Regulation of ARGONAUTE10 Expression Enables Temporal and Spatial Precision in Axillary Meristem Initiation in Arabidopsis

Abstract

Axillary meristems (AMs) give rise to lateral shoots and are critical to plant architecture. Understanding how developmental cues and environmental signals impact AM development will enable the improvement of plant architecture in agriculture. Here, we show that ARGONAUTE10 (AGO10), which sequesters miR165/166, promotes AM development through the miR165/166 target gene REVOLUTA. We reveal that AGO10 expression is precisely controlled temporally and spatially by auxin, brassinosteroids, and light to result in AM initiation only in the axils of leaves at a certain age. AUXIN RESPONSE FACTOR 5 (ARF5) activates while BRASSINAZOLE-RESISTANT 1 (BZR1) and PHYTOCHROME-INTERACTING FACTOR 4 (PIF4) repress AGO10 transcription directly. In axils of young leaves, BZR1 and PIF4 repress AGO10 expression to prevent AM initiation. In axils of older leaves, ARF5 upregulates AGO10 expression to promote AM initiation. Our results uncover the spatiotemporal control of AM development through the cooperation of hormones and light converging on a regulator of microRNA.

Keywords: ARF5; ARGONAUTE10; BZR1; PIF4; REVOLUTA; auxin; axillary meristem; brassinosteroids; light; microRNA.

Figure 1.. ago10 Mutants Show AM Defects

(A and B) Scanning electron micrographs of the basal regions of stage P₁₇ rosette leaves showing an axillary bud (solid arrow) in the leaf axil in wild type (A) and the absence of an axillary bud (dotted arrow) in the pnh-2 mutant (B). Bars, 100 μm. (C) Schematic representation of AM phenotypes in wild type (Ler and Col), pnh-2, ago10–13, and p35S::AGO10 plants. Note that pnh-2 and ago10–13 are in the Ler background while p35S::AGO10 is in the Col background. The black horizontal line represents the border between the youngest rosette leaf and the oldest cauline leaf. Each column represents a single plant, and each square within a column represents an individual leaf axil. The bottom row represents the oldest rosette leaf, with progressively younger leaves above. The color green indicates the presence of an axillary bud, yellow indicates the absence of an axillary bud and orange indicates more than one axillary bud in one leaf axil.

Figure 2.. AGO10 Is Expressed on the Adaxial Side of Young Leaves and in Axillary Meristems

Shoot apices of pZLL::YFP-ZLL zll-1 plants were sectioned and imaged for YFP-ZLL (YFP-AGO10) signals. The numbers indicate the developmental stages of leaves. Asterisks represent the SAM. Bars, 50 μm. (A) A longitudinal section showing that YFP-AGO10 signals are on the adaxial side of young leaves P₄. (B) A cross section showing YFP-AGO10 signals on the adaxial side of young leaves (P₅) and in the middle zone in older leaves (P₆ and P₈). YFP-AGO10 signals begin to appear on the adaxial side of P₉ (blue arrow). The green color represents YFP-AGO10 signals and the red color indicates auto fluorescence from chloroplasts. (C) A cross section showing YFP-AGO10 signals in the first initiated AM (arrow) in the axil of a P₁₁ leaf. (D) A cross section showing YFP-AGO10 signals in an axillary bud developed from an AM (arrow) in the axil of a P₁₅ leaf.

Figure 3.. AGO10 Promotes REV and STM Expression

(A and B) REV (A) and STM (B) expression levels in wild type, pnh-2, and an AGO10 overexpression line (p35S::AGO10) as determined by qRT-PCR. Error bars represent SD from three independent replicates. *p < 0.01, (Student’s t test). (C–F) Cross sections of shoot apices from pREV::REV-Venus (C and E) and pREV::REV-Venus pnh-2 (D and F). REV-Venus signals are present in the AM (arrow) at the axil of a P₁₂ leaf in wild type (C). REV-Venus signals are extremely low in pREV::REV-Venus pnh-2 throughout the apex including at the axil of a P₁₂ leaf (D). REV-Venus signals are present in a larger AM in the axil of a P₁₆ leaf in pREV::REV-Venus (E) but not in the axil of a leaf of a similar stage in pREV::REV-Venus pnh-2 (F). (G–J) Cross sections of shoot apices from pSTM::STM-Venus (G and I) and pSTM::STM-Venus pnh-2 (H and J) plants. STM-Venus signals are present in AMs (arrows) in the axils of P₁₂ (G) and P₁₅ (I) leaves in pSTM::STM-Venus. The signals in the corresponding leaf axils are extremely low in pSTM::STM-Venus pnh-2 (H and J). Note that STM-Venus signals in the axils of young leaves near the SAM in (G) and (H) cannot be directly compared, as the section in (H) does not capture the SAM/leaf boundaries due to morphological differences between pnh-2 and wild type. The inset in (H) shows the SAM/leaf boundaries with STM-Venus signals in young leaf axils in pnh-2. The most drastic difference in STM-Venus expression between wild type and pnh-2 is in axils of P₁₁ and older leaves. Bars, 50 μm in (C–J). Asterisks indicate the SAM.

Figure 4.. ARF5, BZR1, and PIF4 Expression Patterns at the Shoot Apex in Wild Type and AGO10 Expression in Various Mutants

Asterisks indicate the SAM, and the numbers indicate the developmental stages of leaves. Bars, 50 μm in (A–J). (A and B) Longitudinal (A) and cross (B) sections of shoot apices of pARF5::ARF5-GFP plants. ARF5-GFP signals are present on the adaxial side of young leaves (P₄) and at the leaf/SAM boundary (arrow) (A). ARF5-GFP signals are present in AMs (arrows in B). (C and D) Longitudinal (C) and cross (D) sections of shoot apices of pBZR1::BZR1-CFP plants. BZR1-CFP signals are found on the adaxial side of young leaves and at the leaf/SAM boundary (arrow) (C) and in AMs (arrow) (D). (E and F) Longitudinal (E) and cross (F) sections of shoot apices of pPIF4::PIF4-Citrine-HA plants. PIF4-Citrine signals are found on the adaxial side of young leaves and at the leaf/SAM boundary (arrow) (E) and in AMs (arrow) (F). (G–J) Cross sections of shoot apices from plants of various genotypes showing YFP-AGO10 signals in axils of P₁₅ leaves. The pAGO10::YFP-AGO10 transgene in wild type (G) was introduced into ARF5-GFP mp (H), bzr-1D (I), and phyB-9 (J) through crosses. Note that pARF5::ARF5-GFP mp is considered a weak arf5 allele due to lower levels of ARF5 expression than wild type. (K) qRT-PCR analysis to determine AGO10 transcript levels in Col (wild type) and various other genotypes as indicated. Error bars indicate the SD of three independent replicates. Transcript levels were compared with those in Col. *p < 0.01, (Student’s t test).

Figure 5.. ARF5, PIF4, and BZR1 Bind the AGO10 Promoter to Directly Regulate AGO10 Expression

(A) Schematic representation of the AGO10 promoter showing the positions of various motifs. ARE, auxin response elements; BRRE, brassinosteroid response element; G-box, a motif with high affinity for PIF4; E-box, a motif with low affinity for PIF4. ATG denotes the translation start site. Sixteen PCR fragments were designed for ChIP analysis. (B–D) ChIP-qPCR analysis to determine the binding of ARF5-GFP (B), PIF4-Citrine-HA (C), and BZR1-CFP (D) to the AGO10 promoter. Error bars indicate the SD of three independent replicates.

Figure 6.. ARF5, BZR1, and PIF4 Regulate AGO10 Transcription to Influence AM Forma-tion

(A) Schematic diagrams of the luciferase reporter (pAGO10::LUC), the internal control (p35S::REN), and the effectors used in transient expression assays in Arabidopsis protoplasts. The reporter gene LUC was driven by the AGO10 promoter (a 5120-bp region upstream of the ATG diagramed in Figure 5A). A minimal 35S promoter was included 3′ of the AGO10 promoter (black square). Renila luciferase was driven by the 35S promoter. The effector was ARF5-GFP, BZR1-GFP, or PIF4-GFP driven by the 35S promoter. The GFP vector alone served as a negative control. (B–D) Relative LUC expression in transcriptional activity assays in Arabidopsis protoplasts. The reporter was co-transformed with the empty vector p35S::GFP or one of the effectors p35S::ARF5/BZR1/PIF4-GFP constructs. Data are means ± SD for three independent experiments, each performed in triplicate. *p < 0.01. (E) A schematic diagram showing the deleted sites in the AGO10 promoter. (F) Relative (firefly/Renilla) luciferase activity in protoplasts co-transformed with p35S::ARF5-GFP or p35S::GFP, and the reporters driven by the native AGO10 promoter (pAGO10) or promoters with one (Δ1, Δ2) or both (Δ12) AREs deleted. Data are mean ± SD for three independent experiments, each performed in triplicate. One-way ANOVA and post hoc Tukey testing were used for statistical analysis. Different letters indicate significantly different values (p < 0.05). (G) Relative(firefly/Renilla) luciferase activity in protoplasts co-transformed with p35S::BZR1-GFP or p35S::GFP, and the reporters driven by the native AGO10 promoter (pAGO10) or the promoter with the BRRE deleted (Δ3). *p < 0.01, (Student’s t test). (H) Relative (firefly/Renilla) luciferase activity in protoplasts co-transformed with p35S::PIF4-GFP or p35S::GFP, and the reporters driven by the native AGO10 promoter (pAGO10) or the promoter with the G-box deleted (Δ4). *p < 0.01, (Student’s t test). (I) AM phenotypes of the indicated genotypes that were generated to interrogate the genetic relationships between auxin/light/BR and AGO10. Note that pAGO10D1 is a line with a deletion of the ARF5-binding site in the AGO10 promoter generated by Crispr-Cas9. pAGO10D1 showed AM defects in early leaves. AGO10 expression driven by the BZR1-binding-site-deleted promoter (pAGO10Δ3) and the PIF4-binding-site-deleted promoter (pAGO10Δ4) rescued the AM defects in bzr1–1D and PIF4 OX, respectively.

Figure 7.. ARF5, PIF4, and BZR1 Exhibit Different Temporal and Spatial Patterns of Expression during AM Development

(A–F, G, and H) Cross (A–F and H) and longitudinal (G) sections of shoot apices show the signals of fluorescent protein-tagged ARF5, PIF4, and BZR1. Numbers represent the developmental stages of leaves. AD, adaxial side; AB, abaxial side. Bars, 100 μm. (A and B) Cross sections of shoot apices of pARF5::ARF5-GFP plants. ARF5-GFP signals are on the adaxial side of young leaves (P₄), in the middle zone in older leaves before AM initiation (P₆, A) and in AMs in mature leaf axils (P₁₁, B). (C and D) Cross sections of shoot apices of pPIF4::PIF4-Citrine-HA plants. PIF4-Citrine signals are on the adaxial side of both young (P₄ and P₆ in C) and mature (P₁₁ in C and P₁₇ in D) leaves. (E and F) Cross sections of shoot apices of pBZR1::BZR1-CFP plants. BZR1-CFP accumulates in a high-to-low gradient from the adaxial side to the abaxial side in young leaves (P₈ in E); it is localized more in the nucleus on the adaxial side and more in the cytoplasm on the abaxial side. BZR1-CFP accumulates to higher levels in the epidermis (E). Temporally, BZR1-CFP is present in young leaves before AM initiation (P₅–P₇ in F) and absent in mature leaves immediately before or during AMs initiation (P₁₀ and P₁₁ in F). (G and H) Longitudinal (G) and cross (H) sections of apices of pBZR1::BZR1m-CFP plants (a phosphorylation site in BZR1 is mutated in BZR1m). BZR1m-CFP signals are on the adaxial side and in the epidermis of young leaves (P₈–P₁₀ in H). BZR1m-CFP signals persist in leaf stages when AMs would initiate (P_11, H). (I) Diagrams of gene expression patterns during AM development. P₃, P₆, and P₁₁ represent newly emerged leaf primordium, leaf primordium beginning to develop the vasculature, and an older leaf with AM initiation in its axil, respectively. STM is expressed in the P₃ leaf/SAM boundary; its expression decreases at the P₆/SAM boundary and begins to increase in the axil of P₁₁. REV, AGO10, and ARF5 are expressed on the adaxial side of P₃, in the middle zone as the leaves develop the vasculature and eventually in the vasculature and in the center of leaf axils where AMs initiate. AGO1 is expressed throughout P₃, predominantly in the middle zone, epidermis, and on the adaxial side of P₆ and in the vasculature, epidermis, and AMs in P₁₁. BZR1 is expressed on the adaxial side of P₃ and P₆, and in the vasculature in P₁₁. PIF4 expression persists on the adaxial side of all leaves and appears in the vasculature in P₁₁. After AM initiation, all the genes are expressed in axillary buds of P₁₅ with specific patterns not depicted in the diagrams.