Disruption of the rice 4-DEOXYOROBANCHOL HYDROXYLASE unravels specific functions of canonical strigolactones

Abstract

Strigolactones (SLs) regulate many developmental processes, including shoot-branching/tillering, and mediate rhizospheric interactions. SLs originate from carlactone (CL) and are structurally diverse, divided into a canonical and a noncanonical subfamily. Rice contains two canonical SLs, 4-deoxyorobanchol (4DO) and orobanchol (Oro), which are common in different plant species. The cytochrome P450 OsMAX1-900 forms 4DO from CL through repeated oxygenation and ring closure, while the homologous enzyme OsMAX1-1400 hydroxylates 4DO into Oro. To better understand the biological function of 4DO and Oro, we generated CRISPR/Cas9 mutants disrupted in OsMAX1-1400 or in both OsMAX1-900 and OsMAX1-1400. The loss of OsMAX1-1400 activity led to a complete lack of Oro and an accumulation of its precursor 4DO. Moreover, Os1400 mutants showed shorter plant height, panicle and panicle base length, but no tillering phenotype. Hormone quantification and transcriptome analysis of Os1400 mutants revealed elevated auxin levels and changes in the expression of auxin-related, as well as of SL biosynthetic genes. Interestingly, the Os900/1400 double mutant lacking both Oro and 4DO did not show the observed Os1400 architectural phenotypes, indicating their being a result of 4DO accumulation. Treatment of wild-type plants with 4DO confirmed this assumption. A comparison of the Striga seed germinating activity and the mycorrhization of Os900, Os900/1400, and Os1400 loss-of-function mutants demonstrated that the germination activity positively correlates with 4DO content while disrupting OsMAX1-1400 has a negative impact on mycorrhizal symbiosis. Taken together, our paper deciphers the biological function of canonical SLs in rice and reveals their particular contributions to establishing architecture and rhizospheric communications.

Keywords: Striga; arbuscular mycorrhizal fungi; cytochrome P450; plant architecture; strigolactones.

Fig. 1.

SL quantification of the CRISPR/Cas9 mediated OsMAX1 mutants. (A and B) Analysis of SLs in root exudates and root tissues of WT, Os900-KO line, Os900/1400-KO line, Os1400-KO lines, and d17 mutant grown under constant low-Pi conditions. 4-oxo-MeCLA is present in root exudates of WT and Os900-KO (at a quite low, less than 4% of the WT level; SI Appendix, Fig. S4C). The data are presented as means ± SD of 4 biological replicates for (A) and (B). Significant values determined by one-way ANOVA are shown with different letters (P < 0.05) when compared to WT, and asterisks indicate statistically significant differences as compared to control by the two tailed unpaired Student t test (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001). (C) Scheme of the rice SL biosynthesis. The pathway starts with a reversible isomerization of all-trans- into 9-cis-β-carotene, catalyzed by DWARF27 (D27). Next, 9-cis-β-carotene is transformed into carlactone (CL) through cleavage and rearrangement reactions mediated by the CAROTENOID CLEAVAGE DIOXYGENASE 7 and 8 (CCD7 and CCD8). OsMAX1 enzymes can oxygenate CL into carlactonoic acid (CLA), which is further converted into the canonical SLs 4-deoxyorobanchol (4DO) and then orobanchol (Oro) by sequential actions of Os900 and Os1400. In addition, CL is transformed by unknown enzymes into metabolites with a 14 Da (CL+14, putative Oxo-CL) and a 30 Da (putative 4-Oxo-19-hydroxy-CL) higher molecular mass. Os900 and a postulated methyltransferase convert CL+30 into the noncanonical SL methyl 4-oxo-carlactonoate (4-oxo-MeCLA, previously described as a methoxy-5-deoxystrigol isomer), which represents the major route for its formation. The detection of 4-oxo-MeCLA in exudates of Os900-KO mutant at a low level [not visible at the scale shown in (A), SI Appendix, Fig. S4C] indicates the presence of an additional, Os900-independent and minor route for its biosynthesis (21). Abbreviations: D27, Dwarf27; CCD, Carotenoid Cleavage Dioxygenase; MAX1, More Axillary Growth 1; CYP, Cytochrome P450; 4DO, 4-Deoxyorobanchol, 4-oxo-MeCLA, Methyl 4-oxo-carlactonoate.

Fig. 2.

Phenotypic characterization of OsMAX1 mutants. (A and B) Shoot phenotypes of WT, Os900-KO line, Os900/1400-KO line, Os1400-KO lines, and d17 mutant grown in soil (Scale bar, 10 cm). (C and D) Panicle phenotypes of WT, Os900-KO line, Os900/1400-KO line, Os1400-KO lines, and d17 mutant (Scale bar, 5 cm). The data are presented as means ± SD of six biological replicates. Significant values determined by one-way ANOVA are shown with different letter (P < 0.05) when compared to WT, and asterisks indicate statistically significant differences as compared to control by the two tailed unpaired Student t test (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).

Fig. 3.

Exogenous 4DO application as well as transcriptome and hormone analysis of Os1400-KO lines. (A) Shoot and root phenotypes of WT, and Os1400-KO lines grown in hydroponic culture with or without (Mock) 900 nM 4DO (Scale bar, 5 cm). (B) Heat map analysis of DEGs involved Auxin pathways. AUXIN1/LIKE-AUX1 (AUX1/LAX) are major auxin influx carriers; AUXIN/indole-3-acetic acid (AUX/IAA) are transcriptional repressors; the PIN-FORMED (PIN) proteins are secondary transporters in the efflux of auxin; GRETCHEN HAGEN 3 (GH3) gene family encodes auxin-amido synthetases. The expression pattern was shown in log₂FoldChange (Log₂FC). Statistically significant differences are indicated by adjusted P-value (*< 0.05). (C) Analysis of IAA (auxin) in root and shoot of WT, and Os1400-KO lines grown under constant low-Pi and +Pi conditions. Abbreviation: 4DO, 4-deoxyorobanchol; WT, wild-type; ns, nonsignificant; IAA, indole-3-acetic acid. The data are presented as means ± SD of 5 (A), 3 (B), and 4 (C) biological replicates. Significant values determined by one-way ANOVA are shown with different letter (P < 0.05) when compared to WT, and asterisks indicate statistically significant differences as compared to control by the two tailed unpaired Student t test (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).

Fig. 4.

Assessment of rhizospheric interactions. Effect of Os1400-KO lines on (A) the germination of root parasitic weed Striga and (B and C) the arbuscule formation. The R. irregularis colonization was quantified by measuring the expression of an AM marker gene (OsPT11) (B). Arbuscule formation at 10 dpi and 40 dpi. Arrows indicate arbuscule containing cells (Scale bars, 50 µm). (C) The data are presented as means ± SD of 4 (A), and n ≥ 3 (B) biological replicates. Significant values determined by one-way ANOVA are shown with different letter (P < 0.05) when compared to WT, and asterisks indicate statistically significant differences as compared to control by the two tailed unpaired Student t test (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).