Construction of a male sterility system for hybrid rice breeding and seed production using a nuclear male sterility gene

Abstract

The breeding and large-scale adoption of hybrid seeds is an important achievement in agriculture. Rice hybrid seed production uses cytoplasmic male sterile lines or photoperiod/thermo-sensitive genic male sterile lines (PTGMS) as female parent. Cytoplasmic male sterile lines are propagated via cross-pollination by corresponding maintainer lines, whereas PTGMS lines are propagated via self-pollination under environmental conditions restoring male fertility. Despite huge successes, both systems have their intrinsic drawbacks. Here, we constructed a rice male sterility system using a nuclear gene named Oryza sativa No Pollen 1 (OsNP1). OsNP1 encodes a putative glucose-methanol-choline oxidoreductase regulating tapetum degeneration and pollen exine formation; it is specifically expressed in the tapetum and miscrospores. The osnp1 mutant plant displays normal vegetative growth but complete male sterility insensitive to environmental conditions. OsNP1 was coupled with an α-amylase gene to devitalize transgenic pollen and the red fluorescence protein (DsRed) gene to mark transgenic seed and transformed into the osnp1 mutant. Self-pollination of the transgenic plant carrying a single hemizygous transgene produced nontransgenic male sterile and transgenic fertile seeds in 1:1 ratio that can be sorted out based on the red fluorescence coded by DsRed Cross-pollination of the fertile transgenic plants to the nontransgenic male sterile plants propagated the male sterile seeds of high purity. The male sterile line was crossed with ∼1,200 individual rice germplasms available. Approximately 85% of the F1s outperformed their parents in per plant yield, and 10% out-yielded the best local cultivars, indicating that the technology is promising in hybrid rice breeding and production.

Keywords: OsNP1; breeding; hybrid rice; hybrid seed production; male sterility.

Fig. 1.

Phenotypes of osnp1-1 and molecular identification of OsNP1. (A) A WT plant and osnp1-1 mutant plant after bolting. (Scale bar, 10 cm.) (B) Spikelets of WT and osnp1-1 with the palea and lemma removed. (Scale bar, 1 mm.) (C) I₂-KI staining of the WT and osnp1-1 pollen grains. (Scale bar, 100 μm.) (D) Inflorescence of osnp1-1, showing stigma extrusion. (Scale bar, 2 mm.) (E) The identification of osnp1 alleles. The Top section showed distributions of Euclidean distance (ED) scores of SNP sites along chromosomes in osnp1-1. OsNP1 gene structure and mutation sites of osnp1-1, osnp1-2, and osnp1-3 are shown in the Bottom section. The genomic fragment for gene complementation is shown by the double-headed black arrow. Empty boxes represent 5′- and 3′-UTRs, blue boxes represent exons, and the lines between boxes represent introns.

Fig. S1.

Alignment of the Nipponbare, WT HHZ, and osnp1-1 mutant OsNP1 protein sequences. Sequence codes for Nipponbare and HHZ OsNP1 sequences are KX066198 and KX066199, respectively. A₂₀A₂₁ and A₁₅₁ in HHZ indicated by red letter and asterisks are polymorphic amino acids compared to Nipponbare. Amino acid D₅₆₃ in osnp1-1 is the mutation site.

Fig. S2.

Phenotype of other osnp1 alleles and transgenic lines. (A–D) Phenotype of WT and osnp1-2 in the Zhonghua11 background. (E–H) Phenotype of WT and osnp1-3 in the Wuyunjing7 background. (I–L) Transgenic complementation line in osnp1-1. Complementation line was in osnp1-1 background and transformed with OsNP1_pro::OsNP1 construct. (M) qPCR analysis of OsNP1 expression in spikelets at anther developmental stage 12 of WT, osnp1-1, and OE transgenic lines. The OE transgenic line was in osnp1-1 background and transformed with Ubi_pro::OsNP1 construct. OsACTIN1 served as a control. Data were shown as means ± SD (n = 3). (N–P) Phenotype of WT and transgenic plant OE22-1. Plants after bolting (A, E, and I), spikelets with the palea and lemma removed (B, F, J, and N), and I₂-KI staining of the pollen grains (C, D, G, H, K, L, O, and P) are shown. (Scale bars: A, E and I, 10 cm; B, F, J, and N, 1 mm; C, D, G, H, K, L, O, and P, 100 μm.)

Fig. S3.

Protein sequence alignment of OsNP1 homologs in 15 species. Bases in red show the conserved amino acid residues.

Fig. S4.

A phylogenetic tree of OsNP1 and related proteins. The neighbor-joining tree was performed using MEGA6 based on the alignment of OsNP1 and the most similar sequences in rice and Arabidopsis. The numbers at the branches are bootstrap values provided as percent over 1,000 replications. Four groups of genes can be defined as follows: group A is represented by At1g72970, which is involved in the biosynthesis of long-chain α-,ω-dicarboxylic fatty acids; group B comprises putative mandelonitrile lyases; group C includes two proteins predicted to be aldehyde lyases; and group D comprises several putative LCFA alcohol dehydrogenases.

Fig. 2.

Expression pattern of OsNP1. (A) qPCR analysis of OsNP1 expression in different organs of WT. OsACTIN1 served as a control. Data are shown as means ± SD (n = 3). (B) GUS expression (blue staining) patterns of developmental spikelets on the OsNP1_pro::GUS transgenic line. (Scale bar, 2 mm.) (C) In situ hybridization of OsNP1 in WT anthers from stages 7–8b. Anthers hybridized with OsNP1 sense probe served as controls. PMC, pollen mother cell; T, tapetal layer; Tds, tetrads. (Scale bars, 50 μm.)

Fig. 3.

Histological features of anther development in the WT and osnp1-1. Anther sections of WT (A, C, E, G, I, and K) and osnp1-1 (B, D, F, H, J, and L) plants from developmental stages 8a to 12 are shown. BMs, binuclear microspores; DMs, degenerated microspores; E, epidermis; En, endothecium; M, middle layer; Mp, mature pollen; Ms, microspores; PMC, pollen mother cell; ST, swollen tapetal layer; T, tapetal layer; Tds, tetrads. (Scale bars, 20 μm.)

Fig. 4.

TEM analyses of the WT and osnp1-1 anthers from stages 9–12. The transverse sections of the WT (A, C, E, G, I, K, M, O, and Q) and osnp1-1 (B, D, F, H, J, L, N, P, and R) anthers at stage 9 (A–F), early stage 10 (G–L), late stage 10 (M–P), and stage 12 (Q and R) are compared. (C, D, I, and J) Higher magnification of the tapetum showing Ubisch body. (E, F, K, and L) Higher magnification showing deposition of sporopollenin in pollen exine. (O and P) Higher magnification of the tapetum and pollen exine. (Q and R) The outer anther epidermis. Ba, bacula; C, cuticle; CW, cell wall; DMs, degenerated microspores; E, epidermis; En, endothecium; Ex, exine; M, middle layer; Ms, microspores; Nu, nucleus; Ne, nexine; T, tapetum; Te, tectum; Ub, Ubisch body. (Scale bars: A, B, G, H, M, and N, 10 μm; C, D, I–L, O, and P, 1 μm; E, F, Q, and R, 500 nm.)

Fig. S5.

Observation of the anther and pollen grain in the WT and osnp1-1 by SEM. The microspores, inner surface of anther wall, and anther epidermis of WT and osnp1-1 from stages 8b to 12 are shown. (Scale bars: rows 1, 2, 5, and 6, 10 μm; rows 3 and 4, 2 μm.)

Fig. S6.

Expression of genes involved in pollen development in the WT and osnp1-1. Expression of GAMYB (A), UDT1 (B), TDR (C), PTC1 (D), OsCP1 (E), RTS (F), RAFTIN1 (G), WAD1 (H), DPW (I), CYP703A3 (J), CYP704B2 (K), OsNP1 (L), OsC4 (M), OsC6 (N), and OsABCG15 (O) at stage 7–12 in the WT and osnp1-1 was analyzed using qPCR. OsACTIN1 served as a control. Data are shown as means ± SD (n = 3).

Fig. S7.

A nuclear male sterility system in rice via a transgenic approach. (A) Schematic diagram of the nuclear male sterility system. The male sterile mutant was transformed with a transgene containing three functional modules, including the following: (i) the WT fertility gene (MS) to restore the male fertility, (ii) the α-amylase gene (AA) to devitalize transgenic pollens, and (iii) the red fluorescence protein (RFP) gene to differentiate the transgenic seeds from the nontransgenic seeds. Because the male fertility gene is a recessive sporophytic gene, a hemizygous transgene in the male sterile mutant plant can fully restore the male fertility. Whereas the α-amylase gene driven by a pollen-specific promoter disrupts starch accumulation only in the transgenic pollen, only the transgenic pollen grains produced by the hemizygous transgenic plant are defective, and the nontransgenic pollen grains are viable for pollination. The resulting transgenic maintainer plant (MS-AA-RFP; ms/ms) produces male gametes (MG) of one genotype (ms), and female gametes (FG) of two genotypes (ms and MS-AA-RFP, ms). Self-pollination of the transgenic maintainer generates the transgenic seed (MS-AA-RFP; ms/ms) and the male sterile seed (ms/ms) in 1:1 ratio, and the seeds can be sorted based on red fluorescence. The male sterile seed can also be propagated via cross-pollination of the transgenic maintainer with the male sterile plant. The male sterile plants are pollinated by the paternal line for production of hybrid seeds. (B) The pZhen18B plasmid for transformation. The osnp1-1 mutant was transformed with a double T-DNA binary vector pZhen18B. The first T-DNA contained the NPTII gene under the 35S promoter for transformation selection. The second T-DNA contained three functional modules, including the following: (i) OsNP1 under its native promoter for restoration of male fertility, (ii) the maize α-amylase gene (ZM-AA1) under PG47 promoter to devitalize the transgenic pollens, and (iii) the DsRed gene under LTP2 promoter for detection of the transgenic seeds.

Fig. 5.

Phenotype of the nuclear male sterility system. (A) A WT plant and Zhen18B plant after bolting. (Scale bar, 10 cm.) (B) Spikelets of WT and Zhen18B with the palea and lemma removed. (Scale bars, 1 mm.) (C) I₂-KI staining of the pollen grains in WT and Zhen18B. (Scale bars, 100 μm.) (D) Zhen18B panicle under bright field (BF) and a red fluorescence filter (RFP), respectively. (Scale bars, 2 cm.)

Fig. 6.

Heterosis of Zhen18A-derived F1 hybrids. Representative F1s from Zhen18A cross with various germplasms out-performed their parents in individual yield (A) and plot yield (300 plants) (B). The number above the column represents the yield increase of the F1 over the better yielding parent.