Functional redundancy of transcription factors SlNOR and SlNOR-like1 is required for pollen development in tomato

Abstract

In tomato, SlNOR and SlNOR-like1, members of the NAC family of transcription factors (TFs), are known to play critical roles in regulating fruit ripening and are highly expressed in floral organs. However, their role in flower development remains unclear. In this study, we generated and functionally characterized a double knockout mutant, nor/nor-like1. Our findings reveal that the pollen abortion of the nor/nor-like1 impedes ovarian enlargement, resulting in fruit formation failure. Histological analyses demonstrate that the pollen wall collapse occurs during the mature pollen stage and leads to the abnormal pollen wall component deposition at the microspore stage, resulting in the male sterility in the double knockout mutant lines. Kyoto Encyclopedia of Genes and Genomes enrichment pathway analyses further suggest that the loss of SlNOR and SlNOR-like1 function affects several metabolic pathways related to pollen development, including 'ABC transporters', 'lipid metabolism', 'phenylpropanoid biosynthesis', 'hormone signal transduction', 'starch and sucrose metabolism', and 'cutin, suberine, and wax biosynthesis'. Furthermore, our results demonstrate that SlNOR and SlNOR-like1 could directly bind to the promoters of key genes associated with pollen wall formation and activate their expression, including ATP-binding cassette transporters of the G family (SlABCG8/9/23), ECERIFERUM (SlCER1), and glycine-rich protein (SlGRP92). These findings suggest that SlNOR and SlNOR-like1 may play a redundant role in the biosynthesis and transport of sporopollenin precursors, cuticular wax biosynthesis, and exine formation. In summary, our study highlights a previously uncharacterized role of SlNOR and SlNOR-like1 in tomato pollen wall formation and male fertility.

Figure 1

The expression pattern of SlNOR and SlNOR-like1 genes in the WT flower, and the floral phenotype of nor/nor-like1 compared with WT, nor#11, and nor-like1#1. a) The relative expression of SlNOR and SlNOR-like1 at different flower organs. (b) The relative expression of SlNOR and SlNOR-like1 in WT stamen at the different developmental stages. MMC, microspore mother cell stage; Tds: tetrad stage; MUM: middle uninucleate microspore stage; BM: binucleate microspore stage; MP: mature pollen stage. Error bars (a and b) indicate ±SD of three biological replicates. (c) The phenotype of nor/nor-like1 fruits after pollination compared with WT, nor#11, and nor-like1#1. (d) The pollen grains released on the mature anthers’ inner surface of WT, nor#11, nor-like1#1, and nor/nor-like1. Red marked as pollen grains on the inner surface of the anther. (e) The quantity of pollen released from the stamens through artificially assisted vibrations. Five floral pollen grains from WT, nor#11, nor-like1#1, and nor/nor-like1 were collected into the 200-μl centrifuge tube. (f) Pollen grains in clumps were dissected out from nor/nor-like1 anthers. Marker indicated pollen grains in clumps. Scale bars: (d and f) 1 cm.

Figure 2

Subcellular features of mature pollens in the nor/nor-like1 mutant lines compared with WT, nor#11 and nor-like1#1single mutant lines. (a and b) SEM indicated that the nor/nor-like1 pollen grains were collapsed and adhesive. There was no obvious difference between WT and single-gene knockout mutants. (c and d) TEM indicated that the structure of the nor/nor-like1 pollen wall was aberrant. There was no obvious difference between WT and single-gene knockout mutants. IF: intracellular fluid; Ex: exine; Se: sexine; Ne: nexine; Ba: baculum; In: intine; CPG: collapsed pollen grain. (e) 4′,6-diamidino-2-phenylindole staining of mature pollen grains. The nuclei were clearly present in the WT and single-gene knockout mutant pollen, but disappeared in the nor/nor-like1. SN: sperm nucleus; VN: vegetative nucleus

Figure 3

Transverse sections of flower buds at different pollen development stages in nor/nor-like1, compared with WT, nor#11, and nor-like1#1. MMC: microspore mother cell stage; Tds: tetrad stage; MUM: middle uninucleate microspore stage; BM: binucleate microspore stage; MP: mature pollen stage

Figure 4

Pollen viability of nor/nor-like1, compared with WT, nor#11, and nor-like1#1. (a and c) The in vitro TTC staining assay indicates that pollen viability of nor/nor-like1 is significantly lower. (b and d) The in vitro germination assay reveals that the pollen germination rate of nor/nor-like1 is significantly lower. (e) The in vitro germination assay shows a slower pollen tube elongation of nor/nor-like1 compared to WT and single mutant lines. Asterisks (c and d) indicate that the statistical rate is significantly higher than the actual situation due to adhering pollen influencing the counting process. Significant differences (P < 0.01) are indicated by lowercase letters. (f) The phenotypes from the back-crosses between nor/nor-like1 and WT, nor#11, nor-like1#1. The female pistils of nor/nor-like1 as female parents were successfully pollinated with pollens from WT and single knockout mutants. No observable developmental anomalies were detected in the nor/nor-like1 fruits. The pollinated flowers were marked with arrows for easy identification

Figure 5

Identification of differentially expressed genes (DEGs) in WT, nor#11, nor-like1#1, and nor/nor-like1 stamens before the mature pollen stage by transcriptome analyses. (a) DEG status between nor/nor-like1 and WT/nor#11/nor-like1#1, nor#11/nor-like1#1 and WT; (b and c) Venn diagram and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment pathway analysis; (d) The expression level of DEGs related to pollen development in WT, nor#11, nor-like1#1, and nor/nor-like1 stamens

Figure 6

SlNOR and SlNOR-like1 directly bind to SlABCG8/9/23, SlCER1, and SlGRP92 promoters and activate their expression. The EMSA shows that SlNOR (a) and SlNOR-like1 (d) directly bind to the target gene promoters containing the NACRS (NAC recognized sequence). The biotin probes undergo a significant shift upon binding with proteins. The dual-luciferase reporter assay (DLR) indicates that SlNOR (b) and SlNOR-like1 (e) interact with promoter regions of five selected genes and activate their expression. ChIP-qPCR analysis confirms the direct binding of SlNOR (c) and SlNOR-like1 (f) to the target gene promoters. The values represent the percentage of DNA fragments in over-expressing (OE)-GFP (NOR/NOR-like1 native promoter::GFP) (CK) or over-expressing (OE)-NOR/NOR-like1 (NOR/NOR-like1 native promoter: NOR/NOR-like1-GFP) fruit that coimmunoprecipitated with an anti-GFP tag antibody relative to the corresponding input DNA. The error bars indicate ±SD of three biological replicates. Asterisks indicate significant differences determined by Student’s t-test (^*P < 0.05, ^**P < 0.01, ^***P < 0.001)

Figure 7

A working model for the SlNOR and SlNOR-like1 mediated regulation of tomato pollen development. In detail, SlNOR and SlNOR-like1 positively regulate the ABCG transporter genes SlABCG8/9/23 for transporting sporopollenin precursor, the SlGRP92 gene for pollen exine formation, and the SlCER1 gene for the cuticular wax biosynthesis for pollen exine composition and stamen development. SlNOR and SlNOR-like1 double knockout mutant (nor/nor-like1) exhibit the pollen abortion resulted from the pollen wall collapse and hindered pollen transmission. Positive effects are indicated by arrows