P, Wang Y, Qian Q. et al. LAZY 1 controls rice shoot gravitropism through regulating polar auxin transport. Cell Res. 2007;17:402–1017468779 10.1038/cr.2007.38 · doi ↗ · pubmed ↗