Exploration of the molecular mechanism behind a novel natural genic male-sterile mutation of 1205A in Brassica napus

Abstract

The use of a male sterility hybrid seed production system has resulted in a significant increase in rapeseed yields by over 20%. Nevertheless, the mechanisms underlying male sterility remain largely unexamined. This study presents a spontaneous recessive genic male-sterile (RGMS) mutant of 1205A, which was employed to establish two two-line hybrid production systems: 1205AB and NT7G132AB. Cytological investigations reveal that the mutation occurs at the early microspore stage, resulting in premature degradation of pollen. Through inheritance analysis, linkage mapping, and bulked-segregant analysis sequencing (BSA-Seq), a single gene locus, designated Bna1205ams1, was identified within the QTL region on chrC03 (15.36–18.90 Mb). The development of three newly co-segregated kompetitive allele-specific PCR (KASP) markers, in conjunction with two traditional co-segregated markers, allowed for the refinement of the QTL of Bna1205ams1 to a segment of 181.47 kb. This refinement facilitated the identification of a candidate gene, BnaC03g27700D, through functional and expression analyses. Furthermore, the subcellular localization of BnaC03g27700D was examined. Metabolic fluctuations associated with the fertility gene were observed, particularly in processes related to aborted tapetal programmed cell death (PCD), which may contribute to reduced pollen fertility with abnormal pollen exine. A strong correlation was also established between BnaC03g27700D and thirteen metabolites. This study not only offers valuable insights into the research and practical application of plant male sterility but also serves as a case study on the genetic regulatory mechanisms governing male sterility.

Introduction

Rapeseed (Brassica napus L., AACC, 2n = 38) is currently recognized as one of the most significant oilseed crops for vegetable oil production worldwide. The implementation of hybrid breeding strategies has resulted in an increase in the yield of B. napus by over 20% [1]. In China, hybrid seeds constitute more than 70% of the total cultivated area [2]. Given the crop's importance, further research and the development of additional hybrid production systems are imperative. Presently, two primary systems for controlling pollination—cytoplasmic male sterility (CMS) and genic male sterility (GMS)—are employed in B. napus for the creation of hybrid varieties. Among these, CMS is the most widely utilized breeding method in hybrid production, as exemplified by the Ogura system [3,4,5] and the Polima system [2]. Conversely, GMS has garnered significant interest among rapeseed breeders owing to its pure and stable male sterility, which remains unaffected by cytoplasmic factors. Furthermore, GMS enables a more efficient breeding process, facilitating a transition from a three-line system to a two-line system in comparison to CMS. It is noteworthy that GMS lines have been documented in over 610 species of flowering plants [6, 7] including B. napus [8,9,10], Brassica oleracea [11, 12], Gossypium hirsutum [13, 14], Cryptomeria japonica [15], Cucumis sativus [16], and Capsicum annuum [17], as well as many other crop species.

GMS is characterized by the inability to produce functional anthers and pollen grains at various stages of anther microsporogenesis and dehiscence, which are processes that follow cell division, differentiation, and subsequent degeneration [18,19,20,21,22]. The abnormal anthers are programmed to develop into a four-lobed structure composed of highly specialized reproductive tissues necessary for pollen production, as well as non-reproductive tissues, including the epidermis, endothecium, middle layer and tapetum, which are essential for normal pollen development and the release of pollen grains [18, 23]. A mutation in a single nuclear gene, typically governed by a recessive allele, can lead to the manifestation of GMS [24]. Numerous single-gene mutations have been documented in the model organism Arabidopsis thaliana. For instance, mutations in the genes DYSFUNCTIONAL TAPETUM1 (DYT1), Defective in Tapetal Development and Function 1 (TDF1), ABORTED MICROSPORES (AMS), and AtMYB103/MS188 have been shown to result in male sterility [25]. These genes primarily regulate the tapetum, which is the principal tissue that provides precursors for pollen development and pollen wall formation. Furthermore, the entire process of pollen development and pollination is crucial for the male fertility of flowering plants. This process can be modulated by CALLOSE SYNTHASE-LIKE genes, GLUCAN SYNTHASE-LIKE genes, ARABINOGALACTAN PROTEIN 11 (AGP11), REVERSIBLY GLYCOSYLATED POLYPEPTIDE 1 and 2 (RGP1 and 2), and CYSTEINE ENDOPEPTIDASE 1 (CEP1) [25,26,27,28,29,30,31,32,33,34,35]. Mutations in these essential genes could also lead to the development of a male sterility phenotype.

The majority of reported male-sterile mutants have arisen spontaneously and consistently exhibit a Mendelian pattern of inheritance [24]. This conclusion is substantiated by a review of 25 male sterility systems, which encompasses 20 from Brassica oleracea and five from Brassica rapa [36]. Among these twenty-five mutations in Brassicaceae vegetables, twenty-two (88%) were classified as spontaneous mutations. Seventeen (68%) of these were categorized as monogenic recessive, while one mutation (4%) involved a single locus with three alleles [36]. These findings suggest that most spontaneous or induced male-sterile mutants are governed by monogenic recessive genes, although instances of control by multiple alleles may occur [6, 36, 37]. In B. napus, several GMS systems have been discovered and studied in detail, each regulated by different genes. A spontaneous mutant was identified from a distant hybridization between B. napus and B. rapa, which facilitated the identification and mapping of one gene locus, Bnmfs, through whole-genome resequencing and RNA sequencing techniques [38]. The most widely used breeding system among GMS systems reportedly involves a recessive nuclear gene (RGMS) that has mutated from a fertility-dominant gene, such as 117A [39] and S45A [40]. The sterility associated with S45A is attributed to a mutation of BnCYP704B1, which possesses two duplicates of BnMs1 and BnMs2, both of which are essential for pollen exine formation and tapetal development in B. napus [40]. Furthermore, several studies have identified multiple genes that regulate sterility. In the case of 9012A, sterility is controlled by two recessive male-sterile genes, ms3 and ms4, in conjunction with a recessive epistatic suppressor gene (esp) [41, 42]. Additionally, the sterility of RG206A is also regulated by two male sterility alleles (BnRf and BnRf^b) alongside the restorer allele BnRf^a [43].

Although various RGMS lines are employed in the hybrid production of B. napus, the mechanisms underlying male sterility remain largely unexamined. Additionally, the identification of 50% male-fertile plants during the maintenance of the msms line through crossing with Msms (where Ms denotes the dominant fertility gene and ms signifies the recessive gene) poses a significant challenge for the application of this system in commercial hybrid seed production. Consequently, it is essential to identify additional RGMS lines and investigate their regulatory mechanisms to mitigate these limitations. The advent of the rapeseed reference map [44] has enabled the development of innovative novel methods based on sequencing, complementing traditional genetic research such as quantitative trait locus (QTL) mapping. The sequencing-based bulked segregant analysis (BSA-Seq) strategy provides a more efficient approach to QTL mapping approach compared to conventional QTL mapping methods. By integrating the traditional BSA method with next-generation sequencing (NGS) technology, the BSA-Seq technique has gained widespread application in QTL mapping across various crops, including rapeseed [45, 46], rice [47], corn [48] and so on. Furthermore, sequence analysis of the QTL region utilizing BSA-Seq can reveal numerous single nucleotide polymorphisms (SNPs) as well as small insertions and deletions (InDels). Specific markers can be designed and utilized for the fine mapping of candidate genes within primary QTL regions, facilitating the identification of candidate genes [49].

In the current study, we explored the mechanisms underlying a natural GMS line, 1205A, utilizing a combination of cytological identification, genetic analysis, traditional QTL mapping, BSA-Seq analysis, fine mapping and quantitative reverse transcription polymerase chain reaction (qRT-PCR) analysis. Additionally, we investigated the metabolic variations induced by the male sterility gene. This research may enhance the annotation of the candidate gene and offers new resources for fundamental investigations into the genetic regulation of male sterility.

Materials and Methods

Plant materials and field experiment

In 2010, a spontaneous recessive male sterility mutant was identified within the breeding line 1205, was identified, subsequently the sterility RGMS line was designated as 1205A. Fertile plants from line 1205 were crossed with plants from 1205A, resulting in progeny lines that exhibited a segregation ratio of 1:1 between fertile and sterile plants. This led to the establishment of a two-line hybrid production system, referred to as 1205AB, developed by the Oil Crops Research Group at Guizhou University in Guiyang, Guizhou, China [50, 51]. Within this system, each generation consistently displayed a 1:1 segregation ratio of fertile to sterile plants. Consequently, the 50% of sterile plants were designated as the 1205A population, while the remaining 50% fertile plants were designated as the 1205B population. Annually, 1205A-type plants were crossed with 1205B-type plants within the 1205AB line, and the harvested seeds were subsequently planted in the following year. This seed production process has been continuously repeated to the present day. In the years 2016, 2017, and 2018, three populations resulting from the cross between 1205A and 1205B were cultivated and evaluated for plant sterility, yielding 730 plants (designated as 1205AB-2016), 830 plants (1205AB-2017), and 1,424 plants (1205AB-2018), respectively. Moreover, in 2020, the sterile gene, designated as Bna1205ams1, was successfully transferred from the line 1205A into a new-type B. napus line, NT7G132, which was resynthesized from Brassica carinata and B. rapa, as documented in our previous studies [52, 53]. Following this transfer, a new two-line hybrid production system, NT7G132AB, was established, as illustrated in Fig. 1. The nomenclature for NT7G132AB, NT7G132A, and NT7G132B is consistent with that of the above referenced 1205AB. Furthermore, since the NT7G132AB incorporates the male sterility gene of Bna1205ams1 from 1205A, these lines were designated as NT7G132AB^Bna1205ams1, NT7G132A^Bna1205ams1, and NT7G132B^Bna1205ams1, indicate the source of the infertility genes. In the F₃ generation of NT7G132AB^Bna1205ams1, seven hybrid combinations demonstrated a precise 1:1 segregation ratio, with a total of 173 plants (NT7G132AB-F₃ population) were further used for sib-mating between NT7G132A^Bna1205ams1 and NT7G132B^Bna1205ams1. Additionally, a total of 1,847 F₄ plants (NT7G132AB-F₄ population) resulting from the cross NT7G132A^Bna1205ams × NT7G132B^Bna1205ams1 (F₃ generation) were cultivated and evaluated for sterility and fine mapping. The planting occurred at the Teaching and Practice Base of Guizhou University (106.683651°N, 26.410435°W) in Guiyang, China, with each row comprising 30 plants and a plot size of 4 m × 0.4 m.

Fig. 1

Schematic and workflow for the development of the two-line hybrid seed production system of NT7G132AB^Bna1205ams1. The letter “A” in brackets indicates the fertile allele of Bna1205AMS1, compared to the recessive allele of Bna1205ams1 by the letter of “a” in parentheses. Genotypes “AA” and “Aa” were fertile and the genotype “aa” was sterile

Paraffin section analysis

During the flowering period of 2023, flower buds of four distinct sizes were sampled from NT7G132A^Bna1205ams1 and NT7G132B^Bna1205ams1. The sampled flower bud sizes were classified into the following categories: X (X representing bud size) < 1 mm (Group S), 1 ≤ X < 2 mm (Group SM), 2 ≤ X < 3 mm (Group M), and 3 ≤ X < 4 mm (Group L), in accordance with previously published studies [43, 54]. A total of fifty flower buds were collected for each size category, immediately frozen in liquid nitrogen, and subsequently stored at −80℃ for future transcriptome and metabolome analyses. Paraffin section analysis was performed following established protocols [55] with minor modifications. Fertile and sterile flower buds from the four size groups were dissected promptly after sampling. The dissected tissue samples were fixed in FAA fixative within a chemical fume hood for a duration exceeding 24 h. Following fixation, the tissues underwent a series of dehydration and cleaning steps utilizing ethanol and xylene in varying volume ratios (3:1, 1:1, and 1:3). After incubation in molten paraffin, the tissue samples were extracted from the mold and embedded in a paraffin wax solution at 60 °C, allowing it to solidify at room temperature. Thin paraffin Sects. (10 μm) were prepared using an automated microtome, stained with aniline blue, and examined under a microscope. Images of the anthers at various developmental stages were captured using a camera.

Pollen fertility testing

Pollen fertility was evaluated in fertile plants utilizing an acetocarmine staining technique [52]. Two freshly opened flowers from each plant were collected, and the pollen grains were stained with a 1% acetocarmine solution. The stained pollen grains were then examined under an OLYMPUS CKX41 microscope (Olympus Corporation, Japan). Pollen grains that exhibited dark red, big, and round were classified as fertile, whereas those that appeared dark yellow, small, and shriveled were classified as sterile. For each plant, the number of fertile and sterile pollen grains was counted in two fields of view. For each plant, the counts of fertile and sterile pollen grains were recorded in two fields of view. The ratio of fertile pollen grains to the total number of pollen grains assessed was documented as the pollen fertility ratio for each plant evaluated.

Regional linkage mapping

A total of 1,272 Intron Length Polymorphism (ILP) primers, as reported in our previous studies [56], were employed to screen the sterile bulk (A-bulk) of the 1205A and the fertile bulk (B-bulk) of the 1205B for polymorphic markers. The A-bulk was generated by combining equal quantities of DNA from 10 sterile plants of 1205A line, while the B-bulk composed of DNA from 10 fertile plants from the lines of the 1205B line. Polymerase chain reaction (PCR) for the ILP markers was carried out in accordance with the protocols established in our earlier research [57]. All PCR products were analyzed via electrophoresis in 2% agarose gels in 1 × tri-acetate-ethylene diaminetetra acetic acid (TAE) buffer. The gels were visualized by staining with GelRed Nucleic acid dye (in water, 10,000 ×) (Tsingke Biotechnology Co., Ltd., Beijing, China) and photographed on a digital gel documentation system. Then, the polymorphic primers identified between 1205A and 1205B, as well as between NT7G132A^Bna1205ams1 and NT7G132B^Bna1205ams1 were utilized to genotype the individual plants of 1205AB (1205AB-2018) and NT7G132AB^Bna1205ams1 (NT7G132AB-F₄) populations, respectively. Regional linkage maps of the Bna1205ams1 gene were constructed based on the genotyping data from the 1205AB population (Table S1) and the NT7G132AB population (Table S2) using JoinMap 4.0 [58].

DNA library construction and BSA-Seq analysis

Total genomic DNA was extracted from the young leaves of 30 randomly selected fertile and 30 randomly selected sterile plants of the 1205AB population in 2018 (1205AB-2018), utilizing the cetyltrimethylammonium bromide (CTAB) method [59] with minor modifications. Equal proportions of DNA from the 30 fertile plants and 30 sterile plants were mixed and processed into a fertile pool (FP) and a sterile pool (SP), respectively. The sequencing library was constructed from 5 μg of DNA from each of the FP and SP samples, which were subsequently sequenced in 76 cycles on an Illumina Novaseq 6000 platform using the PE150 sequencing model, provided by LC-BIO TECHNOLOGIES CO., LTD. (www.lc-bio.com, Hangzhou, China). Adaptor reads, reads containing more than 5% undetermined nucleotides, and low-quality nucleotides (Q ≤ 10) that comprised over 20% of the total were trimmed. The high-quality clean reads were then aligned to the B. napus reference genome of Darmor-bzh” (https://www.genoscope.cns.fr/brassicanapus/) [60] using the Burrows–Wheeler Aligner software (BWA; version 0.7.13) [61]. SNPs and InDels were identified across the FP and SP by using GATK software version 3.8.1 [62]. Subsequently, the Euclidean distance (ED) method was employed to analyze the candidate regions between the FP and SP. The parameters for the ED calculation were determined as follows:

$$\text{ED}=\sqrt{{({\text{A}}_{\text{FP}}-{\text{A}}_{\text{SP}})}^{2}+{({\text{T}}_{\text{FP}}-{\text{T}}_{\text{SP}})}^{2}+{({\text{C}}_{\text{FP}}-{\text{C}}_{\text{SP}})}^{2}+{({\text{G}}_{\text{FP}}-{\text{G}}_{\text{SP}})}^{2}}$$

In the aforementioned equation, A, T, C, and G represent the four nucleotide bases. A sliding window analysis was employed to illustrate the distribution of ED values across the 19 chromosomes of B. napus, utilizing widows of 1 megabase (Mb) in width and advancing by1 kilobase (kb) at each step. This analysis was conducted using an in-house-developed Python script (available on request) based on the developed program [63]. A higher ED value indicates a greater likelihood that the SNPs and InDels are associated with the male sterility trait or are linked to a gene that regulates this trait.

Development and validation of KASP markers

BSA-Seq methodology enabled the identification of multiple SNPs and InDels between the female parent (FP) and male parent (SP) within the quantitative trait locus (QTL) region located on chromosome C03. The genetic variations identified were subsequently employed in the development of Kompetitive Allele Specific PCR (KASP) markers, utilizing the well-established reference genome of B. napus (https://www.genoscope.cns.fr/brassicanapus/) [60]. On the QTL region on C03 and C04, a total of twenty-one and ten evenly distributed markers were developed, respectively. The design of these markers was facilitated by the online tool BatchPrimer3 v1.0 (https://probes.pw.usda.gov/batchprimer3/), with specific parameters that included a minimum primer length of 15 base pairs, a maximum length of 30 base pairs, and an annealing temperature range of 59℃ to 61℃. Each KASP marker comprised two allele-specific forward primers (F₁ and F₂) and a common reverse primer (R), which were designed based on the flanking sequences surrounding the variant position. The two allele-specific forward primers shared the identical sequences, differing solely at the 3' end to facilitate nucleotide discrimination of the specific SNP. Furthermore, the fluorescent tag sequences FAM (5'-GAAGGTGACCAAGTTCATGCT-3') and VIC (5'- GAAGGTCGGAGTCAACGGATT-3') were incorporated at the 5' end of F₁ and F₂, respectively. The synthesis of these markers were conducted by Tsingke Biotech Co., Ltd (Beijing, China). Standard genotyping procedures for KASP assays were followed [64], and the fluorescence values were recorded using a Bio-Rad CFX96 Real-Time System (Bio-Rad, Hercules, CA, USA). All 21 markers were utilized to amplify DNA samples from 30 fertile and 30 sterile plants of the BAS-Seq population 1205AB. Subsequent analysis of the amplified sequences revealed that only three markers (ID07, ID10, and ID21) on chromosome C03 exhibited potential co-segregation with fertility in the 1205AB population (Table S1). These three markers were then employed to amplify DNA samples from the NT7G132A^Bna1205ams × NT7G132B^Bna1205ams1 plants in 2023 for further fine mapping studies.

RNA extraction and qRT-PCR

During the flowering period of 2023, flower buds of four distinct sizes were collected from the lines NT7G132A^Bna1205ams1 and NT7G132B^Bna1205ams1. Total RNA (2 µg) was extracted from various stages of groups S, SM, M and L, utilizing three independent plants from each of the 1205A and 1205B lines, following the manufacturer's instructions provided with the TRIzol kit (Invitrogen, Carlsbad, CA). The purity of the RNA was evaluated using a Kaiao K5500® Spectrophotometer (Kaiao, Beijing, China), while the RNA integrity and concentration of the RNA were assessed using the RNA Nano 6000 Assay Kit for the Bioanalyzer 2100 system (Agilent Technologies, CA, USA). Subsequently, the cDNA was synthesized through reverse transcription employing the PrimeScript RT Reagent Kit with gDNA Eraser (Takara, Dalian, China), in accordance with the manufacturer’s instructions. Twenty-two candidate gene-specific primers for qRT-PCR were designed based on reference unigene sequences using Primer Premier 5.0. Real-time PCR was conducted using SsoAdvanced™ Universal SYBRR Green Supermix (Hercules, CA), following the methodology established in our previous research (Khattak et al. 2019). The 2^−ΔΔCt method was employed to calculate the relative levels of gene expression, with the β-actin gene serving as an internal control. All qRT-PCR experiments were conducted with three biological replicates and were executed on a Bio-Rad CFX96 Real Time System (Bio-Rad, Hercules, CA, USA).

Subcellular localization

To construct plasmids for subcellular localization, the full-length ORFs of BnaC03g27700D were amplified using specific primers (Table S1) and subsequently cloned into 16,318-hGFP vector (Biovector, Beijing, China) to generate 16,183-BnaC03g27700D-GFP. Mesophyll protoplasts from four-weeks-old Arabidopsis thaliana (ecotype Columbia) plants were prepared from their rosette leaves and transiently transformed according to a previously published protocol [65]. The transformed protoplasts were incubated in the dark for 16 h at 20℃. The subcellular distribution of BnaC03g27700D, tagged with GFP and expressed in the Arabidopsis mesophyll protoplasts, was visualized using a confocal laser scanning microscope equipped with the LSM 510 system (Zeiss, Oberkochen, Germany) at excitation wavelengths of 405 nm and 488 nm.

Construction of the protein–protein interaction network

The protein–protein interaction (PPI) network of the candidate gene in relation to other genes was constructed utilizing STRING (version 12.0) [66]. Following this, hub genes were systematically identified based on their degree centrality using Cytoscape_v3.9.1 [67]. In this network representation, each node corresponds to a gene, with the size of the nodes reflecting the number of connections to other genes. Notably, genes that exhibit a greater number of connections are represented with larger node sizes, signifying their enhanced importance in interacting with a wider array of genes.

Differential metabolites (DMs) analysis

A total of six flower bud samples from four distinct groups from each of NT7G132A^Bna1205ams1 and NT7G132B^Bna1205ams1 were utilized for untargeted metabolomic analysis. The analysis was conducted using a Liquid Chromatography Triple Quadrupole Mass Spectrometer (LC–MS/MS) at APTBIO in Shanghai, China. To extract metabolites from the samples, 100 mg of each sample was combined with 400 μL of a cold extraction solvent composed of methanol, acetonitrile, H₂O in a ratio of 2:2:1 (v/v/v) and thoroughly vortexed. The mixture was then allowed to rest on ice for 20 min before being centrifuged at 4℃ and 14,000 × g for an additional 20 min. The supernatant obtained was collected and subsequently dried using a vacuum centrifuge at 4 °C. For LC–MS analysis, the dried samples were reconstituted in 100 μL of a solvent mixture of acetonitrile and water in a 1:1 (v/v) ratio. The extracts were separated using Agilent 1290 Infinity LC ultra-high-performance liquid chromatography (UHPLC) system. Primary and secondary spectra were acquired using a quadrupole time-of-flight mass spectrometer (Sciex TripleTOF 6600). The raw data were converted into MzXML files using ProteoWizard. The data were then imported into the XCMS software (Scripps Research Institute, La Jolla, USA) for peak picking, peak grouping, and extraction of ion features. Metabolite identification via MS/MS was performed using an in-house database established with authentic standards. Principal Component Analysis (PCA) was employed to discern the characteristics of metabolite variances between the two groups. An unpaired Student’s t-test was conducted to determine statistical significance, with a variable influence on projection (VIP) threshold of > 1 and a p-value threshold of p < 0.05 established to identify significantly different metabolites. Additionally, Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of the distinct metabolites was performed using MetaboAnalyst (https://www.metaboanalyst.ca/) [68]. The Mantel test assessing the correlation between gene expression levels and metabolic abundance across the four stages (S, SM, M and L) was calculated using the R package ggcor, while the visual representation of the correlation heatmap was generated using chiplot [69].

Results

Male sterility initiates during the early stages of flower bud development

In our laboratory, a spontaneous genic male sterility mutation was employed to establish a two-hybrid production system, designated as 1205AB, in B. napus [50, 51]. The mutated genic male sterility gene was designated as Bna1205ams1 and was subsequently incorporated into a novel B. napus line, NT7G132, which was derived from a cross between B. carinata and B. rapa [52, 53]. This process resulted in the creation of a new two-hybrid production system, NT7G132AB^Bna1205ams1 (Fig. 1). To elucidate the morphological changes induced by the mutations and their specific onset periods, we observed the development of anthers and pollens across four stages (S, SM, M, and L) characterized by varying flower bud sizes within each line of NT7G132A^Bna1205ams1 and NT7G132B^Bna1205ams1.

During the tetrad stage in group S, meiocytes undergo meiosis, resulting in the formation of tetrads composed of newly generated haploid microspores, which are subsequently slightly separated and encased by a callose wall deposited on the primexine of the microspores. A comparative analysis of the morphology of NT7G132B^Bna1205ams1 (Fig. 2A) indicated no significant alterations in the vascular region, tetrads, middle layer, epidermis, and endothecium of the anther in NT7G132A^Bna1205ams1 (Fig. 2B). Advancing to the early microspore stage, free haploid microspores are released from the tetrads as the callose wall is degraded by callase secreted from tapetal cells in group SM of NT7G132B^Bna1205ams1 (Fig. 2C). Concurrently, in NT7G132A^Bna1205ams1, the appearance of anthers begin to exhibit deformation at this stage, while other characteristics of the cell wall remained consistent with those observed in NT7G132B^Bna1205ams1 (Fig. 2D). During the early anther development of group M in NT7G132B^Bna1205ams1, the released microspores display a spherical morphology with thin exines and an increased vacuole volume, resulting in a round-shaped microspore (Fig. 2E). Notably, a significant morphological change occurs at this stage in the anther of NT7G132A^Bna1205ams1 (Fig. 2F), where the anther continues to shrink and collapse, indicating a progressive degradation of the endothecium and tapetum. In group L of NT7G132B^Bna1205ams1, the development and maturation of pollen advance gradually, with spherical microspores that are rich in reserve substances (Fig. 2G). In contrast, in the sterile anther of NT7G132A^Bna1205ams1, the anther has undergone complete degeneration (Fig. 2H).

Fig. 2

Transverse sections of four sets of S, SM, M and L of NT7G132B^Bna1205ams1 (A, C, E and G) and NT7G132A^Bna1205ams1 (B, D, F and H), respectively. Representative anther pictures of buds of different sizes [S (X < 1 mm), SM (1 ≤ X < 2 mm), M (2 ≤ X < 3 mm) and L (3 ≤ X < 4 mm)], respectively. A/B, C/D, E/F, and G/H represent the anther and pollen development stages of tetrad stage (S), early microspore stage (SM), vacuolated microspore stage (M), and mature pollen stage (L) in NT7G132B^Bna1205ams1 and NT7G132A^Bna1205ams1, respectively. V: vascular region, Tds: tetrads, ML: middle layer, E: epidermis, En: endothecium, T: tapetum, MP: mature pollen

Male sterility is determined by a single gene locus

The segregation ratio of fertile to sterile plants was analyzed across three distinct populations: 1205AB-2016, which consisted of 730 plants, which includes 730 plants; 1205AB-2017, comprising 830 plants; and 1205AB-2019, which included 830 plants. All three populations were all derived from the cross 1205A × 1205B. The observed numbers of fertile and sterile plants in these populations was 356 and 374 in 1205AB-2016, 402 and 428 in 1205AB-2017, and 705 and 719 in 1205AB-2018, respectively, fitting a segregation ratio of 1:1 (χ2 = 0.444, p = 0.505; χ2 = 0.138, p = 0.711; χ2 = 0.814, p = 0.367, respectively) (Table S3). This data suggests that the male sterility of 1205A is controlled by a gene locus in these 1205A × 1205B-derived populations. Additionally, the segregation ratio of fertile and sterile plants was examined in two populations from the cross NT7G132A^Bna1205ams × NT7G132B^Bna1205ams1. In the NT7G132AB-F₃ population, the counts of fertile and sterile plants were 91 and 82, respectively, resulting in a segregation ratio of 1:1 (χ2 = 0.468, p = 0.494) (Table S3). In the NT7G132AB-F₄ population, the numbers were 956 and 891, respectively, again indicating a segregation ratio of 1:1 (χ2 = 0.468, p = 0.494; χ2 = 2.287, p = 0.130) (Table S3). This further corroborates that male sterility of 1205A is regulated by a gene locus of Bna1205ams1.

Moreover, to eliminate the possibility of partial fertility, 30 fertile plants from each genotype, 1205B and NT7G132B^Bna1205ams1, were randomly selected for pollen fertility assessment (Fig. S1; Table S4). The results revealed that the fertile plants in 1205B exhibited an average fertility ratio of 99.88%, while those in NT7G132B^Bna1205ams1 demonstrated an average fertility ratio of 99.90%, indicating that both genotypes were completely fertile (Table S4).

Male sterility gene locus was mapped by linkage mapping

The populations 1205AB and NT7G132AB^Bna1205ams1 were utilized for regional linkage mapping of the genic male sterility gene Bna1205ams1. Initially, a total of 1,272 ILP primers were employed to screen both the sterile bulk (A-bulk) and fertile bulk (B-bulk), leading to the identification of 23 polymorphic markers that facilitated the mapping of the genic male sterility gene Bna1205ams1. Subsequently, these 23 polymorphic markers were utilized to ascertain the genotypes of 144 randomly selected plants (comprising 72 sterile and 72 fertile individuals) from the 1205AB-2018 population. Ultimately, 11 ILP markers (Table S5) were mapped onto a single linkage group with a length of 10.6 cM, wherein two markers of TAT244050 (Fig. S2) and TAT126180, exhibited co-segregation with the genic male sterility gene Bna1205ams1 (Fig. 3). These 11 ILP markers were subsequently employed to determine the genotypes of 186 randomly selected plants (comprising 84 sterile and 84 fertile individuals) from the NT7G132AB-F₄ population, resulting in identification of a linkage group with a length of 13.3 cM, which also demonstrated co-segregation with the two markers TAT244050 and TAT126180. Most markers between the two linkage groups exhibited strong collinearity with the results obtained from the 1205AB-2018 population. The products from these 11 ILP markers were sequenced and compared with the B. napus reference map of Darmor.v4.1 [70], ultimately positioning the two linkage groups on the chromosome C03 of B. napus.

Fig. 3

Bna1205ams1 were mapped on chrC03 by linkage mapping in 1205AB and NT7G132AB.^Bna1205ams1. The grey dash lines between markers of two linkage groups indicates the same marker. The primer sequences can be found in Table S1

Male sterility gene locus were mapped by BAS-Seq

To validate the linkage mapping results and fine-map the gene locus of Bna1205ams1, the BSA-Seq strategy method was employed. A total of thirty randomly selected plants from each of 1205A and 1205B were employed for DNA extraction. Then equal quantities of DNA from these plants were combined to form a sterile pool (SP) and a fertile pool (FP) for the purpose of whole-genome resequencing. After filtering the raw data, 32.45 Gb of clean data were obtained for the FP, demonstrating a Q30 percentage of 94.36%, while 36.28 Gb clean data were acquired for the SP, with a Q30 percentage of 94.10%. Among these clean reads, a total of 210,153,972 reads (97.13%) for the FP and 235,697,816 reads (97.46%) for the SP were aligned to the reference genome of B. napus [60]. The sequencing depth of coverage exceeding 20 × was 43.84% for the FP and 62.43% for the SP (Fig. S3). A total of 6,706,272 SNPs were identified between the FP and SP, which included 54.84% classified as “intergenic_region”, 11.62% as “upstream_gene_variant”, 11.45% as “downstream_gene_variant” (Table S6). Additionally, a total of 1,155,082 InDels were detected, comprising 49.48% of the “intergenic_region” type, 15.65% of the “upstream_gene_variant” type, 13.87% of the “intron_variant” type, and 13.90% of the “downstream_gene_variant” type between the FP and SP (Table S7).

The SNP/InDel index between the FP and SP was computed using the Euclidean distance algorithm (ED) and subsequently employed for QTL mapping (Fig. 4). Two potential QTL regions were identified on chromosomes chrC03 and chrC04 associated with the male sterility gene locus of Bna1205ams1. The first QTL located on chrC03 encompassed a genomic region of 3.54 Mb (ranging from 15.36 to 18.90 Mb) (Fig. 4), which included 31,751 SNPs, 4,684 InDels, and 416 putative genes in comparison to the B. napus reference genome. The second QTL identified on chromosome chrC04 spanned 2.85 Mb (ranging from 2.43 to 5.28 Mb) (Fig. 4), containing 32,117 SNPs, 4,979 InDels, and 575 candidate genes relative to the B. napus reference genome.

Fig. 4

QTL mapping results by BSA-Seq method. The dashed red line indicates the threshold for QTL detection. chrC01 to chrC9 and chrA01 to  chrA10 on the x-axis indicate the 19 chromosomes of B. napus genome. The number on the y-axis indicates the values calculated by the Euclidean Distance algorithm. The colored boxes on chrC03 and chrC04 indicate the two QTLs identified by the BSA-Seq method

The locus of the male sterility gene was validated and finely mapped

To fine-map the Bna1205ams1 locus, we developed a set of twenty-one and ten KASP markers, which were evenly distributed across chromosomes chrC03 and chrC04, respectively, utilizing SNP and InDel variants. The efficacy of the 31 SNP markers was validated through the amplification of genomic DNA extracted from 30 fertile and 30 sterile plants that had previously been utilized in the BSA-Seq analysis. Notably, only three markers (ID07, ID10, and ID21) (Table S5) located on chromosome chrC03 exhibited co-segregation with the male sterility trait within the 1205AB population. These findings further corroborate that the Bna1205ams1 gene is situated within the QTL region on chromosome chrC03 (3.54 Mb, 15.36–18.90 Mb) (Fig. 5A). Additionally, the two co-segregating ILP markers, TAT244050 and TAT126180, along with three newly developed KASP markers, were utilized to genotype 571 randomly selected plants from NT7G132AB-F₄ population. Sixteen recombinant plants exhibiting six distinct genotypes were identified, specifically labeled as Sabc (three plants), Sabcd (two plants), Fab (two plants), Fde (four plants), Fcde (one plant), and Sbcde (four plants) (Table S8; Fig. 5B; Fig. S4). In this context, the letters S and F denote sterile (S) and fertile (F) plants, respectively, while the letters a, b, c, d, and e represent the recombination sites of the five markers: TAT244050 (a), ID07 (b), ID10 (c), ID21 (d), and TAT126180 (e) (Table S8; Fig. 5B; Fig. S4). Ultimately, the QTL region encompassing Bna1205ams1 on chromosome C03 was refined to a 181.47 kb segment defined by the recombination genotypes Fab and Fcde (Fig. 5C). A comprehensive analysis of all genes within this refined region revealed the presence of 22 genes when compared to the reference genome of B. napus (Fig. 5D). Furthermore, the sequence of the QTL region containing Bna1205ams1 was compared with the Arabidopsis genome, revealing a high degree of sequence similarity to a fragment located on chromosome 5, specially between At5g4940 and At5g50320 of A. thaliana (Fig. 5D and E). Additionally, the sequences of the 22 genes within the fine mapping region of chrC03 were subjected to a BLAST search against the Arabidopsis genome in the TAIR database (https://www.arabidopsis.org/index.jsp), resulting in the identification of thirteen Arabidopsis alleles (Fig. 5D and 5E).

Fig. 5

Fine mapping of the male sterility gene. A indicate the QTL mapping results obtained through BSA-Seq. B and C present the findings from the fine mapping analysis, with the numbers under C03 representing the count of recombinants at the corresponding sites/markers. Among the five markers utilized for fine mapping, TAT244050 and TAT126180 co-segregated with the male sterility gene, while three KASP markers: ID07, ID10, and ID21, were newly developed in this study (B). C exhibits the recombinant plants of six types (Sabc, Sabcd, Fab, Fde, Fcde, and Sbcde). The names these recombinant plants follow a specific nomenclature: S and F indicate sterile and fertile plants, respectively, while a, b, c, d, and e denote the recombination sites of the five markers: TAT244050, ID07, ID10, ID21, and TAT126180. For example, “Sabcd” indicates that the recombinant plant is sterile (S) and recombinated at the sites/markers: TAT244050 (a), ID07 (b), ID10 (c), ID21 (d). The number “571” represents the total number of plants used for traditional fine mapping in (C). The number at the end of each recombination type indicates the count of recombinant plants. The twenty-two candidate genes based on the B. napus reference genome of “Darmor-bzh” in the fine-mapped QTL region were displayed in the 181.47 kb region under the chromosome (D). The sequences within the fine-mapped QTL region were compared with the Arabidopsis genome, and the corresponding alleles in Arabidopsis genome were marked under chromosome region (E)

The candidate genes for male sterility were identified

To further investigate the candidate genes associated with Bna1205ams1, we conducted a qRT-PCR analysis to assess the expression levels of 22 genes located within the fine mapping region of chrC03 at four distinct stages of S, SM, M, and L of flower buds. A comprehensive list of all primers utilized in this study is provided in Table S3. Among the analyzed genes, only three genes—BnaC03g27600D, BnaC03g27670D, and BnaC03g27700D—demonstrated significant differences in expression levels at least at one developmental stage (Fig. 6). Specifically, the expression of BnaC03g27600D exhibited a significant difference at the L stage (p = 0.01) (Fig. 6A). BnaC03g27670D showed significant expression differences at the SM (p = 0.05) and L stages (p = 0.001) (Fig. 6B). Furthermore, the expression level of BnaC03g27700D was significantly elevated in the SM, M, and L stages of the 1205A genotype compared to the 1205B genotype (p = 0.001) (Fig. 6C). The considerable variation in the expression of BnaC03g27700D across different stages was further illustrated through a heatmap (Fig. 6D). Notably, the expression levels of BnaC03g27700D in other tissues, including stems, leaves, and roots, were found to be negligible (data not shown), indicating its specific functional role in the flower buds of B. napus.

Fig. 6

qRT-PCR analysis of the candidate genes in the fine-mapped QTL region. The genes with no difference in any of the four stages of S, SM, M and L between 1205A and 1205B were not shown. A, B, and C indicate the expression level of BnaC03g27600D, BnaC03g27670D, and BnaC03g27700D, respectively. The expression level of BnaC03g27700D was specifically analyzed using a heatmap, R₁, R₂ and R₃ indicate three replicates. *, ** and *** indicate the significant difference at levels of p = 0.05, p = 0.01 and p = 0.001

The functions of three candidate genes were analyzed in this study. At5g50020, which corresponds to Protein S-acyltransferase 9 (PAT9), has been associated with the plant immune response as an Arabidopsis allele of BnaC03g27600D. Mutations in this gene may enhance the plant immune response [71]. At5g50180, corresponding to BnaC03g27670D, is predicted to belong to the mitogen-activated protein kinase (MAPK) family and clusters with ATN1, a predicted signal transduction module [72]. However, the specific function of At5g50180 remains undetermined according to the current literature. Additionally, the Arabidopsis allele of BnaC03g27700D (At5g50260, CEP1, CYSTEINE ENDOPEPTIDASE 1) has been identified as a crucial factor in tapetal programmed cell death and the regulation of pollen development in Arabidopsis [26]. Consequently, BnaC03g27700D exhibits significant potential as a candidate gene for Bna1205ams1.

Subcellular localization and protein–protein interaction (PPI) network analysis

To improve the analysis of the subcellular localization of BnaC03g27700D, the gene was cloned into a vector and subsequently introduced into A. thaliana. Laser confocal microscopy analysis indicates that the BnaC03g27700D protein is predominantly localized within the nucleus, exhibiting a patchy fluorescence signal (Fig. 7). Furthermore, a faint protein localization signal for the protein is detected on the membrane, suggesting that the BnaC03g27700D protein is situated on the endoplasmic reticulum (Fig. 7).

Fig. 7

The subcellular localization of BnaC03g27700D. The fluorescence co-localization of p35S-BnaC03g27700D-GFP fusion proteins and DAPI indicated the subcellular localization of BnaC03g27700D. p35S-GFP was used as a positive control, and the fluorescence was extensively detected in both the nucleus and cytoplasm. Bar = 10 um

In addition, a protein–protein interaction (PPI) network for the candidate gene BnaC03g27700D was constructed utilizing STRING (version 12.0) (Fig. 8A) [66]. Among the 17 associated genes identified, five have been recognized as potentially significant in the genetic regulation of male sterility (Table S9). These genes include BnaC03g29860D (APA3), which is linked to developmental programmed cell death; BnaC05g04310D (F9P14.12), which is involved in cysteine peptidase activity within floral structures; BnaC07g33540D (RD19C), which plays a role in anther development post-maturation through the action of γVPE; as well as BnaC03g49600D (IDL2) and BnaC03g40640D (IDL4), both of which are associated with floral organ abscission. The remaining 13 genes have not yet been documented as participants in the regulation of male sterility; however, they may represent potential candidates for future investigations, as they interact with the candidate gene BnaC03g27700D.

Fig. 8

Protein–protein interaction (PPI) network and heatmap analysis for the candidate gene of BnaC03g27700D and its linked genes. The PPI network was constructed using STRING (v12.0) (A). The larger circle indicates that more genes were associated with the gene. The colors from light red to blue red indicates that more genes were associated with the candidate gene. The 2^−ΔΔCt algorithm was used to calculate the relative level of gene expression in (B), and β-actin gene served as an internal control

To further investigate the interconnected genes within the network, we conducted an analysis of the expression levels of all 18 genes across four developmental stages (S, SM, M, and L) in both NT7G132A^Bna1205ams1 and NT7G132B^Bna1205ams1 (Fig. 8B). Among these genes, BnaC03g29860D (APA3) exhibited a significantly elevated expression level during the L stage of NT7G132B^Bna1205ams1 when compared to NT7G132A^Bna1205ams1. Based on the functional analysis performed, BnaC03g29860D (APA3) and BnaC07g33540D (RD19C) emerge as the most probable candidates for involvement in the genetic regulation of male sterility through their interaction with the candidate gene BnaC03g27700D. This interaction will be the primary focus of future research in our laboratory.

Large metabolic fluctuations were caused by the male sterility gene mutation

To investigate and annotate the functions and consequences of the Bna1205ams1 mutation, a comprehensive metabolic analysis was performed across four developmental stages: S, SM, M, and L, comparing the NT7G132A^Bna1205ams1 genotype with the NT7G132B^Bna1205ams1 genotype. A total of 48 samples were analyzed, comprising four sets of NT7G132A^Bna1205ams1 and NT7G132B^Bna1205ams1, each with six replicates. The analysis identified 784 metabolites were identified using the positive ion mode (POS) and 569 metabolites using the negative ion mode (NEG) (Table S10). A comparison assessment of the metabolic profiles between NT7G132A^Bna1205ams1 and NT7G132B^Bna1205ams1 at each developmental stage revealed a total of 200 differential metabolites (DMs) (Table S11), exhibiting a nearly equal distribution between upregulated and downregulated metabolites (Fig. 9A). Notably, the number of DMs in the S stage was relatively low, and the fold change values for the majority of DMs falling below one (87.0%), indicating that the expression of most genes remains consistent between NT7G132A^Bna1205ams1 genotype and NT7G132B^Bna1205ams1 genotype (Fig. 9A). This observation is consistent with the findings from the cytological analysis (Fig. 9A). Subsequently, the total number of DMs increased, peaking in the M stage, before declining in the L stage (Fig. 9A). Throughout all stages, the most significant KEGG pathway identified was that associated with lipids and lipid-like molecules (Fig. 9B).

Fig. 9

Analysis of the metabolic fluctuations resulted from the mutation of male sterility gene. The number of differential metabolites (DMs) between 1205A and 1205B at four stages of S, SM, M and L were analyzed in (A). The functional classes were analyzed for the DMs in the four stages of S, SM, M and L, the larger bubble size indicates more DMs in (B). The correlation heatmap between the expression level of BnaC03g27700D and the abundance of metabolites at four stages of S, SM, M and L was analyzed and shown using the Mantel test, while the correlation heatmap between different identified metabolites at four stages of S, SM, M and L was analyzed using the Pearson method in (C)

The study performed a correlation analysis to examine the relationship between the expression levels of the candidate gene BnaC03g27700D across four developmental stages (S, SM, M, and L) and the abundance of various differentially expressed metabolites (DMs). The findings indicated a significant correlation between the expression level of BnaC03g27700D and 13 specific metabolites. These metabolites included four lipids and lipid-like molecules (NEG_12546, NEG_5420, NEG_9009, and POS_12529), three phenylpropanoids and polyketides (POS_30211, POS_8691, and NEG_11481), two organoheterocyclic compounds (POS_7834, and NEG_14166), as well as one each of organic nitrogen compounds (POS_7834), organic oxygen compounds (NEG_14166), and nucleosides/nucleotides (POS_4492) (Fig. 9C; Table S12). Notably, among these specific metabolites, four exhibited upregulation at more than two developmental stages: NEG_11481 (fold change: 6.54 at S; 3.79 at SL; 6.54 at M), NEG_14166 (fold change: 1.26 at SL; 3.47 at M; 1.55 at L), and POS_7834 (fold change: 1.66 at SL; 5.02 at M; 3.55 at L). These metabolites, along with their highly regulated genes, may serve as significant subjects for future interaction analyses involving BnaC03g27700D.

Discussion

In China, the RGMS system has emerged as a significant two-line hybrid system, with several RGMS-based lines already registered and employed for hybrid production [8, 10, 73]. When compared to the three-line hybrid system, the two-line hybrid system presents several advantages: it ensures complete and stable male sterility, facilitates the transfer of the sterility trait across diverse genetic backgrounds, provides abundant recovery resources [74], and simplifies the hybrid production process. These attributes render the two-line system particularly advantageous for regions experiencing unstable temperatures during the flowering period of B. napus, such as certain areas in Guizhou Province, China, where temperatures frequently fluctuate between 0 °C and 10 °C. Such temperature variations may result in the emergence of fertile pollens in the CMS lines.

However, the large-scale application of RGMS lines in hybrid production necessitates addressing the primary challenge of extracting 50% fertile plants. One potential solution involves the development of associated markers linked to the fertility and sterility genes present in these RGMS lines. Numerous previous studies have endeavored to create associated markers for marker-assisted selection (MAS) of the fertility/sterility genes just prior to flowering in B. napus. For instance, two sequence characterized amplified region (SCAR) markers, SC1 and SC7, which are associated with the male sterility regulation gene Bnms1, were developed based on polymorphic Amplified Fragment Length Polymorphism (AFLP) markers identified between the lines S45A and S45B [10]. Similarly, the identification of two AFLP fragments linked to a recessive genic male sterility gene (Ms) within a two-line hybrid production system, 117AB, and their subsequent conversion to SCAR markers facilitated the marker-aided selection of fertile plants [41]. In another study, four ILP markers, designed based on a common microsyntenic region with Arabidopsis chromosome 3, were found to be tightly linked to the male sterility regulation genes BnMs4 and BnRf in the line7635AB, which could also be employed in MAS breeding [75]. Although all of these linked markers are traditional molecular markers, their application for large-scale screening and extraction of fertile plants in RGMS lines remains a significant challenge. The development of KASP markers presents an opportunity for large-scale and robotic screening and extraction of fertile plants in RGMS lines, as evidenced by the three KASP markers developed in this study.

Another potential approach for the extraction of fertile plants within the two-line hybrid production system involves the identification of additional sterile lines and the execution of comprehensive research on their regulatory mechanisms. The extensively investigated and conserved regulatory pathways associated with male sterility mechanisms, particularly those related to the anther and pollen development, have been documented in model species such as Arabidopsis [76,77,78,79]. These pathways encompass 82 genes related to lipid metabolism [80] and 63 genes associated with sugar metabolism [81], both of which are classified as GMS genes. The elucidation of these GMS genes is essential for a deeper understanding of the regulatory mechanisms governing fertility in rapeseed. The recessive gene BnCYP704B1 in the line S45AB was cloned and characterized based on the collinearity between Arabidopsis and Brassica napus, as well as the functional insights regarding CYP704B1 derived from studies in Arabidopsis [40]. Additionally, a recessive candidate gene, BnaC03g56870D, which regulates male sterility in one RGMS line, was identified from the progeny resulting from a cross between B. napus and B. rapa. This gene was mapped using BSA-Seq and was ultimately characterized through transcriptome analysis and its functional annotation, revealing homology with the CYP86C4 gene in Arabidopsis thaliana [38]. In the line 9012AB, the final selection of the candidate gene BnRf^b was also informed by studies involving a nucleus-localized chimeric protein, HSP70-1-like, in Arabidopsis [43]. In the current study, we identified a QTL region measuring 181.47 kb that exhibited collinearity with the genomic fragment located between At5g49940 and At5g50320 on chromosome 5 in Arabidopsis. This discovery facilitated the identification of 13 candidate genes, ultimately leading to the identification of the candidate gene BnaC03g27700D.

Apoptosis-like PCD is a critical process that can initiate cellular degeneration, thereby positioning the tapetum as a vital secretory component in microspore development. The tapetum contributes enzymes necessary for microspore release, nutrients essential for pollen development, and structural components for the pollen wall [43, 82]. Therefore, the timely occurrence of tapetum PCD is imperative for the proper development of microspores and pollen, a process governed by a complex network [83]. The orthologous gene corresponding to the identified candidate gene BnaC03g27700D is CYSTEINE ENDOPEPTIDASE 1 (CEP1), which has been documented to play a significant role in tapetal PCD and is implicated in the regulation of pollen development in Arabidopsis [26]. In the current study, cytological analyses indicate that the anthers begin to undergo premature degradation during the early microspore stage (SL stage), potentially attributable to an aberrant process of tapetal PCD. Expression analysis further corroborates this hypothesis, revealing that the expression level of BnaC03g27700D is several times higher in the 1205A line compared to the 1205B line, commencing from the SL stage. The markedly elevated expression level of BnaC03g27700D in 1205A may lead to severely premature tapetum PCD, which could lead to male sterility by disrupting the normal regulatory mechanisms of PCD. A comparable scenario has been documented in the 9012A line, where the abnormal expression of the candidate gene BnaA7.mtHSP70-1-like results in the improper timing of tapetal PCD and subsequent male sterility [43]. Notably, in the BnaA7.mtHSP70-1-like transformation lines of Arabidopsis, the expression levels of CEP1, along with another protease gene, UNDEAD, which are proposed executors of tapetum PCD [26, 84], exhibited significantly reduced expression levels [43]. This suggests that the gene BnaA7.mtHSP70-1-like may function as a positive regulator of CEP1. Furthermore, a transcription factor gene, MYB2, which is crucial for tapetal PCD and pollen development, has been reported to activate CEP1 in Arabidopsis [85]. Therefore, the abnormal variation observed in BnaC03g27700D may arise from intricate regulatory network interactions. This study presents a significant case for exploring the regulatory mechanisms underlying RGMS and provides valuable insights for enhancing the functional annotation of this gene in rapeseed.

Two-line hybrid systems have typically undergone more than ten generations of sibling mating, rendering them particularly suitable for the mapping of the male sterility gene. The “A” and “B” types are nearly isogenic, differing primarily at loci that regulate male sterility and a few tightly linked loci [43]. Concurrently, other loci that are not closely associated with the male sterility locus have become homozygous following extensive sibling mating [43]. This phenomenon has resulted in a bottleneck within two-line seed production systems, as the limited diversity of fertility-associated phenotypes impedes the rapid identification of fertile plants. Consequently, this presents a significant challenge to the broader application of two-line sterile lines and diminishes research interest in this domain. A recent study has introduced a visible seedling-stage screening system for B. napus hybrid breeding, utilizing the hypocotyl length-regulated gene BnHL [86]. This gene, which is linked to the fertility gene BnMs2, functions as a marker for seedling morphology. Notably, hypocotyl length at seven days post-germination exhibited significant differences between sterile and fertile plants, facilitating effective identification under both white (W) and red/far-red (R/FR) light conditions [86]. This research has stimulated further investigations aimed at identifying fertile strains through phenotype and metabolite analyses. Analyzing DMs or correlating with the expression of BnaC03g27700D may provide valuable insights into the effects of genes on metabolites and offer methodologies for distinguishing fertile plants from nuclear sterile plants in the 1205AB context.