The GRAS protein RAM1 interacts with WRI transcription factors to regulate plant genes required for arbuscule development and function

Abstract

During arbuscular mycorrhiza (AM) symbiosis AM fungi form tree-shaped structures called arbuscules in root cortex cells of host plants. Arbuscules and their host cells are central for reciprocal nutrient exchange between the symbionts. REQUIRED FOR ARBUSCULAR MYCORRHIZATION1 (RAM1) encodes a GRAS protein crucial for transcriptionally regulating plant genes needed for arbuscule development and nutrient exchange. Similar to other GRAS proteins, RAM1 likely does not bind to DNA and how RAM1 activates its target promoters remained elusive. Here, we demonstrate that RAM1 interacts with five AM-induced APETALA 2 (AP2) transcription factors of the WRINKLED1-like family called CTTC MOTIF-BINDING TRANSCRIPTION FACTOR1 (CBX1), WRI3, WRI5a, WRI5b, and WRI5c via a C-terminal domain containing the M2/M2a motif. This motif is conserved and enriched in WRI proteins encoded by genomes of AM-competent plants. RAM1 together with any of these WRI proteins activates the promoters of genes required for symbiotic nutrient exchange, namely RAM2, STUNTED ARBUSCULES (STR), and PHOSPHATE TRANSPORTER 4 (PT4), in Nicotiana benthamiana leaves. This activation as well as target promoter induction in Lotus japonicus hairy roots depends on MYCS (MYCORRHIZA SEQUENCE)-elements and AW-boxes, previously identified as WRI-binding sites. The WRI genes are activated in two waves: Transcription of RAM1, CBX1, and WRI3 is coregulated by calcium- and calmodulin-dependent protein kinase-activated CYCLOPS, through the AMCYC-RE in their promoter, and DELLA, while WRI5a, b, and c promoters contain MYCS-elements and AW-boxes and can be activated by RAM1 heterocomplexes with CBX1 or WRI3. We propose that RAM1 provides an activation domain to DNA-binding WRI proteins to activate genes with central roles in AM development and function.

Keywords: GRAS transcription factor; arbuscular mycorrhiza; root; symbiosis; transcriptional regulation.

Fig. 1.

WRI-target sites in the RAM2 promoter are required for RAM2 activation by RAM1 and for arbuscule development. (A) Transactivation assay in N. benthamiana leaves. Genomic sequence encoding proteins indicated at the y-axis were controlled by the LjUbiquitin10 promoter. The schematic representation on Top illustrates position and sequence of AW-boxes (yellow marks) and MYCS-elements (red marks) and their mutated versions (light yellow marks and light red marks, respectively). Bold letters indicate the basepairs, defining the respective motif. 4-MU, 4-methylumbelliferone. Bold black line, median; box, interquartile range; whiskers highest and lowest data point within 1.5 interquartile range; dots, actual values. Different letters indicate different statistical groups (ANOVA; post hoc Tukey; n = 6; P < 0.05). The experiment was performed twice with similar results. (B) Representative images of GUS activity resulting from pRAM2 and pRAM2m activation in wild-type hairy roots colonized with Rhizophagus irregularis at 8 wpi. (C) Root length colonization parameters of wild-type and ram2-1 hairy roots, colonized with R. irregularis at 8 wpi, transformed with the indicated expression cassettes. Different letters indicate different statistical groups (ANOVA; post hoc Tukey; P < 0.05). (D) Representative laser scanning confocal images of wild-type and ram2-1 hairy roots colonized by R. irregularis at 8 wpi, transformed with the indicated expression cassettes. Numbers indicate the root systems that displayed the phenotype shown in the image, among the number of analyzed root systems. The fungus is stained with wheat germ agglutinin (WGA)-Alexa-Fluor488. Scale bars: Upper panel 20 µm, Lower panel 10 µm. (E) Representative images of GUS activity resulting from pRAM2 and pRAM2m activation in noncolonized wild-type hairy roots at 6 wpp, coexpressing pUbi:RAM1 (+RAM1) or not (−RAM1). (B and E) Roots were stained with X-Gluc for 6 h. Numbers indicate the number of root systems that displayed staining as shown in the image, among the total number of analyzed root systems. (B–E), the experiments were performed once.

Fig. 2.

RAM1 interacts with AM-induced WRI transcription factors via its GRAS domain. (A) GAL4-based Y2H assay to test interaction of RAM1 as prey (AD) and five AM-induced WRI transcription factors as bait (BD). The schematic representation on Top illustrates the truncated F1 and M5 RAM1 versions and their length in amino acids. Expressed proteins, optical density of dropped yeast cultures, amino acids lacking from the medium and 3-AT concentration are provided in the figure. (B) Western blot images of Co-IP assays showing interaction of CBX1 and WRI5b with RAM1 in N. benthamiana leaves. Fractions loaded are input (Input) and immunoprecipitation (IP). Molecular weights of the bands of the protein standard, left side. Primary antibodies used for detection, right side. (A and B) The experiments were performed twice with similar results. (C) Interaction of CBX1 or WRI5b with RAM1 based on BiFC in cortex cells of colonized L. japonicus hairy roots 5wpi. Green fluorescence indicates interaction. Magenta fluorescence indicates arbuscules. All gene fusions were expressed under the control of their endogenous (RAM1, CBX1, or WRI5b) promoters. See SI Appendix, Table S1 for the number of observed cells. Size bars, 20 µm. nCit, N-terminal half of Citrine; cCit, C-terminal half of Citrine.

Fig. 3.

The conserved M2 motif in the C-terminus of AM-induced WRI transcription factors is important for the interaction with RAM1. (A and B) GAL4-based Y2H assay to test interaction of RAM1 as prey (AD) and CBX1 and its truncated versions as bait (BD). Schematic representations on Top illustrate the truncated versions of CBX1 and their length in amino acids. The experiments were performed twice with similar results. Truncations in (A) are based on (34). Truncations in (B) are based on the sequence of CBX1_C. All coding-sequence-containing plasmids were used in combination with the complementary EV as negative controls. Expressed proteins, optical density of dropped yeast cultures, amino acids lacking from the medium and 3-AT concentration are provided in the figure. (C) Alignment of the amino acid sequences of five AM-induced WRI transcription factors shown for CBX1 to be important for the interaction with RAM1 in (B). The color code indicates amino acids with similar physico-chemical properties (blue nonpolar, green polar, red basic, purple acidic, yellow proline, orange glycine). Dashes show gaps in the alignment. The alignment was performed in seaview5 (49) with the muscle algorithm (50). Alignment of full-length protein sequences is shown in SI Appendix, Fig. S4. Amino acid numbers on top of the alignment refer to the amino acid sequence of CBX1.

Fig. 4.

Structural implications of residues involved in the CBX1–RAM1 interaction (A) RoseTTafold-derived CBX1–RAM1 complex structure shown as cartoon. RAM1 is colored in light blue and CBX1 in salmon. (B) Close-up view of CBX1–RAM1 interacting helices. Residues in CBX1 interaction helix (r1) correspond to CBX1 291-304 SALGLLLQSSKFKE, which encompass the CBX1 M2/M2b domain and are colored in magenta. Residues highlighted in marine relate to RAM1 residues (r3) within 3 Å of CBX1 291-304 that may potentially form interactions. Red sphere is provided for spatial orientation. (C–G) Interacting residues are highlighted between CBX1 (magenta) and RAM1 (marine). The additional box in (E) represents the top view showing space filling by leucine residues as spheres. Measurements are shown in panels (F and G) to highlight (F) correct hydrogen bonding distance and (G) correct van der Waals distance network. (H) Table contains the codes of the amino acids of CBX1 and RAM1 that are highlighted in (B–G).

Fig. 5.

Mutation of RAM1 residues predicted to be involved in interaction with WRIs reduces interaction capability. (A) GAL4-based Y2H assay to test interaction of mutant versions of RAM1 (bait) and five AM-induced WRI transcription factors (prey). All coding-sequence-containing plasmids were used in combination with the complementary EV as negative controls. Expressed proteins, optical density of dropped yeast cultures, amino acids lacking from the medium and 3-AT concentration are provided in the figure. The schematic representation on top shows the positions of amino acid substitutions of the different mutant versions of RAM1. The experiment was performed three times with similar results. (B) Transactivation assay in N. benthamiana leaves showing activity of wild-type and mutated versions of RAM1 in combination with 5 AM-induced WRI proteins on the PT4 promoter. The genomic sequence encoding the proteins indicated at the y-axis was driven by the LjUbiquitin10 promoter. Bold black line, median; box, interquartile range; whiskers highest and lowest data point within 1.5 interquartile range; dots, actual values. Different letters indicate different statistical groups (ANOVA; post hoc Tukey; n = 12; P < 0.05). (C) Root length colonization parameters of wild-type and ram1-4 hairy roots, colonized with R. irregularis at 6 wpi, transformed with the indicated expression cassettes. Different letters indicate different statistical groups (ANOVA; post hoc Tukey; P < 0.05). (D) Representative laser scanning confocal images of wild-type and ram1-4 hairy roots colonized with R. irregularis at 6 wpi, transformed with the indicated expression cassettes. The numbers indicate the number of root systems that displayed the phenotype shown in the image, among the total number of analyzed root systems. Black scale bar, 50 µm. Insets show a close-up of arbuscules; white scale bar, 10 µm. The fungus is stained with WGA-Alexa-Fluor488. (B–D) The experiments were performed once.

Fig. 6.

Expression of WRI encoding genes during AM symbiosis. (A and B) Transcript accumulation of CBX1 and WRI3 in roots of the indicated genotypes colonized by R. irregularis at 6 wpi. Transcript accumulation was determined by RT-qPCR. The housekeeping gene Ubiquitin10 was used for normalization. Corresponding colonization data are displayed in SI Appendix, Fig. S13 A and B. (C) Transcript accumulation of WRI5a, WRI5b, and WRI5c in roots of the indicated genotypes colonized by R. irregularis at 5 wpi. Transcript accumulation was determined by RT-qPCR, and the housekeeping gene Ubiquitin10 was used for normalization. Corresponding colonization data are displayed in SI Appendix, Fig. S14. Statistical analysis: Welch t test [n = 3 (A), n = 4 (B and C), #P < 0.1, *P < 0.05, **P < 0,01, ***P < 0.001]. All experiments were performed twice with similar results.

Fig. 7.

Model of gene regulation by RAM1 and WRI transcription factors during AM. The expression of RAM1, CBX1 and WRI3 is activated by the CCaMK–CYCLOPS–DELLA complex via the AM-CYCRE in their promoters. DELLA can also activate the transcription of the three genes in the absence of CYCLOPS, suggesting an additional, so far unknown, DNA-binding transcription factor X (14). The transcription of CBX1 and WRI3 can also be induced by ectopic function of dominant active CCaMK³¹⁴ in the absence of CYCLOPS, suggesting that their promoters can be targeted by (a) currently unknown CCaMK phosphorylation target(s) Z. Subsequently the GRAS protein RAM1 interacts with the WRI transcription factors CBX1 and WRI3 and can induce transcription of RAM2, STR, and PT4 to enable lipid for phosphate exchange between plant and fungus. Likely, also the WRI transcription factor encoding genes WRI5a, b, and c are transcriptional targets of CBX1, WRI3, and RAM1. WRI5a, b, and c also interact with RAM1 and activate RAM2, STR, and PT4 promoters and potentially their own promoters.