ZAXINONE SYNTHASE 2 regulates growth and arbuscular mycorrhizal symbiosis in rice

Abstract

Carotenoid cleavage, catalyzed by CAROTENOID CLEAVAGE DIOXYGENASEs (CCDs), provides signaling molecules and precursors of plant hormones. Recently, we showed that zaxinone, a apocarotenoid metabolite formed by the CCD ZAXINONE SYNTHASE (ZAS), is a growth regulator required for normal rice (Oryza sativa) growth and development. The rice genome encodes three OsZAS homologs, called here OsZAS1b, OsZAS1c, and OsZAS2, with unknown functions. Here, we investigated the enzymatic activity, expression pattern, and subcellular localization of OsZAS2 and generated and characterized loss-of-function CRISPR/Cas9 (clustered regularly interspaced short palindromic repeats and associated protein 9)-Oszas2 mutants. We show that OsZAS2 formed zaxinone in vitro. OsZAS2 was predominantly localized in plastids and mainly expressed under phosphate starvation. Moreover, OsZAS2 expression increased during mycorrhization, specifically in arbuscule-containing cells. Oszas2 mutants contained lower zaxinone content in roots and exhibited reduced root and shoot biomass, fewer tillers, and higher strigolactone (SL) levels. Exogenous zaxinone application repressed SL biosynthesis and partially rescued the growth retardation of the Oszas2 mutant. Consistent with the OsZAS2 expression pattern, Oszas2 mutants displayed a lower frequency of arbuscular mycorrhizal colonization. In conclusion, OsZAS2 is a zaxinone-forming enzyme that, similar to the previously reported OsZAS, determines rice growth, architecture, and SL content, and is required for optimal mycorrhization.

Figure 1

Phylogenetic analysis of ZAS enzymes and analysis of OsZAS2 enzymatic activity. A, Phylogenetic tree analysis of ZAS orthologs from selected monocot and dicot plants, showing bootstrap values on nodes of each cluster. Dashed rectangles represent rice ZAS members. The scale bar represents the number of amino acid change per site. B, HPLC chromatogram of in vitro incubation of OsZAS2 with apo-10′-zeaxanthinal (I) yielded zaxinone (II) and a presumed C₉-dialdehyde. The maximum absorbance (mAU) peak for substrate and product is shown at 347 and 450 nm (mAU), respectively. The representation of zaxinone production in the Figure B adapted from Wang et al. (2019), which is permitted to adaptation under a Creative Commons Attribution 4.0 International License. C, Verification of OsZAS2 in vitro product, based on retention time, (D) accurate mass and MS/MS pattern and in comparison to zaxinone standard.

Figure 2

OsZAS2 subcellular localization and expression pattern. A, Subcellular localization of OsZAS2 transiently expressed in N. benthamiana epidermis leave tissue. B, GUS staining of roots of two independent pZAS2:GUS reporter lines (pZAS2::GUS-L11, pZAS2::GUS-L18) under normal (+Pi) and low Pi conditions. Dash rectangle emphasizes the root tip. C, Cross-section of pZAS2:GUS11 line primary root was examined with two different resolutions under microscope: in the first resolution (parts I and II, bars correspond to 50 μm) all types of tissues were observed while the second resolution of the same samples (parts III and IV, bars correspond to 25 μm) showed a close view of epidermal and cortex cells. Exo, exoderms; epi, epidermis; c, cortex; cc, central cylinder. D, Normalized expression value of OsZAS2 under normal (+Pi) and low Pi conditions in root and shoot tissue of 21-day-old rice plants. Values in (D) are means ± sd (n = 4). Student’s t test used for the statistical analysis (****P ≤ 0.0001).

Figure 3

Characterization of CRISPR/Cas9-mediated Oszas2 mutant lines at the seedling stages. A, Schematic representation of three individual mutations of OsZAS2 gene generated by CRISPR–Cas9. B, The seedling phenotype of WT (DJ) and three independent Oszas2 mutants. The scale bar in the pictures represents 7.5 cm. C, Quantification of zaxinone content in WT and Oszas2 mutants roots. D–H, Root biomass (D), shoot biomass (E), root length (F), shoot length (G), and tiller number (H) of the WT and Oszas2 mutants are shown in (B). Boxes in boxplots represent the median, first and third quartile. The minimum and maximum values are shown with the length of the whiskers. Dots represent the biological replicates. Values in (C–H) are means ± sd (n ≥ 4). Student’s t test used for the statistical analysis (***P ≤ 0.001).

Figure 4

Characterization of Oszas2 mutant lines at the maturing stage. A, The picture of the 3-month-old WT and Oszas2 mutants grown in the greenhouse. The scale bar in the pictures represents 7.5 cm. B–E, Grain weight per plant (B), shoot biomass (C), plant height (D), and tiller number (E) of the WT and Oszas2 mutants represented in (A). Values in (B–E) are means ± sd (n ≥ 7). Student’s t test was applied for the statistical analysis (***P ≤ 0.001).

Figure 5

OsZAS2 is required for AM establishment. A, GUS staining of roots of pZAS2:GUS-L18 reporter line inoculated (I, II, and III) for 35 days with F. mosseae and noninoculated (IV, V, and VI). B, Localization of OsZAS2 mRNA in sections from differentiated regions of inoculated roots by cold in situ hybridization. Sections of mycorrhizal roots treated with OsZAS2 antisense probe are shown in parts I and II; arrows highlight the strong chromogenic signal, which mirrors the presence of the OsZAS2 transcript in arbuscule-containing cells. Sections of mycorrhizal roots treated with the OsZAS2 sense probe are shown in parts III and IV; arrows in part IV indicate arbusculated cells that are not labeled. C, Relative expression of fungal genes; RiEF and RiPeip1 in WT and Oszas2 mutants at 50 dpi. D, Relative expression of OsPT11 at 14 and 40 dpi in WT and Oszas2 mutants. E, Frequency of mycorrhizal colonization (F%), the intensity of colonization (M%), and a total number of arbuscules (A%) in WT and Oszas2 mutants at 50 dpi. cc, central cylinder; c, noncolonized cortical cells; e, epidermal cells; vt, vascular tissue; ar, arbuscule containing cells; rc, root cap. Bars (A and B) correspond to 50 µm. Values in (C–E) are means ± sd (n ≥ 3). Student’s t test was applied for the statistical analysis (**P ≤ 0.01; ***P ≤ 0.001).

Figure 6

SL biosynthesis increased in Oszas2 mutants. A, Relative quantification of 4DO and 4-oxo-MeCLA in root tissue of Oszas2 mutants. B, Normalized expression value of SL biosynthetic genes in Oszas2 mutants. C, Relative quantification of canonical (4DO, Orobanchol) and noncanonical (4-oxo-MeCLA) SL in root exudate of Oszas2 mutants. D, Striga seed germination assay conducted with root exudate and tissue of Oszas2 mutants. For both root exudate and tissue bioassay, 1 μM of GR24 was used as control, which showed about 65% and 67% of Striga seed germination, respectively. Boxes in boxplots in (A) and (C) represent the median, first and third quartile. The minimum and maximum values are showed with the length of the whiskers. Dots represent the biological replicates. Values in (A–D) are means ± sd (n ≥ 4) and student’s t test was applied for the statistical analysis (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001).

Figure 7

Zaxinone treatment reduced SL biosynthesis in Oszas2 mutant. A, SL biosynthetic genes; OsD27, OsCCD7, OsCCD8, and OsCO expression in WT and Oszas2 mutant upon zaxinone (5 μM) treatment. B, Relative content of 4-oxo-MeCLA after zaxinone (5 μM) treatment in root tissue and exudate of WT and Oszas2 mutant. C, Relative content of Orobanchol after zaxinone (5 μM) treatment in root exudate of WT and Oszas2 mutant. D, Striga seed germination assay with exudate of WT and Oszas2 mutant upon zaxinone (5 μM) treatment. About 1 μM of GR24 was used as control, which showed about 63% of Striga seed germination. Values in (A–D) are means ± sd (n ≥ 4). Student’s t test was applied for the statistical analysis (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001).

Figure 8

Exogenous zaxinone application rescued the growth defects of Oszas2 mutant. A, The images of the WT (DJ) and Oszas2-d mutant grown for 2 weeks in soil supplemented with 10 μM of zaxinone and tap water (0.01% [v/v] acetone) as mock. The white bar represents 10 cm of scale. The white arrows represent main tillers. B–F, Tiller number (B), root biomass (C), shoot biomass (D), root length (E), and shoot length (F) of the WT and Oszas2 mutants are shown in (A). Boxes in boxplots represent the median, first and third quartile. The minimum and maximum values are showed with the length of the whiskers. Dots represent the biological replicates. Values in (B–F) are means ± sd (n ≥ 7). Student’s t test used for the statistical analysis (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ns, not significant).