Working smarter, not harder: silencing LAZY1 in Prunus domestica causes outward, wandering branch orientations with commercial and ornamental applications

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 branch and petiole angles. LAZY1-antisense lines also displayed 'wandering' or weeping branch trajectories. These phenotypes were not associated with decreases in branch strength or stiffness. We evaluated the utility of LAZY1-antisense trees for use in two planar orchard systems by training them according to super slender axe and espalier methods. We found that the LAZY1-antisense trees had more open canopies and were easier to constrain to the trellis height. This work illustrates the power of manipulating gene expression to optimize plant architecture for specific horticultural applications.

Figure 1

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.

Figure 2

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.

Figure 3

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.

Figure 4

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.

Figure 5

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.

Figure 6

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.

Figure 7

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 Genotype^*Timepoint combination.