ZmEREB92 plays a negative role in seed germination by regulating ethylene signaling and starch mobilization in maize

Abstract

Rapid and uniform seed germination is required for modern cropping system. Thus, it is important to optimize germination performance through breeding strategies in maize, in which identification for key regulators is needed. Here, we characterized an AP2/ERF transcription factor, ZmEREB92, as a negative regulator of seed germination in maize. Enhanced germination in ereb92 mutants is contributed by elevated ethylene signaling and starch degradation. Consistently, an ethylene signaling gene ZmEIL7 and an α-amylase gene ZmAMYa2 are identified as direct targets repressed by ZmEREB92. OsERF74, the rice ortholog of ZmEREB92, shows conserved function in negatively regulating seed germination in rice. Importantly, this orthologous gene pair is likely experienced convergently selection during maize and rice domestication. Besides, mutation of ZmEREB92 and OsERF74 both lead to enhanced germination under cold condition, suggesting their regulation on seed germination might be coupled with temperature sensitivity. Collectively, our findings uncovered the ZmEREB92-mediated regulatory mechanism of seed germination in maize and provide breeding targets for maize and rice to optimize seed germination performance towards changing climates.

Fig 1. Loss-of-function of ZmEREB92 enhanced seed germination in maize by promoting embryo growth during imbibition.

(A) The germination performance for KN5585 and ereb92 mutants (ereb92-2 and ereb92-6) at 5 days after imbibition (DAI) on the filter paper infiltrated with sterile water. (B) The time-course germination of KN5585 and ereb92 mutants for 1–7 DAI under normal condition. Error bars indicate mean ± SE (n = 3). (C) The phenotype of seedling emergence of KN5585 and ereb92 mutants after 10 days sowing in the soil. (D) The seedling emergence rates of KN5585 and ereb92 mutants were counted at 4, 5, 6, 8, 10 days after sowing in the soil. Error bars indicate mean ± SE (n = 3). (E) The longitudinal section of KN5585 and ereb92 mutant seeds at 0, 6, 24 and 36 HAI. The embryo region was marked with yellow dash line. (F) The percentage of embryo for KN5585 and ereb92 mutant seeds at 0, 6, 24 and 36 HAI. The relative embryo proportion is calculated with the formula: Embryo area/ whole seed area *100%. The areas were calculated by ImageJ software. Error bars indicate mean ± SE (n = 5). (B, D, F) Asterisks indicate significant difference from the control (KN5585) at each time point (Two-way ANOVA followed by Tukey test, **P <0.01). (G) The schematic longitudinal view of the embryo structure in maize. (H) The histological sectioning and cytological analysis of the radicle and plumule region of KN5585 and ereb92-6 seeds at 36 HAI. (I, K) The cell size in radicle (I) and plumule (K) of KN5585 and ereb92-6 indictaed in (G), respectively. A total of 50 cells from three different sections of each line were counted. (J, L) The cell number of radicle and plumule of KN5585 and ereb92-6 seeds indicated in (G). (G-J) The distribution of cell size is displayed by boxplot. Error bars indicates the value range and the box shows the medium and the upper and lower quartiles. The circles represent for individual datapoints of biological replicates in each line. Number marked for each data is the exact P value (Student’s t-test, *P<0.05, **P<0.01).

Fig 2. Transcriptomic profile reveals the role of ZmEREB92 in regulating hormone-related pathways.

(A, B) A and B, Number of DEGs (Log₂ (|FC|) ≥1, P<0.05) (A) and the Venn diagram shows the number of overlapped DEGs (B) in the comparison groups of ereb92_6h vs KN5585_6h, ereb92_6h vs ereb92_0h, KN5585_6h vs KN5585_0h and ereb92_6h vs KN5585_6h. Three independent experiments were performed for each sample at each time point. (C, D) GO (C) and KEGG (D) analysis of the DEGs in the comparison groups of ereb92_6h vs ereb92_0h and KN5585_6h vs KN5585_0h. (E) The heatmap shows the log₂FC (Fold change) of differentially expressed hormone-related genes in the comparison groups of ereb92_6h vs ereb92_0h or ereb92_6h vs KN5585_6h. (F) qPCR analysis for several selected hormone-related genes in seeds of ere92 mutant and KN5585 at 6 HAI. Error bars indicate mean ± SE (n = 4). Ef1a was used as the reference gene and relative expression level was normalized to one biological replicate of KN5585. Asterisks indicate significant difference (Student’s t-test, *P<0.05, **P<0.01).

Fig 3. Ethylene signaling plays a major to promote seed germination in ereb92 mutants.

(A) The germination performance at the for KN5585 and ereb92 mutants at 3 DAI under normal condition (CK), 50 μM ABA treatment, 100 mg/L Paclobutrazol (PAC, inhibitor of GA biosynthesis) treatment and 200 mg/L 1-Methylcyclopropene (1-MCP, inhibitor of ET receptor). (B) The germination rate of KN5585 and ereb92 mutants counted at 1, 1.5, 2 and 3 DAI under the treatments described in (A). Error bars indicate mean ± SE (n = 4). (C) The heatmap showed the Log₂FC of the hormone level relative to KN5585_0h. (D) The longitudinal section of the seeds of KN5585 and ereb92 mutants at 0, 24 and 36 HAI under 1-MCP treatment. Embryo region was sketched with yellow dash line. (E) The percentage of embryo for seeds of KN5585 and ereb92 mutant at 0, 24 and 36 HAI under 1-MCP treatment. The relative embryo proportion is calculated by ImageJ software. The circles are represented for individual datapoints of biological replicates in each line. Error bars indicate mean ± SE (n = 7). n. s. indicates no significant difference using one-way ANOVA followed by Tukey tests (P>0.05).

Fig 4. ZmEREB92 inhibits the transcription of ZmEIL7 by directly binds to the G-boxes in ZmEIL7 promoter.

(A) Dual-Luciferase Reporter (DLR) assays in maize protoplast showing that ZmEREB92 negatively regulated ZmEIL7 transcription through the -1188 to -1500 bp region upstream to the start codon. The different promoter region of ZmEIL7 were co-transformed with empty vector (EV) or ZmEREB92. p35S-REN was used as the internal control. Error bars indicate mean ± SE (n = 3). Asterisks indicate significant difference (Student’s t-test, **P<0.01). (B) The repression of ZmEREB92 on ZmEIL7 transcription was dependent on two EAR-motif at C terminus. The EAR-motif was indicated by black line. Single or double mutations of two EAR-motifs (ΔEAR-1, ΔEAR-2 and ΔEAR-1/2) in ZmEREB92 were generated and co-transformed with ZmEIL7 promoter, respectively. Error bars indicate mean ± SE (n = 3). Different lowercase letters represent significant differences (one-way ANOVA followed by Tukey tests, P<0.05). (C) Yeast one hybrid assays suggest the direct binding of ZmEREB92 to ZmEIL7 promoter. The promoter of ZmEIL7 was co-transformed with AD-ERBE92 or AD-EREB92(ΔEAR1/2) and grown on selective medium (SD/-Ura/-Leu/AbA). AbA, Aureobasidin, 400 ng·mL^-1. Empty pGADT7 vector (AD) was also co-transformed as the negative control. (D) EMSA to show the direct binding of ZmEREB92 to the GCC-boxes in ZmEIL7 promoter. Two fragments containing GCC-boxes in ZmEIL7 promoter were labeled with biotin and incubated with ZmERBE92 purified recombinant proteins. 200-fold excess of unlabeled probes were used for competition. (E, F) The expression pattern of ZmEIL7 in KN5585 and ereb92-6 mutant at 2, 6, 12, 24 and 36 HAI under control (E) and 1-MCP treatment (F). Error bars indicate mean ± SE (n = 3). Ef1a was used as the reference gene and relative expression level was normalized to one biological replicate of 2 HAI of KN5585. Different lowercases or majuscules represent significant difference in KN5585 or ereb92-6 mutant, respectively (one-way ANOVA followed by Tukey tests, P<0.05). Asterisks indicate significant difference between KN5585 and ereb92-6 mutant at each time point (Student’s t-test, **P<0.01, n.s. no significant difference).

Fig 5. ZmEREB92 directly inhibit ZmAMYa2 transcription and resulted in accelerated starch mobilization in imbibed seeds of ereb92 mutants.

(A) The germination performance at 24 HAI using separated embryos of KN5585 and ereb92-6 mutant. (B) The germination rate at 0, 24, 36 and 48 HAI of KN5585 and ereb92-6 mutant using separated embryos. Error bars indicate mean ± SE (n = 3). (C) α-amylase activity in KN5585 and ereb92 mutants at 24 HAI. Error bars indicate mean ± SE (n = 7). Asterisks indicate significant difference (Student’s t-test, *P<0.05). (D) The relative expression of ZmAMYa2 in ereb92-6 mutant and KN5585 at 6 HAI. Error bars indicate mean ± SE (n = 3). Ef1a was used as the reference gene and relative expression level was normalized to one biological replicate of KN5585. Asterisks indicate significant difference (Student’s t-test, **P<0.01). (E) DLR assays in maize protoplast showing that ZmEREB92 negatively regulated ZmAMYa2 transcription and such repression was partially dependent on two EAR-motifs. The promoter of ZmAMYa2 were co-transformed with EV, ZmEREB92 or ZmEREB92(ΔEAR1/2), respectively. p35S-REN was used as the internal control. Error bars indicate mean ± SE (n = 3). Different lowercases indicate significant difference (one-way ANOVA followed by Tukey tests, P<0.05). (F) EMSA to show the direct binding of ZmEREB92 to the GCC-boxes in A and C fragments of ZmAMYa2 promoter. Three fragments containing GCC-boxes in ZmAMYa2 promoter were labeled with biotin and incubated with ZmERBE92. 200-fold excess of unlabeled probes were used for competition. (G, H) The expression pattern of ZmAMYa2 in KN5585 and ereb92-6 at 2, 6, 12, 24 and 36 HAI under control (G) and 1-MCP treatment (H). Error bars indicate mean ± SE (n = 3). Ef1a was used as the reference gene and relative expression level was normalized to one biological replicate of 2 HAI of KN5585. Different lowercases or majuscules represent significant difference in KN5585 or ereb92-6 mutant, respectively (one-way ANOVA followed by Tukey tests, P<0.05). Asterisks indicate significant difference between KN5585 and ereb92-6 mutant at each time point (Student’s t-test, **P<0.01, n.s. no significant difference).

Fig 6. Rice ortholog of ZmEREB92 negatively regulates seed germination and underwent selection during rice domestication.

(A) Syntenic analysis of ZmEREB92 and OsERF74 from maize chromosome 8 (B73 RefGen_v5, pink) and rice chromosome 5 (Nipponbare IRGSP-1.0, light blue), respectively. (B) Phylogenetic analysis of ZmEREB92 and its orthologs in major cereal crops and Arabidopsis using neighbor-joining method. The numbers shown next to the branches indicate bootstrap values from 1000 replicates. (C) The germination performance of Kitaake and Oserf74 mutants at 3 DAI under normal condition (25 °C). (D) Time-course seed germination rate from 0–7 DAI for Kitaake and Oserf74 mutants under normal condition. Error bars indicate mean ± SE (n = 3). Asterisks indicate significant difference from the control (Kitaake) at each time point (Two-way ANOVA followed by Tukey test, *P <0.05, **P <0.01). (E) Nucleotide diversity across the OsERF74 locus among O. sativa japonica (blue), O. sativa indica (green) and O. Rufipogon (pink). The genomic region of OsERF74 is shown at the X axis. (F) Nucleotide diversity across the ZmEREB92 locus among maize (blue) and teosinte (pink). The genomic region of ZmEREB92 is shown at the X axis.

Fig 7. The orthologous gene pair ZmEREB92/OsERF74 show conserved functions in seed germination under cold stress.

(A) Relative expression analysis of cold-related genes in seeds of ere92-6 mutant and KN5585 at 6 HAI. Ef1a was used as the reference gene and relative expression level was normalized to one biological replicate of KN5585. (B) DLR assays for the promoters of ZmCESA1 and ZmCESA6 co-transformed with EV or ZmEREB92. (C) The expression pattern of ZmEREB92 during imbibition under normal (28 °C) and cold (12 °C) condition, respectively. Ef1a was used as the reference gene and relative expression level was normalized to one biological replicate of control-2HAI. (A-C) Error bars indicate mean ± SE (n = 3). Asterisks indicate significant difference (Student’s t-test, *P<0.05, **P<0.01). (D) The seed germination performance of KN5585 and ereb92 mutants at 6 DAI under cold condition (12 °C). (E) Time-course seed germination rate from 1–7 DAI for KN5585 and ereb92 mutants. Error bars indicate mean ± SE (n = 3). (F) The germination performance of Kitaake and Oserf74 mutants at 6 DAI under cold condition (15 °C). (G) Time-course seed germination rate from 0–7 DAI for Kitaake and Oserf74 mutants under cold condition. Error bars indicate mean ± SE (n = 3). (E, G) Asterisks indicate significant difference from the control (Kitaake) at each time point (Two-way ANOVA followed by Tukey test, **P <0.01).

Fig 8. Proposed model for the temperature-sensitive regulation on seed germination by ZmEREB92.

When maize seeds are germinating in normal condition, inhibited expression of ZmEREB92 leads to the release of ZmEIL7 and ZmAMYa2 transcription, which enhance ethylene signaling and endosperm starch degradation, thereby promoting embryo growth and ensuring timely germination for maize seeds. When the condition is cold, the expression of ZmEREB92 is drastically induced in imbibed seeds, which in turn suppresses the transcription of ZmEIL7 and ZmAMYa2 by directly binding to the GCC-box enriched region in their promoters, ultimately resulting in blocked seed germination.