Silencing the flavonoid pathway in Medicago truncatula inhibits root nodule formation and prevents auxin transport regulation by rhizobia

Abstract

Legumes form symbioses with rhizobia, which initiate the development of a new plant organ, the nodule. Flavonoids have long been hypothesized to regulate nodule development through their action as auxin transport inhibitors, but genetic proof has been missing. To test this hypothesis, we used RNA interference to silence chalcone synthase (CHS), the enzyme that catalyzes the first committed step of the flavonoid pathway, in Medicago truncatula. Agrobacterium rhizogenes transformation was used to create hairy roots that showed strongly reduced CHS transcript levels and reduced levels of flavonoids in silenced roots. Flavonoid-deficient roots were unable to initiate nodules, even though normal root hair curling was observed. Nodule formation and flavonoid accumulation could be rescued by supplementation of plants with the precursor flavonoids naringenin and liquiritigenin. The flavonoid-deficient roots showed increased auxin transport compared with control roots. Inoculation with rhizobia reduced auxin transport in control roots after 24 h, similar to the action of the auxin transport inhibitor N-(1-naphthyl)phthalamic acid (NPA). Rhizobia were unable to reduce auxin transport in flavonoid-deficient roots, even though NPA inhibited auxin transport. Our results present genetic evidence that root flavonoids are necessary for nodule initiation in M. truncatula and suggest that they act as auxin transport regulators.

Figure 1.

RT-PCR for CHS in Control and Silenced Hairy Roots. (A) RT-PCR of individual roots showing examples of CHS expression in fluorescent (FL) empty vector control hairy root, fluorescent CHS-silenced root, and nonfluorescent (non-FL) CHS-silenced root. Actin was used as a loading control. Numbers of roots in each category showing these RT-PCR results are shown in Table 1. (B) RT-PCR for CHS in roots and leaves of empty vector control (with fluorescent roots) and CHS-silenced (with nonfluorescent roots) hairy root plantlets. Approximately 10 roots or leaves were pooled for these experiments. Actin was used as a loading control.

Figure 2.

HPLC Chromatograms and Absorbance Spectra of Flavonoid Extracts from Roots. All chromatograms show absorbance at 300 nm. (A) Chromatograms of root extracts from empty vector control (black line; top) and CHS-silenced hairy roots (gray line; bottom) grown on Fåhraeus (F) medium. Daidzein (Da), formononetin (Fo), and medicarpin (Me) could be detected. Naringenin (Na) and liquiritigenin (Li) were below the detection limit. The gray and black lines were offset to make it easier to view both. (B) Chromatograms of extracts from empty vector control (black line; top) and CHS-silenced roots (gray line; bottom) grown on medium complemented with 100 nM naringenin and 100 nM liquiritigenin in F medium. The gray and black lines were offset to make it easier to view both. (C) to (E) Retention times of daidzein, formononetin, and medicarpin standards. (F) to (H) Absorbance spectra of daidzein, formononetin, and medicarpin standards (gray lines) compared with the absorbance spectra of the respectively labeled peaks in ([A]; black lines).

Figure 3.

Fluorescence Microscopy of Control and CHS-Silenced Plants. (A) Autofluorescence in a hairy root tip of a control hairy root. (B) Autofluorescence in a hairy root tip of a CHS-silenced hairy root. (C) Intracellular autofluorescence in a fresh vibratome section of a control hairy root. (D) Autofluorescence in a fresh vibratome section of a CHS-silenced hairy root is limited to cell wall fluorescence. (E) Enhanced and red-shifted intracellular and root hair fluorescence after addition of DPBA in a control hairy root section. (F) No intracellular fluorescence is evident in a CHS-silenced hairy root section treated with DPBA. (A) and (B), (C) and (E), and (D) and (F) were taken with the same exposure settings, respectively. Bars = 1 mm ([A] and [B]) and 100 μm ([C] to [F]).

Figure 4.

Effects of Flavonoid Deficiency on Nodulation. (A) Root hair curling on a control hairy root 48 h after inoculation with GFP-labeled S. meliloti. (B) Root hair curling on a CHS-silenced hairy root 48 h after inoculation with GFP-labeled S. meliloti. (C) Cortical cell divisions in control hairy roots 72 h after inoculation, stained with toluidine blue. (D) Autofluorescence in a CHS-silenced root grown for 1 week on flavonoid-supplemented medium. This photo was taken at the same time with the same exposure settings as Figures 3A and 3B. (E) Nodule developing on a CHS-silenced root grown on flavonoid-supplemented medium 10 d after inoculation with GFP-labeled S. meliloti, which can be seen on the nodule surface (arrowheads). (F) Autofluorescence of a fresh vibratome section of a 10-d-old nodule of a control hairy root grown on flavonoid-supplemented medium. Infection threads of GFP-labeled bacteria can be seen at the arrowhead. (G) Autofluorescence of a fresh vibratome section of a 10-d-old nodule of a CHS-silenced hairy root grown on flavonoid-supplemented medium. Infection threads of GFP-labeled bacteria can be seen at the arrowhead. Bars = 50 μm ([A] and [B]), 100 μm (C), 1 mm ([D] and [E]), and 500 μm ([F] and [G]).

Figure 5.

Acropetal Auxin Transport Measurements in Hairy Roots. (A) Effect of 24 h treatment with NPA on auxin transport in control seedling roots. The asterisk indicates a significant difference at the 0.01 level in the NPA treatment compared with control (Student's t test; n = 10). (B) Comparison of auxin transport in untreated control and CHS-silenced hairy roots. The asterisk indicates a significant difference at the 0.05 level in the CHS-silenced roots compared with control roots (Student's t test; n = 19 to 20). (C) Effect of S. meliloti and NPA on auxin transport in control and CHS-silenced hairy roots. Any two pairwise comparisons of treatments labeled with different letters are significantly different from each other at the 0.05 level (one way analysis of variance; n = 10 to 24). All error bars indicate standard errors of the mean. For the experimental setup, see Supplemental Figure 4 online.