Regulation of Jasmonate-Mediated Stamen Development and Seed Production by a bHLH-MYB Complex in Arabidopsis

Abstract

Stamens are the plant male reproductive organs essential for plant fertility. Proper development of stamens is modulated by environmental cues and endogenous hormone signals. Deficiencies in biosynthesis or perception of the phytohormone jasmonate (JA) attenuate stamen development, disrupt male fertility, and abolish seed production in Arabidopsis thaliana. This study revealed that JA-mediated stamen development and seed production are regulated by a bHLH-MYB complex. The IIIe basic helix-loop-helix (bHLH) transcription factor MYC5 acts as a target of JAZ repressors to function redundantly with other IIIe bHLH factors such as MYC2, MYC3, and MYC4 in the regulation of stamen development and seed production. The myc2 myc3 myc4 myc5 quadruple mutant exhibits obvious defects in stamen development and significant reduction in seed production. Moreover, these IIIe bHLH factors interact with the MYB transcription factors MYB21 and MYB24 to form a bHLH-MYB transcription complex and cooperatively regulate stamen development. We speculate that the JAZ proteins repress the bHLH-MYB complex to suppress stamen development and seed production, while JA induces JAZ degradation and releases the bHLH-MYB complex to subsequently activate the expression of downstream genes essential for stamen development and seed production.

Figure 1.

JAZ Proteins Interact with MYC5. (A) Schematic diagrams show domain constructs of MYC5 and JAZ11. The diagrams display the conserved JAZ-interaction domain (JID; green), bHLH domain (orange), ZIM domain (purple), and Jas domain (blue). The numbers indicate positions of the first and last amino acids of the domain constructs. (B) Y2H assay to test the interactions of 12 Arabidopsis JAZs with MYC5 and its related domains (as shown in [A]). JAZs were fused with the LexA DNA binding domain in pLexA, and MYC5 and its domains were fused with the activation domain (AD) in pB42AD. Interactions (represented by blue color) were assessed on 2% Gal/1% raffinose/SD/-Ura/-His/-Trp/-Leu/X-β-Gal medium. (C) BiFC assay to detect the interactions of JAZ1 and JAZ10 with MYC5. JAZ1, JAZ10, and MYC5 were fused with the N-terminal fragment of YFP (nYFP) or the C-terminal fragment of YFP (cYFP) to generate JAZ1-nYFP, JAZ10-nYFP, and cYFP-MYC5. YFP fluorescence was detected 50 h after coexpression of the indicated construct pairs in leaves of N. benthamiana. The nuclei are indicated by 4,6-diamidino-2-phenylindole dihydrochloride (DAPI) staining. (D) Y2H assay to test the interactions of JAZ11 domain constructs (as shown in [A]) with MYC5. MYC5 was fused with the activation domain, and different JAZ11 domains were fused with the DNA binding domain.

Figure 2.

The bHLH Subgroup IIIe Factors Function Redundantly to Regulate Stamen Development. (A) Phenotypic observation of anther dehiscence of the indicated genotypes at floral stage 13, 14, or 15. Red arrowheads represent indehiscent anthers. (B) In vitro germination of pollen grains from the flowers at floral stage 13, 14, or 15 in the indicated genotypes. Red arrowheads indicate pollen grains deficient in germination in vitro. (C) Comparison of flowers at floral stage 13, 14, or 15 in the Col-0 wild type, myc5, myc2 myc5 (myc2/5), myc2 myc3 myc5 (myc2/3/5), myc2 myc3 myc4 myc5 (myc2/3/4/5), coi1-1, and myb21-3 myb24 (myb21/24). Red arrowheads indicate shorter filaments. (D) Staining of pollen grains from Col-0 and myc2 myc3 myc4 myc5 (myc2/3/4/5) flowers at floral stage 13 or 15 with fluorescein diacetate and propidium iodide. Viable pollen grains are green, and unviable pollen grains are red. (E) Filament length and pistil length at floral stage 13 in the indicated genotypes. Values are means ± se from three biological replicates. Asterisks represent Student’s t test significance compared with the Col-0 wild type (**P < 0.01). (F) Silique length of Col-0 and myc2 myc3 myc4 myc5 (myc2/3/4/5). Values are means ± se from three biological replicates. Asterisks represent Student’s t test significance compared with the Col-0 wild type (**P < 0.01). (G) Siliques from Col-0, myc5, myc2 myc5 (myc2/5), myc2 myc3 myc5 (myc2/3/5), and myc2 myc3 myc4 myc5 (myc2/3/4/5). (H) Seed numbers per silique of Col-0 and myc2 myc3 myc4 myc5 (myc2/3/4/5). Values are means ± se from three biological replicates. Asterisks represent Student’s t test significance compared with the Col-0 wild type (**P < 0.01).

Figure 3.

Overexpression of MYC5 Partially Restores Stamen Development in coi1-1. (A) Comparison of flowers, anthers, and in vitro germination of pollen grains from flowers at floral stage 13 in the Col-0 wild type, coi1-1, and coi1-1 plants overexpressing MYC5 (coi1-1 MYC5OE1 and coi1-1 MYC5OE2). (B) Filament length and pistil length in flowers at floral stage 13 from Col-0, coi1-1, and coi1-1 plants overexpressing MYC5 (coi1-1 MYC5OE1 and coi1-1 MYC5OE2). Values are means ± se from three biological replicates. Asterisks represent Student’s t test significance compared with coi1-1 (**P < 0.01). (C) Main inflorescences in the indicated genotypes. Red arrows indicate fertile siliques in coi1-1 MYC5OE1 and coi1-1 MYC5OE2. (D) Comparison of seed set in the indicated genotypes.

Figure 4.

Overexpression of MYC3 Partially Restores Stamen Development in coi1-1. (A) Comparison of flowers, anthers, and in vitro germination of pollen grains from flowers at floral stage 13 in the Col-0 wild type, coi1-1, and coi1-1 plants overexpressing MYC3 (coi1-1 MYC3OE1 and coi1-1 MYC3OE2). (B) Filament length and pistil length in flowers at floral stage 13 from Col-0, coi1-1, and coi1-1 plants overexpressing MYC3 (coi1-1 MYC3OE1 and coi1-1 MYC3OE2). Values are means ± se from three biological replicates. Asterisks represent Student’s t test significance compared with coi1-1 (**P < 0.01). (C) Main inflorescences in the indicated genotypes. Red arrows indicate fertile siliques in coi1-1 MYC3OE1 and coi1-1 MYC3OE2. (D) Comparison of seed set in the indicated genotypes.

Figure 5.

MYC2, MYC3, MYC4, and MYC5 Are Required for JA-Inducible Expression of MYB21, MYB24, MYB57, and MYB108. (A) RT-qPCR analysis for MYB21, MYB24, MYB57, and MYB108 in young flower buds of the Col-0 wild type, coi1-1, and myc2 myc3 myc4 myc5 (myc2/3/4/5) treated without (Mock) or with methyl jasmonate (JA). ACTIN8 was used as the internal control. Values are means ± se from three biological replicates. (B) Comparison of flowers, anther dehiscence, and in vitro germination of pollen grains at floral stage 13 in Col-0, coi1-1, opr3, myc2 myc3 myc5 (myc2/3/5), and myc2 myc3 myc4 myc5 (myc2/3/4/5) treated without (Mock) or with methyl jasmonate (JA). White arrowheads indicate fully elongated filaments. (C) and (D) RT-qPCR analysis of gene expression in Col-0, coi1-1, and coi1-1 plants overexpressing MYC5 (coi1-1 MYC5OE1 and coi1-1 MYC5OE2) (C) or coi1-1 plants overexpressing MYC3 (coi1-1 MYC3OE1 and coi1-1 MYC3OE2) (D). Expression of MYC5 (C) and MYC3 (D) was detected in 3-week-old plants, and expression of MYB21, MYB24, MYB57, and MYB108 was tested in young flower buds of the indicated genotypes using ACTIN8 as the internal control. Values are means ± se from three biological replicates. Asterisks represent Student’s t test significance compared with the Col-0 wild type (**P < 0.01).

Figure 6.

MYC2, MYC3, MYC4, and MYC5 Interact with MYB21 and MYB24. (A) Schematic diagrams of MYB21, MYB24, and their N-terminal domains (MYB21NT and MYB24NT). MYB21NT and MYB24NT contain the conserved R2 (red) and R3 (blue) domains but not the NYW^G/_S^M/_VDD^I/_LW^S/_P motif (orange). (B) Y2H assay to show interactions between MYB21NT/MYB24NT and MYC5. MYB21NT and MYB24NT were fused with the DNA binding domain (BD), and MYC5 was fused with the activation domain (AD). (C) Co-IP assay to verify the interaction between MYC5 and MYB21 in vivo. Flag-MYC5 was coexpressed without (empty control) or with myc-MYB21 or myc-COI1 (negative control) in N. benthamiana leaves. The total protein extracts from N. benthamiana leaves transiently expressing flag-MYC5, flag-MYC5 plus myc-MYB21, or flag-MYC5 plus myc-COI1 were immunoprecipitated using the anti-c-myc antibody-conjugated agarose and were blotted with anti-flag or anti-c-myc antibody. (D) BiFC assay to detect interactions of MYC2, MYC3, MYC4, and MYC5 with MYB21 and MYB24. MYC2, MYC3, MYC4, and MYC5 were fused with nYFP, and MYB21 and MYB24 were fused with cYFP. The nuclei are indicated by 4,6-diamidino-2-phenylindole dihydrochloride (DAPI) staining.

Figure 7.

JAZ1 Attenuates the Transcriptional Activation Function of MYC2, MYC3, MYC4, MYC5, MYB21, and MYB24. (A) Schematic diagrams of the constructs used in the transient expression assays in (B). (B) Transient expression assays show that JAZ1 inhibits the transcriptional activation function of MYC2, MYC3, MYC4, MYC5, MYB21, and MYB24. Values are means ± se from three biological replicates. Asterisks represent Student’s t test significance between pairs indicated with brackets (**P < 0.01).

Figure 8.

The bHLH and MYB Factors Cooperatively Mediate JA-Regulated Stamen Development. (A) and (B) Phenotypes of flowers, stamens, and in vitro pollen germination in the Col-0 wild type, myc2 myc3 myc4 myc5 (myc2/3/4/5), myb21-t1 myb24, and myc2 myc3 myc4 myc5 myb21-t1 myb24 (myc2/3/4/5 myb21-t1/24) at floral stage 13 (A) and floral stage 15 (B). (C) Inflorescences in the indicated genotypes. Red arrows indicate fertile siliques of myc2 myc3 myc4 myc5 and myb21-t1 myb24.

Figure 9.

A Simplified Model for JA-Regulated Stamen Development. JAZ repressors interact with and repress the bHLH-MYB complex consisting of the IIIe bHLH factors MYC2, MYC3, MYC4, and MYC5 (indicated as bHLH) and the R2R3 MYB factors MYB21 and MYB24 (indicated as MYB). JA is perceived by COI1 to induce JAZ degradation and release the bHLH-MYB complex, which regulates the expression of downstream genes (as well as some members of the bHLH-MYB complex) to mediate stamen development. Gibberellin (GA) may also act through the bHLH-MYB complex to regulate stamen development: DELLAs repress the biosynthesis of JA in flowers (Cheng et al. 2009) to suppress JA-regulated stamen development. In addition, DELLAs directly interact with and repress MYC2 (Hong et al. 2012). Further experiments are required to verify whether the interaction of DELLAs with MYC2 or other unidentified components (indicated as Others) modulates gibberellin-regulated stamen development.