The strigolactone biosynthesis gene DWARF27 is co-opted in rhizobium symbiosis

Abstract

Background: Strigolactones are a class of plant hormones whose biosynthesis is activated in response to phosphate starvation. This involves several enzymes, including the carotenoid cleavage dioxygenases 7 (CCD7) and CCD8 and the carotenoid isomerase DWARF27 (D27). D27 expression is known to be responsive to phosphate starvation. In Medicago truncatula and rice (Oryza sativa) this transcriptional response requires the GRAS-type proteins NSP1 and NSP2; both proteins are essential for rhizobium induced root nodule formation in legumes. In line with this, we questioned whether MtNSP1-MtNSP2 dependent MtD27 regulation is co-opted in rhizobium symbiosis.

Results: We provide evidence that MtD27 is involved in strigolactone biosynthesis in M. truncatula roots upon phosphate stress. Spatiotemporal expression studies revealed that this gene is also highly expressed in nodule primordia and subsequently becomes restricted to the meristem and distal infection zone of a mature nodules. A similar expression pattern was found for MtCCD7 and MtCCD8. Rhizobium lipo-chitooligosaccharide (LCO) application experiments revealed that of these genes MtD27 is most responsive in an MtNSP1 and MtNSP2 dependent manner. Symbiotic expression of MtD27 requires components of the symbiosis signaling pathway; including MtDMI1, MtDMI2, MtDMI3/MtCCaMK and in part MtERN1. This in contrast to MtD27 expression upon phosphate starvation, which only requires MtNSP1 and MtNSP2.

Conclusion: Our data show that the phosphate-starvation responsive strigolactone biosynthesis gene MtD27 is also rapidly induced by rhizobium LCO signals in an MtNSP1 and MtNSP2-dependent manner. Additionally, we show that MtD27 is co-expressed with MtCCD7 and MtCCD8 in nodule primordia and in the infection zone of mature nodules.

Fig. 1

Bayesian phylogeny of D27 and D27-like proteins. Phylogeny was reconstructed based on an alignment of D27 and D27-like proteins from Arabidopsis (At), soybean (Glycine max) (Glyma), Lotus japonicus (Lj), M. truncatula (Medtr), rice (Os), poplar (Populus trichocarpa) (Potri) and grapevine (Vitis vinifera) (VIT). Branch support is indicated by posterior probabilities. Terminals are labeled by their gene name or genbank identifier. Proteins identified from M. truncatula are highlighted in bold. The D27 orthology group containing rice OsD27 (gi|2549466546) and Arabidopsis AtD27 (At1G03055) is highlighted in blue. Mid-point rooting was applied for better tree visualization

Fig. 2

Effect of MtD27 knock-down on strigolactone biosynthesis. a Relative transcript abundance as determined by qRT-PCR of MtD27 and Medtr7g095920 in M. truncatula transgenic roots expressing an empty vector control construct (EVi) or MtD27 RNAi construct (D27i). Relative transcript abundance was normalized against MtD27 transcript abundance in roots transformed with the empty vector control (EVi). b Relative quantity of the strigolactone didehydro-orobanchol (DDH) (peak area/g FW) in root exudates collected from M. truncatula transgenic roots expressing an empty vector control construct (EVi) or MtD27 RNAi construct (D27i). c Relative quantity of the strigolactone didehydro-orobanchol (DDH) (peak area/g FW) in root extracts collected from M. truncatula transgenic roots expressing an empty vector control construct (EVi) or MtD27 RNAi construct (D27i). Data shown represent means of 4-5 biological replicates ± SEM. Different letters above bars indicate statistical difference (p < 0.05, students’ t-test)

Fig. 3

Relative transcript abundance of MtD27 and Medtr7g095920 after application of rhizobium LCOs. Relative transcript abundance as determined by qRT-PCR of MtD27 and Medtr7g095920 in M. truncatula root susceptible zones 3 h after mock (-LCO) or S. meliloti LCO (10^-9 M) (+LCO) treatment. Data shown represent means of 2 biological replicates that each were analyzed in 3-fold (technical replicates) ± SEM. For each gene, transcript abundance was normalized against that of the mock-treated wild type. Different letters above bars indicate statistical difference (p < 0.05, students’ t-test)

Fig. 4

Symbiotic MtD27 expression is under direct control of S. meliloti LCO signaling. a Relative transcript abundance as determined by qRT-PCR of MtD27 and MtENOD11 in M. truncatula root susceptible zones after 0, 1, 2 or 3 h of LCO treatment (10^-9 M). b Relative transcript abundance as determined by qRT-PCR of MtD27 at 3 h after mock (-LCO) or rhizobium LCO (10^-9 M) (+LCO) treatment in wild type, Mtdmi1, Mtdmi2 and Mtdmi3. c Relative transcript abundance as determined by qRT-PCR of MtD27 at 3 h after mock (-LCO) or rhizobium LCO (10^-9 M) (+LCO) treatment in wild type, Mtnsp1, Mtnsp2 and Mtnsp1 Mtnsp2. d Relative transcript abundance as determined by qRT-PCR of MtENOD11 at 3 h after mock (-LCO) or rhizobium LCO (10^-9 M) (+LCO) treatment in wild type and Mtern1. e Relative transcript abundance as determined by qRT-PCR of MtD27 at 3 h after mock (-LCO) or rhizobium LCO (10^-9 M) (+LCO) treatment in wild type and Mtern1. f Relative transcript abundance as determined by qRT-PCR of MtD27 and MtENOD11 after mock (-LCO) or rhizobium LCO (10^-9 M) (+LCO) treatment, in presence or absence of 50 μM cycloheximide (CHX). Data shown represent means of 3 biological replicates that each were analyzed in 3-fold (technical replicates) ± SEM. For each gene, transcript abundance was normalized against that of the mock-treated wild type. Different letters above bars indicate statistical difference (p < 0.05, students’ t-test)

Fig. 5

Spatial expression pattern of MtD27 in M. truncatula roots and nodules. The MtD27 spatial expression pattern was analyzed in M. truncatula transgenic roots expressing an MtD27 promoter-reporter GUS construct. a Cross-section through a non-inoculated root. Arrowheads indicate casparian strips, which mark the endodermal cell layer. b Mock-treated control root. c Root treated with S. meliloti LCOs (10^-9 M) for 3 h. d Root at four days post inoculation (dpi) with S. meliloti strain 2011. e Root at 7 dpi. f Longitudinal section through a root at 2 dpi. g Longitudinal section through a nodule primordium (7 dpi). The infection thread is indicated with an arrowhead. h Longitudinal section through an eighteen-day-old nodule. Scale bars are equal to 25 μm (a, f and g), 0.5 mm (b-e) and 50 μm (h). Sections were counterstained with Ruthenium Red

Fig. 6

Expression of MtD27, MtCCD7 and MtCCD8 upon treatment with rhizobium LCOs. a Relative transcript abundance as determined by qRT-PCR of MtD27, MtCCD7 and MtCCD8 in M. truncatula root susceptible zones 3 h after mock or rhizobium LCO (10^-9 M) treatment. RNA was isolated from plants grown in Fåhraeus slides. b Relative transcript abundance as determined by qRT-PCR of MtD27, MtCCD7 and MtCCD8 in M. truncatula root susceptible zones 3 or 6 h after mock or rhizobium LCO (10^-9 M) treatment. RNA was isolated from plants grown on agar-solidified Fåhraeus medium supplemented with 1 μM AVG. Data shown represent means of 2-3 biological replicates that each were analyzed in 3-fold (technical replicates) ± SEM. For each gene, transcript abundance was normalized against that of the mock-treated wild type. Different letters above bars indicate statistical difference (p < 0.05, students’ t-test)

Fig. 7

Spatial expression pattern of MtCCD7 and MtCCD8 in M. truncatula nodule primordia and mature nodules. Expression patterns were analyzed in M. truncatula transgenic roots expressing promoter-reporter GUS constructs. a Longitudinal section through a root expressing the pMtCCD7::GUS construct 2 days post inoculation (dpi) with S. meliloti strain 2011. b Longitudinal section through a root expressing the pMtCCD8::GUS construct at 2 dpi. Arrowhead points at an infection thread growing inside the root hair cell. c Longitudinal section through a mature nodule expressing the pMtCCD7::GUS construct. d Longitudinal section through a mature nodule expressing the pMtCCD8::GUS construct. Scale bars are equal to 50 μm. Sections were counterstained with Ruthenium Red

Fig. 8

Spatial expression pattern of MtD27 upon phosphate starvation. a Transgenic M. truncatula root expressing pMtD27::GUS grown in full nutrient condition (200 μM PO₄ ^3-). b Longitudinal section of the root shown in (a). c Expression of pMtD27::GUS in M. truncatula transgenic roots after 5 days of phosphate starvation (0 μM PO₄ ^3-). d Longitudinal section of the root shown in (c). e Relative transcript abundance as determined by qRT-PCR of MtD27 in wild type (WT) and the Mtdmi3 mutant and Mtnsp1 Mtnsp2 double mutant after 2 days of phosphate starvation (0 μM PO₄ ^3-). f Relative transcript abundance as determined by qRT-PCR of MtD27 in WT and the Mtern1 mutant and Mtnsp1 Mtnsp2 double mutant after 2 days of phosphate starvation. Scale bars are equal to 250 μm in (a) and (c) and 50 μm in (b) and (d). Sections were counterstained with Ruthenium Red. Data in (e,f) represent means of 3 biological replicates that each were analyzed in 3-fold (technical replicates) ± SEM. For each gene, transcript abundance was normalized against that of the mock-treated wild type. Different letters above bars indicate statistical difference (p < 0.05, students’ t-test)

Fig. 9

Schematic model depicting the induction of MtD27 expression by phosphate starvation and rhizobium LCO-induced signaling. Rhizobium LCOs activate expression of MtD27 through the common symbiosis signaling module consisting of MtDMI1, MtDMI2 and MtDMI3. Downstream of this module MtERN1 is required for full induction of MtD27. MtERN1 function is partly redundant, suggesting that MtERN1 might function in conjunction with or redundant to another unknown transcriptional regulator, indicated as X. This unknown factor might be MtERN2, a transcriptional regulator closely related to MtERN1 [48]. The pathway leading to activation of MtD27 expression by phosphate starvation remains unknown. The GRAS proteins MtNSP1 and MtNSP2 are required for expression of MtD27 during both phosphate starvation and following rhizobium LCO application. We propose that these proteins function in parallel to both signaling pathways, as previously already suggested for the LCO signaling cascade [7, 15]