Down-Regulation of OsEMF2b Caused Semi-sterility Due to Anther and Pollen Development Defects in Rice

Abstract

Anther and pollen development are crucial processes of plant male reproduction. Although a number of genes involved in these processes have been identified, the regulatory networks of pollen and anther development are still unclear. EMBRYONIC FLOWER 2b (OsEMF2b) is important for rice development. Its biological function in floral organ, flowering time and meristem determinacy have been well-studied, but its role, if only, on male reproduction is still unknown, because null mutants of OsEMF2b barely have anthers. In this study, we identified a weak allele of OsEMF2b, osemf2b-4. The T-DNA insertion was located in the promoter region of OsEMF2b, and OsEMF2b expression was significantly decreased in osemf2b-4. The osemf2b-4 mutant exhibited much more normal anthers than null mutants of OsEMF2b, and also showed defective floret development similar to null mutants. Cytological analysis showed various defects of anther wall and pollen development in osemf2b-4, such as slightly or extremely enlarged tapetum, irregular or normal morphology microspores, and partial or complete sterility. OsEMF2b was highly expressed in tapetum and microspores, and the protein was localized in the nucleus. The expression of 15 genes essential for anther and pollen development was investigated in both WT and osemf2b-4. Fourteen genes including GAMYB was up-regulated, and only PTC1 was down-regulated in osemf2b-4. This suggests that up-regulated GAMYB and down-regulated PTC1 might contribute to the defective anther and pollen development in osemf2b-4. Overall, our work suggests that OsEMF2b plays an essential role during post-meiotic anther and pollen development.

Keywords: OsEMF2b; anther development; male reproductive development; rice; semi-sterility.

FIGURE 1

Phenotypic and RT-qPCR analyses of osemf2b-4. (A) Expression levels of OsEMF2b in the WT or osemf2b-4 leaves were detected by RT-qPCR. Error bars indicate SD. ^∗∗ indicates significant difference by Student’s t-test (P ≤ 0.01). (B) Plant phenotypes of wild type (left) and osemf2b-4 (right) at maturity. Bars = 5 cm. (C) Panicles of wild type (left) and osemf2b-4 (right). Abnormal seeds are marked with white arrow in the osemf2b-4 panicle. Data are given as mean ± SEM (n = 3). Bars = 2 cm. (D) Seed phenotypes of wild type and osemf2b-4. The numbers to the left were calculated as the proportion of slightly defective types of abnormal seeds to spikelets for each plant at maturity. The 10.3% remaining were sterile spikelets. Bars = 5 mm.

FIGURE 2

Floret and pollen fertility comparisons of the wild type and osemf2b-4. (A) A wild type spikelet; (B–E) four types of slight floret defects with stamens including small husk (B), beak-like (C), open husk (D), and long sterile lemma (E) spikelets in osemf2b-4 plants, bars = 2 mm; (F–J) spikelets after removing the lemma and palea, bars = 2 mm; (K–R) I₂-KI pollen staining, (L–O) pollen from yellow anthers, (P–R) pollen from three white anthers (white arrowheads), respectively. Bars = 200 μm. (a–d) Four types of severe floret defects with stamens such as double lemma (a), coiled lemma and palea (b), missing lemma or palea (c), and leaf-like (d) spikelets in osemf2b-4 mutant, Bars = 2 mm; (e–h) spikelets after removing the lemma and palea, bars = 2 mm; (i–l) Pollen fertility using I₂-KI pollen staining. Numbers indicate the ratio of different types of floral organ defects with stamens per plant. Bars = 200 μm.

FIGURE 3

Histological comparison of anther development in the wild type and osemf2b-4. Three stages of anther development in the wild type and osemf2b-4 are compared. The images are of cross sections through a single locule. Wild type sections are shown in (A,E,I), and others show osemf2b-4 sections. (A–D) Young microspore stage; (E–H) vacuolated pollen stage; (I–L) mature pollen stage. Numbers indicate the percentage of different anther defects at one stage. E, epidermis; En, endothecium; ML, middle layer; T, tapetum; Ms, microsporocyte; Msp, microspore; MP, mature pollen. Bars = 15 μm.

FIGURE 4

RNA in situ analysis of OsEMF2b. (A) A wild type anther at the meiosis stage showing OsEMF2b expression in microsporocytes, tapetum and connectives of anther. (B) A wild type anther at the young microspore stage showing stronger OsEMF2b expression in tapetal cells, and weaker expression in microspores and connectives of anther. (C) A wild type anther at young microspore stage with the sense probe. Ms, microsporocyte; T, tapetum; Msp, microspore; Vb, vascular bundle of anther connective. Bars = 15 μm.

FIGURE 5

Structure and nuclear localization of OsEMF2b. Schematic representation of OsEMF2b. The black boxes indicate two putative nuclear localization signals (NLS) with a basic segment (Lys-Lys-Arg-Lys-Arg), and the gray boxes indicate a C₂H₂-type zinc finger motif and the VRN2-EMF2-FIS2-SUZ12 (VEFS) domain, respectively. (A) Nuclear localization of OsEMF2b in rice protoplasts. (B) (a–c) Images of mCherry. (d–f) Images of mCherry-OsEMF2b fusion. (a,d) Bright-field images. (b,e) Red fluorescence images. (c,f) Merged images of (a,b,d,e), respectively. Bars = 2 μm in (a–c), 4 μm in (d–f).

FIGURE 6

RT-qPCR expression analysis of OsEMF2b and genes involved in rice anther and pollen development in the inflorescence of the wild type and osemf2b-4 plants. Stage 1 and stage 2 represent meiosis stage and young microspore stage, respectively. The ubiquitin gene was used as an internal control. Each data point is the average of three biological repeats. Error bars indicate SD. ^∗,^∗∗ indicate significant difference by Student’s t-test (P ≤ 0.05 and 0.01).