The rhizobial type III effector ErnA confers the ability to form nodules in legumes

Abstract

Several Bradyrhizobium species nodulate the leguminous plant Aeschynomene indica in a type III secretion system-dependent manner, independently of Nod factors. To date, the underlying molecular determinants involved in this symbiotic process remain unknown. To identify the rhizobial effectors involved in nodulation, we mutated 23 out of the 27 effector genes predicted in Bradyrhizobium strain ORS3257. The mutation of nopAO increased nodulation and nitrogenase activity, whereas mutation of 5 other effector genes led to various symbiotic defects. The nopM1 and nopP1 mutants induced a reduced number of nodules, some of which displayed large necrotic zones. The nopT and nopAB mutants induced uninfected nodules, and a mutant in a yet-undescribed effector gene lost the capacity for nodule formation. This effector gene, widely conserved among bradyrhizobia, was named ernA for "effector required for nodulation-A." Remarkably, expressing ernA in a strain unable to nodulate A. indica conferred nodulation ability. Upon its delivery by Pseudomonas fluorescens into plant cells, ErnA was specifically targeted to the nucleus, and a fluorescence resonance energy transfer-fluorescence lifetime imaging microscopy approach supports the possibility that ErnA binds nucleic acids in the plant nuclei. Ectopic expression of ernA in A. indica roots activated organogenesis of root- and nodule-like structures. Collectively, this study unravels the symbiotic functions of rhizobial type III effectors playing distinct and complementary roles in suppression of host immune functions, infection, and nodule organogenesis, and suggests that ErnA triggers organ development in plants by a mechanism that remains to be elucidated.

Keywords: Bradyrhizobium; T3SS; legume; nodulation; symbiosis.

Fig. 1.

Identification of T3Es in Bradyrhizobium strain ORS3257 that play a symbiotic role in the interaction with Aeschynomene indica. (A) Genetic organization of putative effector genes identified in strain ORS3257. Deleted regions in the mutants are indicated by horizontal lines. Insertion mutants are indicated by black arrowheads carrying the Ω sign. In yellow, putative effector genes; in orange, nif and fix genes; in blue, nod genes; in red, genes encoding components of the T3SS apparatus; in green, the ttsI gene encoding the T3SS transcriptional regulator; black arrows, tts boxes. (B) Number of nodules formed and nitrogen fixation activity of A. indica plants at 21 d after inoculation with strain ORS3257 and its mutant derivatives. Nitrogen fixation activity was measured by the acetylene reduction assay; A.U., arbitrary unit. Box plots show results of 1 representative experiment out of at least 2 independent experiments per strain (18 plants each). The central rectangle spans the first quartile to the third quartile; the bold segment inside the rectangle shows the median; and the whiskers above and below the box show the locations of the maximum and minimum value, respectively. **P < 0.01, and ***P < 0.001, significant differences between WT ORS3257 and each mutant strain using a nonparametric Kruskal–Wallis test. (C) View of the root and the nodules induced by strain ORS3257 and its mutant derivatives. (Scale bars: Upper, 1.5 cm; Lower, 4 mm.) (D) Cytological analysis of the nodules induced by strain ORS3257 and its mutant derivatives observed by light (Upper) and confocal microscopy (Lower) after staining with SYTO 9 (green; live bacteria), calcofluor (blue; plant cell wall), and propidium iodide (red; infected plant nuclei and dead bacteria or bacteria with compromised membranes). (Scale bars, 500 µm.) In C and D, the white arrowheads indicate necrotic zones.

Fig. 2.

The ernA gene encodes a bona fide T3E, and its expression is under the control of TtsI. (A) Fold change expression of ernA (Brad3257_7701) and 2 T3SS genes (nopX and rhcJ) used as controls in ORS3257 and the ΩttsI mutant after induction with genistein. Bacteria cultivated in the absence of genistein but in presence of DMSO were used as reference. The expression levels were normalized using adhB (Brad3257_3749) transcripts. The level of expression was measured using qRT-PCR. Values represent mean ± SD (n = 3). (B) ErnA of ORS3257 is secreted in the supernatant. Secreted proteins from culture supernatants (sup.) or proteins from cell pellets (pel.) of the indicated strains were subjected to immunoblot analysis with the anti-His₆ (α-His) or anti-GFP (α-GFP) antibodies. (C) ErnA of ORS3257 is secreted via the T3SS. Secreted proteins from culture supernatants (sup.) or proteins from cell pellets (pel.) of the ORS3257∆T3SS mutant strains were subjected to immunoblot analysis with the anti-His₆ antibody (α-His). The artifacts observed in panel C are not the result of an image treatment.

Fig. 3.

Distribution of ernA genes among bradyrhizobia and symbiotic role in other strains after mutation or transfer. (A) Venn diagram representing the number of Bradyrhizobium strains with an available genome sequence and the proportion with a T3SS as well as a homolog of ernA. (B–D) Symbiotic properties on A. indica of 1) B. elkanii strain USDA61 and its mutant derivatives affected in the T3SS apparatus (USDA61ΩrhcJ) and the ernA homolog (USDA61ΩernA₆₁), and 2) Bradyrhizobium sp. strain DOA9 derivatives (WT and the T3SS mutant DOA9ΩrhcN) containing the empty vector pMG103 or ernA₃₂₅₇ cloned into pMG103. (B) Number of nodules formed on A. indica plants at 21 d after inoculation with the indicated strains. Box plots show results of 1 representative experiment out of at least 2 independent experiments per strain (18 plants each). The central rectangle spans the first quartile to the third quartile; the bold segment inside the rectangle shows the median; and the whiskers above and below the box show the locations of the maximum and minimum value, respectively. ***P < 0.001, significant differences between the WT strain and each mutant strain using a nonparametric Kruskal–Wallis test. (C) View of the roots and nodules elicited by the indicated strains. (Scale bars: Upper, 1.5 cm; Lower, 4 mm.) (D) Cytological analysis of the nodules elicited by the various strains tested. (Scale bars, 500 µm.)

Fig. 4.

ErnA is targeted to the plant cell nucleus and interacts with nucleic acids. (A) GFP fluorescence observed in N. benthamiana leaves transiently expressing ErnA-eGFP or ErnAΔNLS-eGFP. GFP was visualized by confocal microscopy 48 h after Agrobacterium infiltration of the leaves. From Left to Right: an overlay of GFP and chlorophyll fluorescence from transformed leaf cells, and the GFP and DAPI fluorescence spectrum of a representative nucleus from a transformed cell. Staining with DAPI was used to visualize nuclei. At least 10 nuclei were observed, and all of them showed the same distribution pattern of DAPI staining and GFP fluorescence. (Scale bars, 15 µm.) (B) Visualization of ErnA-GFP₁₁ in Arabidopsis cells after its secretion and injection by P. fluorescens. Three-week-old Arabidopsis Col-0 plants expressing GFP_1–10 were infiltrated with P. fluorescens expressing either ErnA-GFP₁₁ or ErnA-3HA. Reconstituted GFP was visualized by confocal microscopy 12 h postinfiltration. From Left to Right: an overlay of GFP, DAPI, and chlorophyll fluorescence spectrum, and the GFP and DAPI fluorescence spectrum, respectively. Staining with DAPI was used to visualize nuclei. (Scale bars, 15 µm.) (C) GFP lifetime distribution of nuclear ErnA-eGFP in presence or in absence of SYTOX Orange. FRET-FLIM measurements (SI Appendix, Table S3) were performed as described in Methods. Histograms show the distribution of nuclei (in percentage) according to classes of GFP lifetime value (in nanoseconds) in the absence (red bars) or in the presence (green bars) of the SYTOX Orange acceptor. The measured lifetimes of ErnA-eGFP were clearly shifted to lower values in the presence of SYTOX Orange compared to samples without the acceptor (indicated by partial overlap between green and orange arrows spanning relative GFP lifetime classes). Values were obtained from 8 different foliar discs collected 48 h postinfection and obtained from 2 independent experiments. Notably, upon transient expression of ernA in N. benthamiana leaves, no callus development or any other morphological changes were observed.

Fig. 5.

ErnA induces meristematic protuberances. (A–D) A. indica roots transformed with either the empty vector containing the DsRed marker (A) or p35S-ernA (B–D) and observed by a light (Upper) or a fluorescent (Lower) stereomicroscope equipped with a DsRed filter. Observations were made at 3 wk (B) or 7 wk (A, C, and D) after transformation in the absence of bradyrhizobia. White arrowheads in B indicate the formation of small bumps. (E–I) Lateral root-like structures induced by p35S-ernA observed 7 wk after transformation. View of lateral root-like structures (E). Cross-sections of transformed roots forming lateral root-like structures (F and G). (H and I) Confocal microscopy of lateral root-like structures using staining with SYTO 9 (green; xylem vessels), calcofluor (blue; plant cell wall), and propidium iodide (red; plant nuclei). In H, abbreviations: ct, cortical cells; ed, endoderm; ep, epiderm; pi, pericycle; pl, phloem; xl, xylem vessels. In I, white arrowheads indicate new meristems. (J–N) Root nodule-like primordia structures induced by p35S-ernA. View of the root nodule-like primordia (J). Longitudinal sections of root nodule-like primordia (K–N). In K, black arrowheads indicate necrotic zones. (O and P) Cross-sections of tumor-like structures observed either by light (O) or by confocal microscopy (P) after staining as in H and I. (Scale bars: A, C, and D, 1.5 cm; B, E, and J, 2 mm; F, G, I, and K–P, 500 µm; H, 50 µm.)

Fig. 6.

Proposed model for the NF-independent, T3SS-dependent symbiotic process between Bradyrhizobium ORS3257 and A. indica. The symbiosis between ORS3257 and A. indica does not involve NFs but depends on a mixture of T3Es delivered into the host cell where they act in concert for nodulation. NopP1 and NopM1 suppress plant defense responses resulting from the activation of PTI and/or ETI. Both NopAB and NopT effectors promote bacterial infection of the nodule, directly or indirectly. The ErnA effector triggers organ development either by activating the common symbiosis signaling pathway (CSSP) or by a yet-unknown mechanism. Moreover, we cannot exclude that ErnA is also involved, directly or indirectly, in the infection process.