A β-Ketoacyl-CoA Synthase encoded by DDP1 controls rice anther dehiscence and pollen fertility by maintaining lipid homeostasis in the tapetum

doi:10.21203/rs.3.rs-4822227/v1

Download PDF

Research Article

A β-Ketoacyl-CoA Synthase encoded by DDP1 controls rice anther dehiscence and pollen fertility by maintaining lipid homeostasis in the tapetum

https://doi.org/10.21203/rs.3.rs-4822227/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Anther dehiscence and pollen fertility are crucial for male fertility in rice. Here, we studied the function of Defective in Dehiscence and Pollen1 (DDP1), a novel member of the KCS family in rice, in regulating anther dehiscence and pollen fertility. DDP1 encodes an endoplasmic reticulum (ER)-localized protein and is ubiquitously expressed in various organs, predominately in the microspores and tapetum. The ddp1 mutant exhibited partial male sterility attributed to defective anther dehiscence and pollen fertility, which was notably distinct from those observed in Arabidopsis thaliana and rice mutants associated with lipid metabolism. Mutations of DDP1 altered the content and composition of wax on anther epidermis and pollen wall, causing abnormalities in their morphology. Moreover, genes implicated in lipid metabolism, pollen development and anther dehiscence exhibited significantly altered expression levels in the ddp1 mutant. These findings indicate that DDP1 controls anther dehiscence and pollen fertility to ensure normal male development by modulating lipid homeostasis in the tapetum, thereby enhancing our understanding of the mechanisms underlying rice anther dehiscence and pollen fertility.

DDP1 affects rice anther dehiscence and pollen fertility by maintaining lipid homeostasis in the tapetum.

Rice (Oryza sativa L.), one of the primary food crops globally, feeds approximately 50% of the global population. Increasing rice yields is essential for global food security, but low seed-setting rate impedes rice yield increase. Pollen fertility, which depends on the proper development of both pollen and anthers, significantly influences seed setting and rice yield. The pollen wall, a lipid-rich structure, contains the exine and intine (Shi et al. 2015). The anther wall is made up of four distinct layers, the outermost layer is the epidermis, followed in sequence by the endothecium and the middle layer, the innermost layer is tapetum (Zhang et al. 2011). The tapetum provides lipid precursors crucial for sporopollenin synthesis within the pollen wall and for wax and cutin formation in the anther epidermis. Aberrant tapetum development commonly leads to defective anthers and pollen. In addition to tapetum function, male fertility also depends on the timely release of mature pollen grains, which is strictly regulated by anther dehiscence. However, the precise mechanisms underlying anther dehiscence and pollen fertility remain unclear. Thus, elucidating the molecular mechanism behind these processes will not only enhance our knowledge of rice biology but also significantly propel the genetic improvement of rice fertility and yield traits.

The main components of the anther epidermis and pollen wall include very long-chain fatty acids (VLCFAs) and related lipid derivatives.They provide critical protection against dehydration, environmental stresses and pathogen attacks (Buschhaus and Jetter 2011; Jenks et al. 1994). VLCFAs are initially de novo synthesized in the plastid and then elongated in ER, which is facilitated by the fatty acid elongation (FAE) complex (Ariizumi and Toriyama 2011; Shi et al. 2015). The FAE complex is made up of four distinct enzymes: β-ketoacyl-CoA synthase (KCS) for condensation, β-ketoacyl-CoA reductase (KCR) for reduction, 3-hydroxyacyl-CoA dehydratase (HCD) for dehydration, and trans-2,3-enoyl-CoA reductase (ECR) for secondary reduction. Together, they extend the carbon chain length of the substrate by incorporating two additional carbon atoms in each catalytic cycle (Lee and Suh 2013). Among the FAE enzymes, the KCS family exhibit strict substrate specificity, determining the chain lengths of the substrate (Blacklock and Jaworski 2006). KCS enzymes are also highly conserved across plants, with varying numbers of KCS genes from Arabidopsis thaliana, Zea mays, Oryza sativa, Arachis hypogaea and Gossypium hirsutum (Joubès et al. 2008; Yang et al. 2023; Campbell et al. 2019; Huai et al. 2020; Xiao et al. 2016). Notably, KCS6, characterized by its broad tissue distribution and potent catalytic activity, not only catalyzes the extension of C24 to C28 VLCFAs on its own but also synergizes with a CER2-like protein to promote the elongation of C30 to C34 VLCFAs (Haslam and Kunst 2013; Huang et al. 2022; Haslam et al. 2015).

Abundant studies have explored the roles of KCS genes in Arabidopsis thaliana, where KCS mutations generally lead to the reduction in wax production and various deficiencies in vegetative development, such as KCS6, KCS1, KCS2/DAISY, KCS20, KCS9, KCS16, KCS4, KCS5 (Millar et al. 1999; Fiebig et al. 2000; Todd et al. 1999; Lee et al. 2009; Franke et al. 2009; Lee et al. 2009; Kim et al. 2013; Hegebarth et al. 2017; Kim et al. 2021; Huang et al. 2022). In contrast, functional characterization of KCS enzymes in rice is limited. Mutations or overexpression of rice KCSs can result in diverse phenotypes by affecting wax deposition. For instance, wsl1 displays enhanced sensitivity to drought and low fertility (Yu et al. 2008), while sd38 mutation results in a semi-dwarf phenotype (Zhang et al. 2022); wsl4 exhibits reduced wax accumulation on leaves and its allelic mutant hms1 is characterized by decreased seed setting under low humidity condition (Gan et al. 2017; Chen et al. 2020). Additionally, oni1 and oni2 are seedling-lethal (Tsuda et al. 2013), and overexpression of OsCUT1 is closely associated with enhanced drought resistance (Gao et al. 2022). In conclusion, while KCS homologs share conserved biochemical function in wax deposition, their biological functions exhibit significant differentiation. However, the specific roles of KCS in contributing to male fertility are still unclear.

In our research, we have conducted a functional characterization of DDP1 (Defective in Dehiscence and Pollen1), a novel member of the KCS family in rice. The ddp1 mutant showed a partial male sterile phenotype attributed to defective anther dehiscence and pollen fertility. This phenotype was notably distinct from those observed in Arabidopsis thaliana and rice mutants associated with lipid metabolism, underscoring the unique role of DDP1 in male fertility.

Plant materials and growth conditions

We generated knockout (KO) mutants ddp1-1 and ddp1-2 utilizing CRISPR/Cas9 technology. The target sequence within the DDP1coding sequence was selected with the aid of designer software specifically tailored for CRISPR target prediction (Xie et al. 2017) (Fig. 1a). Subsequently, the target sequence was synthesized and integrated into a CRISPR/Cas9 expression vector. Following its construction, the plasmid was transferred into the japonica rice cultivar Zhonghua 11. The plant materials analyzed in this research were cultivated in an experimental field at Nanning (Guangxi, China) under natural conditions and managed with standard agricultural practices.

Characterization of mutant phenotypes

A Nikon digital camera was utilized to photograph the phenotypes of whole plants and panicles after seed maturation. An sz61TR trinocular microscope (Olympus) was used to photograph the phenotypes of spikelets and anthers. Pollen fertility was examined by staining mature pollen grains using 1% I₂-KI and photographed using an Axioscope 5 microscope (ZEISS). Yield traits were measured using the methods previously outlined by Qin et al (2021). Anther samples collected at various intervals during the flowering period were immersed in a 2.5% glutaraldehyde solution (v/v) to prepare them for subsequent examination via scanning electron microscopy (SEM) and transmission electron microscopy (TEM) following the previously described methods (Qin et al. 2013).

In vitro and in vivo pollen germination assays

In conducting the in vitro pollen germination assay, we harvested pollen grains at flowering stage and place them on a prepared germination medium containing 20% sucrose, 10% polyethylene glycol 4000, 40 mgL^− 1 H₃BO₃, 3 mmolL^− 1 Ca(NO₃)₂, and 10 mgL^− 1 vitamin B1(VBl) on slide glass. Following a 2 hours culture period at 35℃, pollen germination was observed using an Axioscope 5 microscope (ZEISS). To determine the germination rate, we measured the germination of 250 pollen grains from both ddp1 and WT anthers.

For the in vivo assessment of pollen germination, pistils were harvested 3 hours after pollination and subsequently immersed in Carnoy solution (3:1 ethanol-to-glacial acetic acid mixture) for overnight. Following fixation, the pistils underwent three sequential washes using double-distilled water (ddH₂O) and were then softened in 4M sodium hydroxide (NaOH) for 24 hours. After another three washes with ddH₂O, the pistils were immersed in a 0.05% aniline blue solution for 12h in the dark. After staining, the pistils were carefully mounted onto glass slides for microscopic examination using an Axioscope 5 microscope (ZEISS).

Examination of anther dehiscence

The assessment of anther dehiscence was conducted in accordance with the previously published method (Zhang et al. 2004). The anthers were collected and observed with a stereoscopic microscope to assess dehiscence. The degree of anther dehiscence was categorized into five levels as follows:

●Level 5: Anthers were completely dehisced, releasing all pollen grains.

●Level 4: Anthers had large holes at the top and bottom, nearly releasing all pollen grains.

●Level 3: Anthers with small holes or a single chamber dehiscence released a portion of pollen grains.

●Level 2: Anthers with a small hole at the bottom released a minimal amount of pollen grains.

●Level 1: Anthers that were completely indehiscent did not release any pollen grain.

The anther dehiscence phenotype was quantified using the dehiscence index, calculated as follows:

Histochemical analysis

For nuclear staining, the spikelets of ddp1-1 and WT were collected during stage 9 to 12 and then immersed in Carnoy solution for overnight. After being washed three times with 70% ethanol, anthers were dissected from the spikelets and carefully mounted onto a glass slide containing a small drop of 1 mgL^− 1 DAPI solution. A cover slip was placed over the anthers to prevent evaporation and ensure even staining. After a 10 minutes incubation period, the stained anthers were observed under an Axioscope 5 microscope (ZEISS) to visualize nuclear morphology.

For lipidic staining, the spikelets of ddp1-1 and WT at stage 12 were collected and then fixed in the 70% solution of formalin-acetic-alcohol (FAA) for 24 hour. Following fixation, the spikelets underwent dehydration using a sequential ethanol gradient, ranging from 100–30%. The dehydrated spikelets were then stained by Sudan Red 7B solution. The paraffin sections of anthers were prepared and processed by dewaxing, followed by hydration, and subsequently stained using Sudan Red 7B solution. After staining, the sections were washed three times using 1% sodium dodecyl sulfate (SDS) followed by double distilled H₂O (ddH₂O) to eliminate any surplus stain. The anthers, dissected from the spikelets, as well as paraffin sections and microspores, dissected from the anthers, were carefully mounted onto slides using a 50% (v/v) glycerol solution to prevent dehydration and distortion, and then observed under an Axioscope 5 microscope (ZEISS) for detailed examination.

For the preparation of semi-thin sections, fresh panicles were initially fixed in Carnoy solution, followed by dehydration through an ascending series of ethanol concentrations. Following dehydration, the panicles were encased in EMbed812 resin (Hereaus Kulzer) and left to polymerize, ensuring a stable matrix for sectioning. The transverse sections of the embedded samples were cut to a thickness of 3µm using a Leica slicer (LEICA EMUC7FC7) and were stained in 0.1% (w/v) toluidine blue. The transverse sections were examined and captured with an Axioscope 5 microscope (ZEISS).

For the observation of callose walls, paraffin sections underwent dehydration before being stained with a 0.1% (w/v) solution of aniline blue. Subsequently, they were mounted on slides using a 50% (v/v) glycerol medium and captured with an Axioscope 5 microscope (ZEISS).

Wax components analysis

Mature anthers from ddp1-1 and WT at stage 12 were collected with five biological replicates. Following vacuum drying, the dried anthers underwent rapid chloroform extraction to isolate wax components. Subsequently, the isolated components underwent a thorough analytical evaluation using gas chromatography-mass spectrometry (GC-MS), which was conducted at the SJTU-Metabolon Joint Metabolomics Lab in China following the previously described methods (Xu et al., 2017).

Protein sequence analysis

The full-length protein sequences of DDP1 homologs from various plant species were retrieved through a BLASTP search conducted on the NCBI database. The full-length protein sequences of DDP1 and its homologs across various plant species were aligned using MUSCLE version 3.6, a widely recognized software for multiple sequence alignment available at the EBI website. Following the alignment, the sequences were utilized to construct a phylogenetic tree, which was generated using the MEGA X software, a widely recognized tool for phylogenetic analysis. The neighbor-joining method was applied with default settings to infer the evolutionary relationships among the sequences. To ensure the reliability of the tree topology, 1000 bootstrap replications were performed, providing a statistical measure of the confidence in the branching order. To predict protein domains, the full-length protein sequences of DDP1 and ddp1-1/2 were analyzed using the SMART tool and the PSIPRED server. These resources are well-established for identifying protein domains and predicting secondary structures, respectively. For three-dimensional structure prediction, the same full-length protein sequences were deposited into the SWISS-MODEL platform, an online resource for protein modeling accessible at SWISS-MODEL.

Quantitative real-time PCR and RNA in situ hybridization

To determine DDP1 expression profiles, total RNA was extracted from multiple tissues of WT plant, such as roots, stems, leaves, and panicles at distinct stages as well as from panicles at different stages of ddp1 plant. The extraction was performed using a TRIzol kit (Invitrogen, USA). Both Reverse Transcription Polymerase Chain Reaction (RT-PCR) and Reverse Transcription Quantitative Polymerase Chain Reaction (RT-qPCR) were performed following established methods (Ma et al. 2021). OsActin was employed as the internal reference gene. Eample was assayed in triplicate to ensure biological reproducibility. The gene primers used for examination in our research were presented in Table S1.

RNA in situ hybridization was conducted following the established protocols detailed in previous studies (Xue et al. 2018). The DDP1 cDNA fragment was used as a template for probe labeling to generate the sense and antisense probes, which were amplified by specific primers (Table S1). The Axioscope 5 microscope (ZEISS) was used to capture the image of the assay result.

Subcellular localization

The complete cDNA sequences of DDP1 and ddp1-1 were replicated and subsequently inserted into the pCAMBIA2300 vector to produce the constructs 35S-GFP-DDP1 and 35S-GFP-ddp1-1. The ER marker was generated by fusing HDEL ER retention signal with mCherry (Caroli et al., 2011). The construct 35S-GFP-DDP1 and 35S-GFP-ddp1-1 was transiently co-introduced with the ER marker mCherry-HDEL into rice protoplasts, respectively. The transformation was performed using the polyethylene glycol (PEG)-mediated method, a widely used technique for introducing DNA into plant cells. A GFP reporter, serving as a control, was utilized to verify the success of the transformation process. The fluorescence signal, indicative of the localization of the GFP-tagged proteins and the ER marker, was detected and recorded through a confocal laser scanning microscope (Leica TCS SP8MP).

ddp1 mutants exhibit partial pollen abortion

The rice genome contains 22 KCS genes (Yang et al. 2023). To investigate the functions of these KCSs, we employed CRISPR/Cas9 technology to generate a series of knockout (KO) mutants targeting these genes (Table S1). Among all developed mutants, the knockout mutants of LOC_Os02g49920 exhibited a partial male sterility attributed to defective anther dehiscence and pollen fertility. Thus, the mutant of LOC_Os02g49920 was designated as ddp1 (defective in dehiscence and pollen1). two alleles ddp1 namely ddp1-1 and ddp1-2 were identified in T₁ progeny. ddp1-1 and ddp1-2 carried 4-bp and 2-bp deletion in the sole exon of LOC_Os02g49920, respectively (Fig. 1b). These deletions caused reading-frame shifts, generating premature and delayed stop codon, respectively.

Compared to WT, ddp1-1/2 mutants exhibited no significant differences in whole-plant morphology and the structure of flower organs such as spikelets and anthers (Fig. 1c-f). However, at the mature stage, the seed-setting rates of ddp1-1 (41.533 ± 1.050%) and ddp1-2 (32.100 ± 1.400%) were significantly lower than that of WT (95.000 ± 0.816%) (Table 1)(Fig. 1g, k). To determine if the lower seed-setting rate in ddp1-1/2 mutants resulted from a male or female gametophyte impairment, we evaluated the pollen and female fertility. Notably, the fertile pollen rate was significantly lower in ddp1-1 (54.67% ± 3.68%) and ddp1-2 (49.17% ± 1.65%) compared to WT, which exhibited a fertile pollen rate of 95.13% ± 0.84% (Fig. 1h-j, l). When ddp1 mutants were self-pollinated with its pollen, their seed setting rate was only 24.45% ± 2.53%. However, after pollination with WT pollen, the seed setting rate of ddp1 mutants reached 58.60% ± 5.88% (Fig. S1). The findings suggest that the reduced seed-setting rate observed in ddp1 mutants is due to abnormalities in pollen development, rather than any deficiency in the female gametophyte.

Table 1

The agronomic traits of WT and *ddp1* mutant.
Agronomic traits	WT Mean ± SD	ddp1-1 Mean ± SD	ddp1-2 Mean ± SD
Plant height (cm)	86.333 ± 2.055	83.667 ± 1.247	88.000 ± 0.816
Panicle length (cm)	21.500 ± 0.408	23.533 ± 0.500	21.400 ± 0.638
1000-grain weight (g)	26.647 ± 0.509	27.406 ± 0.561	26.810 ± 0.752
Grain length (cm)	0.811 ± 0.009	0.797 ± 0.005	0.811 ± 0.009
Grain width (cm)	0.351 ± 0.002	0.354 ± 0.003	0.352 ± 0.002
Grain thickness (mm)	2.32 ± 0.037	2.303 ± 0.054	2.317 ± 0.049
Seed-setting Rate (%)	95.000 ± 0.816	41.533 ± 1.050**	32.100 ± 1.400*
Number of tillers	10.333 ± 1.247	9.667 ± 1.247	9.667 ± 0923
Number of primary branches	10.000 ± 0.816	10.333 ± 1.247	10.000 ± 1.633
Number of secondary branches	27.667 ± 1.700	23.000 ± 2.60*	17.333 ± 1.700*
Yield per plant (g)	16.000 ± 1.633	6.233 ± 1.481**	4.433 ± 0.419**
, * significant at 5% and 1% levels of probability, respectively.

In addition to partial pollen abortion, ddp1-1/2 mutants also exhibited significant differences in certain yield-related traits compared to WT. For instance, the number of secondary branches and yield per plant were significantly lower in ddp1-1 mutant (23.00 ± 2.16% and 6.23 ± 1.48%) compared to WT (27.67 ± 1.70% and 16.00 ± 1.62%) (Table 1). The ddp1-2 mutant showed similar trends to those observed in ddp1-1 (Table 1). Nevertheless, the comparison revealed no notable differences in plant height, number of tillers and primary branches, panicle length, 1000-grain weight, and grain dimensions (length, width, thickness) between ddp1-1/2 and WT (Table 1).

To confirm whether LOC_Os02g49920 was the target gene responsible for the ddp1 mutants phenotype, a complementation test was performed by transforming a construct DPP1_pro:DPP1CDS-mCitrine containing a 2.5-kb DDP1 promoter region, 1.458-kb DDP1 coding region and 0.72-kb mCitrine coding region into Cas9-free ddp1-1 mutants (Table S1). As expected, the complementary transgenic lines exhibited a restoration of seed-setting rates and pollen fertility to WT levels (Fig. S2), thereby confirming that LOC_Os02g49920 is indeed the target gene whose knockout causes partial male sterility in ddp1 mutants.

Anther dehiscence is defective in ddp1 mutants

To determine whether partial male sterility of ddp1-1/2 mutants was associated with the defects in anther dehiscence, pollen germination, tube growth, and dehydration, we compared these critical reproductive processes in ddp1-1/2 mutants to those in WT. For anther dehiscence analysis, we examined the anther dehiscence phenotype in both ddp1-1/2 and WT. WT anthers typically undergo dehiscence, facilitating the dispersal of mature pollen (Fig. 2a, f), while ddp1-1/2 mutants exhibited various defects in anther dehiscence and pollen grain release (Fig. 2b-e, g-j). The anther dehiscence phenotype was quantified using the dehiscence index as reported previously (Zhang et al. 2004). A higher dehiscence index indicates that anthers have dehisced more completely. Our findings indicated that the anther dehiscence index was significantly lower in ddp1-1 (2.89) and ddp1-2 (2.73) compared to WT (4.09) (Table 2). This impaired anther dehiscence in ddp1-1/2 mutants may result in reduced adhesion and germination of fertile pollen grains on the stigmas. To validate this finding, we performed an in vivo pollen germination experiment to evaluate the pollen germination rates on the stigmas of ddp1-1/2 and WT. The assay demonstrated a markedly lower pollen germination rate on the stigmas in ddp1-1/2 mutants relative to WT (Fig. 2n-r, u). Collectively, these results indicate that defective anther dehiscence in ddp1-1/2 mutants hinders the adhesion and subsequent germination of fertile pollen grains on the stigma, thereby reducing the seed-setting rate.

Table 2

The anther dehiscence index of WT and *ddp1*.
	Level of anther dehiscence					Total number of anther	Dehiscence index
	5	4	3	2	1	Total number of anther	Dehiscence index
WT	8	84	0	0	0	92	4.09
ddp1-1	10	34	26	18	24	112	2.89**
ddp1-2	14	41	36	30	41	162	2.73**
** significant at 1% levels of probability.

For pollen germination analysis, we compared pollen germination rate between ddp1-1/2 and WT using an in vitro pollen germination assay. Because ddp1-1/2 mutants contained sterile pollen grains which could not normally germinate (Fig. 2k-m), the overall in vitro pollen germination rates were significantly lower in ddp1-1 (50.65 ± 1.86%) and ddp1-2 (42.23 ± 2.36%) compared to WT (89.4 ± 1.11%) (Fig. 2s). However, when only considering the germination rate of fertile pollen grains, the comparison revealed no notable differences between ddp1-1/2 and WT (with 82.03 ± 0.82% of ddp1-1 and 81.37 ± 0.45% of ddp1-2 versus 84.90 ± 1.35% of WT) (Fig. 2t). These findings suggest that the germination of fertile pollen in ddp1-1/2 mutants remains unaffected.

For pollen tube growth analysis, we compared pollen tube growth within the pistil of ddp1-1/2 and WT using an in vivo pollen germination assay. In self-pollinated WT pistils, at 5 hours post-pollination, pollen tubes had successfully reached the ovule micropyle (Fig. 2n). However, self-pollinated pistils of the ddp1-1/2 mutants showed a reduced germination rate of pollen grains, yet those that did germinate were able to extend their pollen tubes to the ovule micropyle (Fig. 2o, q), suggesting normal pollen tube growth within ddp1-1/2 pistils.

For pollen dehydration analysis, we compared the in vitro dehydration rate of fertile pollen between ddp1-1 and WT. However, no significant differences in the dehydration rate of fertile pollen were observed between ddp1-1 and WT from 0 second to 30 minutes (Fig. S3). Altogether, these findings suggest that while pollen germination, pollen tube growth and pollen dehydration are not compromised in ddp1-1/2 mutants, anther dehiscence is abnormal. Consequently, we conclude that the decrease in seed-setting rate of ddp1-1/2 mutants was attributed to the aborted pollen and aberrant anther dehiscence.

Cytological comparison of anther development between ddp1 and WT

To determine the morphological defect of ddp1-1/2 anthers, semi-thin sections of both ddp1-1/2 and WT anthers at different development stages were examined using light microscopy. Initial observations revealed that no significant differences were detected between ddp1-1/2 and WT anthers from stage 6 to 9 (Fig. 3a-c, h-j, p-r). However, at stage 10, the epidermis and endothecium of ddp1-1/2 anthers appeared to exhibit a greater thickness compared to their counterparts in WT anthers, and microspores of ddp1-1/2 exhibited collapse and a defect in vacuolation (Fig. 3k, s). At stage 11, the epidermis and endothecium of ddp1-1/2 anthers continued to show swelling, and tapetum degradation seemed to be delayed. Only a fraction of microspores formed falcate-shaped pollen grains filled with starch granules, while the rest developed into irregularly shaped pollen grains lacking starch accumulation (Fig. 3l, t). At stage 12, the WT tapetum had completely degraded, allowing microspores to mature into pollen grains (Fig. 3f). In contrast, in ddp1-1/2 anthers, the incompletely degraded tapetum residues were observed, with only some microspores maturing into fertile pollen grains, while others aborted (Fig. 3m, u). At stage 14, WT anthers were able to dehisce normally as the connective tissue between anthers cracked (Fig. 3g). However, in ddp1-1/2, only a subset of anthers dehisced, while others remained closed (Fig. 3n-o, v-w). These findings show that ddp1-1/2 mutations lead to developmental defects in anther and pollen from stage 10 to 14.

To precisely characterize the differences between ddp1-1/2 and WT, SEM was employed to compare the anthers and pollen grains. The anthers of ddp1-1/2 and WT were comparable in size (Fig. 4a, e, i). WT anthers exhibited an outer epidermis covered by spaghetti-like cuticle layers (Fig. 4b) and an inner epidermis with Ubisch bodies (Fig. 4c, d). In contrast, the cuticle layers of the outer epidermis and the Ubisch bodies of the inner epidermis in ddp1-1/2 anthers seemed to be denser (Fig. 4f-h, g-l). In addition, WT pollen grains were spherical and smooth, uniformly coated by a layer of sporopollenin (Fig. 5a, d). ddp1-1/2 mutants had two types of pollen grains, namely spherical pollen grains and shrunken pollen grains (Fig. 5b-c, e-h). The spherical pollen grains in ddp1-1/2 mutants exhibited a slightly sparse sporopollenin layer (Fig. 5j, l), while the sporopollenin on the shrunken pollen grains in ddp1-1/2 mutants appeared to be denser (Fig. 5k, m). The increased accumulation of sporopollenin on the surfaces of shrunken ddp1-1/2 pollen may be attributed to the shrinkage of the pollen grain surface area. Given the similar phenotypes exhibited by ddp1-1 and ddp1-2 under identical growth conditions, the following analyses within this study were conducted solely on ddp1-1.

To further elucidate the cytological differences between ddp1 and WT, TEM was also employed to observe the ultrastructures of anthers and pollen grains. At stage 11, in accordance with our light microscopy results, the anther wall, particularly the tapetum of ddp1-1 mutant was thicker compared to that of WT (Fig. 6a, e). At stage 13, the WT tapetum fully degraded while the incompletely degraded tapetum residue was still observed in ddp1-1 mutant (Fig. 6i, m). At stage 11 and 13, the cuticle layer and the Ubisch body in ddp1-1 anthers were denser than those of WT (Fig. 6f-g, n-o), aligning with SEM observations. Furthermore, the aborted ddp1-1 pollen grains exhibited a thicker exine and smaller bacula compared to WT (Fig. 6h, p), indicating the alterations in the pollen wall ultrastructure. At stage 10, neither ddp1-1 nor WT microspores showed signs of starch accumulation and intine formation (Fig. 6q, r). By stage 12, WT microspores had developed an intine and were filled with a substantial amount of starch granules (Fig. 6s), whereas aborted pollen grains of ddp1-1 mutant lacked both features (Fig. 6t). In summary, ddp1-1 mutant exhibited abnormalities in their anther cuticle layer, Ubisch bodies, tapetum degradation, anther dehiscence, pollen wall structure, and starch accumulation.

Gametogenesis at stage 11 and 12 is defective in ddp1-1 anthers

To determine whether gametogenesis was affected in ddp1 anthers, we performed the 4’,6-diamidino-2-phenylindole (DAPI) staining on microspores from both ddp1-1 and WT at different stages. At stage 9, WT microspores typically developed into uninucleate microspores. At stage 11, they underwent the first mitotic division, resulting in binucleate microspores. At stage 12, they underwent the second mitotic division, yielding trinucleate microspores with one vegetative nucleus and two generative nuclei. Accordingly, a portion of ddp1-1 microspores successfully developed into uninucleate, binucleate and trinucleate microspores at stage 9, 11 and 12, respectively, similar to WT (Fig. 7a). However, other ddp1-1 microspores exhibited a brightly stained nucleus at stage 9 and 11 but failed to complete the first and second mitotic divisions to form binucleate and trinucleate microspores at stage 11 and 12 (Fig. 7a). The DAPI staining analysis indicates that defective gametogenesis in the aborted ddp1-1 microspores begins at stage 11, which is roughly consistent with the semi-thin section observations that anther and pollen abnormalities occur at stage 10.

Defective gametogenesis was often closely associated with abnormalities in callose biosynthesis and degradation. Callose biosynthesis initiated at stage 7, and its degradation at stage 9 facilitated the release of microspores from the tetrads. In order to investigate whether these processes were affected in ddp1-1 anthers, we performed aniline blue staining on anthers from both ddp1-1 and WT from stage 7 to 9. However, no obvious differences in the fluorescence signal surrounding microspores were observed between ddp1-1 and WT anthers (Fig. 7b). This result suggests that the callose formation and degradation steps in ddp1-1 anthers from stage 7 to 9 are not impaired, consistent with the semi-thin section observations that anther and microspore development in ddp1-1 mutant appeared normal from stage 6 to 9.

Altered wax content and composition of the anther epidermis and pollen wall in the ddp1 mutant

DDP1 encodes a β-ketoacyl-CoA synthase with 485 amino acids residues, catalyzing cuticular wax synthesis. To elucidate the function of DDP1, a phylogenetic analysis was performed using DDP1 and other plant KCS proteins with both known and unknown biochemical and biological functions. Phylogenic analysis revealed that DDP1 shares high sequence similarity with its homologs (Fig. S4a). Domain homology analysis using SMART demonstrated that the DDP1 protein contains three conserved domains: an ACP_syn_III_C domain, a transmembrane region, and a FAE1_CUT1_RppA domain, similar to its counterparts in other plant species (Fig. S4b). The analysis suggests that DDP1 may play a conserved function in cuticular wax synthesis. Given that ddp1-1 mutant also showed structural abnormalities in the lipidic anther epidermis and pollen wall, we speculated that the wax content and composition of these structures in the ddp1-1 mutant might be altered. To test this speculation, we initially examined the lipidic compounds by staining anthers and pollen from ddp1-1 and WT at stage 12 with the lipophilic dye Sudan Red 7B. While the WT anther epidermis and pollen wall showed intense staining (Fig. 8a1-a3,), the ddp1-1 mutant displayed strong staining in anther epidermis and fertile pollen wall, and weak staining in the wall of aborted pollen grains (Fig. 8b1-b3). This suggests a reduction in lipidic compounds within the wall of ddp1-1 aborted pollen. Subsequently, we analyzed the phenolic compounds by comparing the fluorescence intensity emitted by phenolic compounds in pollen walls under ultraviolet radiation between ddp1-1 and WT. The fertile pollen grains of ddp1-1 emitted fluorescence similar in intensity to that of WT, whereas the aborted pollen grains exhibited significantly weaker fluorescence (Fig. 8b4), indicating a substantial decrease in phenolic compounds in the wall of ddp1-1 aborted pollen. In conclusion, these findings indicate that precursors to sporopollenin, including lipidic and phenolic compounds, are significantly reduced in the walls of ddp1-1 aborted pollen.

To ascertain the chemical profiles of anther epidermis and pollen wall, we quantified wax constituents at stage 12 using gas GC-MS. This result indicated that the total wax content in ddp1-1 anthers increased by 34%±0.08% compared to WT (Fig. 8c), primarily due to significant elevations in free fatty acids (C16:0, C18:2, C18:3, C18:0, C20:0, C22:0, C26:0,C28:0 and C30:0), alkenes (C25:0, C27:0, C33:1, C33:0, C35:0 and C35:1), and alcohols (C26:0 and C28:0), as well as sterols such as campersterol, stigmasterol and β-sitosterol (Fig. 8d, e). In contrast, the levels of certain free fatty acid (C24:0), alkenes (C23:0 and C27:1) and inositol were dramatically decreased in ddp1-1 mutants, and other wax constituents remained largely unchanged (Fig. 8e). The altered wax profile in ddp1-1 mutant indicated that DDP1 is essential for maintaining lipid metabolic homeostasis. The alteration in the content and composition of wax in the anther epidermis and pollen wall may underlies the structural abnormalities observed in the ddp1-1 anther epidermis and pollen wall.

Expression pattern and subcellular location of DDP1

To determine DDP1 expression pattern, RT-qPCR was conducted to quantify its expression in various tissues of WT. The result suggested that DDP1 transcription was uniformly detected across all assessed tissues, with peak transcription observed in the leaves (Fig. 9a). DDP1 expression was observed throughout the stages of anther development, peaking at stage 8 (Fig. 9a). To further analyze the expression of DDP1 in anthers, RNA in situ hybridization was conducted on sections of WT anthers. The hybridization signal was first observed in tapetum and pollen mother cells (PMCs) at stage 7, then in tapetum and tetrads at stage 8, and in tapetum and microspores from stage 9 to 11 (Fig. 9b). In control experiments, the sense probe failed to produce any detectable signal, confirming the specificity of the hybridization. These findings indicate that DDP1 is predominantly expressed in the tapetum, PMCs and microspores, which correlates with the phenotypes of delayed tapetum degradation and impaired microspore development in ddp1 mutants.

To further elucidate the spatial distribution of DDP1 protein, we utilized confocal microscopy to observe the mCitrine signal in ddp1-complementary transgenic plants expressing DDP1pro:DDP1cDNA-mCitrine. At stage 10, yellow mCitrine fluorescence was detected in the tapetum and microspores (Fig. 10a-c). By stage 11, yellow mCitrine fluorescence was noted in the partially degraded tapetum residues and microspores (Fig. 10d-f). At stage 12, with the tapetum fully degraded, mCitrine signal was exclusively detected in pollen grains (Fig. 10g-i). These observations indicate that DDP1 expression aligns with DDP1 protein distribution in the tapetum, microspores, and pollen grains.

Protein domain analysis using SMART revealed that the DDP1 protein featured a transmembrane domain at its N-terminus (Fig. S4b). Predictions of secondary structure and three-dimensional models using PSIPRED and SWISS-MODEL indicated that DDP1 possesses coil, helix, and strand domains (Fig. S5a). Mutations in the first helix domain of the ddp1-1/2 mutants led to noticeable alterations in the protein's three-dimensional structure (Fig. S5b, c).

To confirm the localization of DDP1 and ddp1-1 in rice protoplasts, we produced DDP1-GFP and ddp1-1-GFP fusion constructs and introduced them into rice protoplasts alongside the ER marker mCherry-HDEL. DDP1-GFP fluorescence was found to almost completely overlap with the mCherry-HDEL fluorescence (Fig. 9c), suggesting that DDP1 is specifically localized to the ER, as was expected. In contrast, the ddp1-1-GFP fluorescence signal only partially co-localized with the ER marker, with a significant portion of the green fluorescence observed in the cytoplasm (Fig. 9c).

The expression levels of known genes complicated in lipid metabolism, anther and pollen development, and anther dehiscence are significantly changed in ddp1 anthers

Observations of delayed tapetum degradation, defective anther and pollen development, impaired lipid metabolism, and abnormal anther dehiscence in the ddp1-1 mutant have led us to hypothesize that DDP1 mutations may disrupt the expression of genes associated with these processes. To test this speculation, we performed a comparative analysis to evaluate the expression profiles of known genes associated with these processes between ddp1-1 and WT from stage 7 to 10. Compared to WT, the expression levels of six lipid metabolism genes (OsABCG26, OsPKS2, OsNP1, RMS2, CYP703A3 and OsCER2), five genes complicated in the anther, pollen and tapetum development (OsCP1, GAMYB, OsSTRL2, STS1 and DPW3), and two genes involved in anther dehiscence (DAO and OsH1) were significantly decreased in the ddp1-1 mutant (Fig. 11). In contrast, the ddp1-1 mutant displayed a pronounced upregulation in the expression levels of two lipid metabolism genes (OsABCG3 and OsCER1), along with one tapetum development gene (OsAP25), specifically at stages 8 and 9 (Fig. 11). In addition, considering the close association of anther dehiscence with auxin and jasmonic acid (JA) signaling, we also conducted a comparative analysis of gene expression levels related to these signaling pathways between ddp1-1 and WT. Six genes (OsETTIN1, OsETTIN2, OsETTIN3, OsYUCCA1, OsYUCCA4 and OsMP) required for auxin synthesis and response and three JA signaling suppressor genes (JAZ1, JAZ6 and JAZ8) were prominently up-regulated expressed in the ddp1-1 mutant, while a significant downregulation of two JA synthesis genes (LOX2 and AOS3) was detected in the ddp1-1 mutant at stage 9 and 10, respectively (Fig. 11). Together, these findings suggest that mutations in DDP1 alter the expression of known genes complicated in lipid metabolism, anther development and dehiscence, thus causing defective phenotypes associated with these processes.

The anther epidermis contains cutin and wax, while the pollen exine is made up of sporopollenin, with their lipid precursors and derivatives being synthesized by the tapetum. Thus, the homeostasis of lipid metabolism in the tapetum is critical for the development of anther epidermis and pollen wall. This concept is supported by evidence that mutations in several genes associated with the accumulation of lipids in tapetum, including WDA1 (Jung et al. 2006), OsC6 (Zhang et al. 2010), CYP704B2 (Li et al. 2010), DPW (Shi et al. 2011), OsABCG15 (Wu et al. 2014), OsABCG26 (Chang et al. 2016) and CYP703A3 (Yang et al. 2018), cause defective tapetum, anther epidermis and pollen wall, ultimately leading to pollen sterility. In our research, we demonstrated that DDP1, a novel member of the KCS family in rice, is essential for pollen fertility and lipid homeostasis in the tapetum. DDP1 was highly expressed in tapetum and microspore. DDP1 was localized in the ER, and distributed in tapetum, microspore and pollen. ddp1-1 mutant exhibited defects in the anther epidermis and pollen wall. Additionally, compared to WT, the wax content and composition in ddp1-1 were dramatically changed, and a significant shift in the expression levels of gene complicated in lipid metabolism, anther and pollen development was observed in the ddp1-1 anthers. These findings show that DDP1, like many genes associated with lipid metabolism, affects rice pollen fertility by modulating lipid metabolic homeostasis: the knockout of DDP1 disrupts this homeostasis, resulting in partial pollen abortion.

Although DDP1 is conserved across plant species, the phenotype of the ddp1 mutants was significantly distinct from those of its homologs from Arabidopsis and rice. The phenotypes of these homologs included reduced wax accumulation, defective growth of roots and pollen tubes, semi-dwarfism, seedling lethality, and increased drought resistance. However, ddp1 mutants showed no notable vegetative phenotypes, suggesting potential functional diversification between DDP1 and its homologs. HMS1, a DDP1 homolog, has also been implicated in male fertility. However, hms1 mutant exhibited a more severe complete male sterility compared to the partial male sterility of ddp1 mutants. The sterility of hms1 was due to reduced pollen adhesion and germination on stigmas, while sterility of ddp1 mutants arose from defects in the tapetum, anther epidermis, pollen wall, and anther dehiscence. These findings indicate a functional differentiation between DDP1 and HMS1 in the regulation of male fertility. Moreover, there was a difference in the expression patterns of DDP1 homologous genes. For examples, DDP1, OsWSL1, SD38 and OsWSL4/OsHMS1 were expressed in all examined organs, including roots, stems, leaves and panicles, while OsCUT1 was found to has a higher level of transcription in young panicles and leaves. The transcription of ONI1 and ONI2 was confined to specific regions, notably the outer cell layer of the shoot apical meristem and the developmental stages of lateral organs. The significant differences in mutant phenotype and gene expression pattern between DDP1 and its homologs suggest that they may have different biological functions in plant development. Investigating the function of DDP1 will be instrumental in understanding the broader functions of KCS proteins within the plant kingdom.

The ddp1 mutants exhibited a partial male sterile phenotype, suggesting a redundant function for DDP1 in controlling male fertility. The potential redundancy of DDP1 could be due to the widespread presence of VLCFA biosynthesis related genes and their homologs, in addition to the high expression levels of these genes in rice young inflorescences.VLCFA biosynthesis requires four key enzymes, including KCS, KCR, HCD and ECR. Loss of function of members from any of these enzyme families might impair VLCFAs formation, leading to male sterility. In addition, the rice genome contains 22 homologs of DDP1 (Yang et al., 2023). Analysis of gene expression patterns in the Bio-Analytic Resource for Plant Biology database revealed that eight DDP1 homologs with unknown functions, such as OsCUT8, OsKCS16, OsCUT6, OsKCS15, OsKCS13, OsKCS22, OsKCS2, and OsKCS9, are highly expressed in young inflorescences (Fig. S6). This suggests that these homologs may play crucial roles in VLCFA biosynthesis and male fertility, and DDP1 may interact directly or indirectly with other proteins to coordinately regulate rice male fertility. Despite attempts to screen for DDP1-interacting proteins in the rice anther cDNA library using yeast two-hybrid technology, no interacting protein was identified. Further research is necessary to elucidate the regulatory networks of DDP1 in male fertility by identifying its interacting proteins or upstream transcription factors.

Mutations in certain genes associated with lipid metabolism result in reduced anther wax, as observed in KCS6 (Fiebig et al. 2000), WDA1 (Jung et al. 2006), WSL1 (Yu et al. 2008), OsGL1-2 (Islam et al. 2009), OsGL1-1 (Qin et al. 2011), OsGL1-6 (Zhou et al. 2013), ONI1 and ONI2 (Tsuda et al. 2013), OsGL1-3 (Zhou et al. 2015), WSL4/HMS1 (Gan et al. 2017; Chen et al. 2020), and KCS5 (Huang et al. 2022), SD38 (Zhang et al. 2022). In contrast, mutations of other genes associated with lipid metabolism cause a dramatical increase of anther wax, such as DPW2 (Xu et al. 2017), OsPKS2 (Zou et al. 2018), PEM1 (Song et al. 2022), Os12BGlu38 (Shim et al. 2022), KCS3 (Huang et al. 2023a), and KCS12 (Huang et al. 2023b). These findings suggest that changes in anther wax content may be an indirect consequence of mutations in genes related to lipid metabolism. In this study, the ddp1 mutants showed a denser anther cuticle. Consistent with this, elevated levels of free fatty acids, alkenes, alcohols and sterols were detected in the ddp1 mutants. Moreover, two lipid metabolism genes (OsABCG3 and OsCER1) had up-regulated expression in the ddp1 mutants during stages 8 and 9. Given the role of DDP1 in lipid precursor formation, it is plausible to hypothesize that the knockout of DDP1 disrupts lipid metabolism homeostasis, potentially triggering a feedback mechanism that results in a thicker anther cuticle.

In addition to the aforementioned defects in anthers and pollen, the ddp1 mutants also exhibited a distinct defect in anther dehiscence, a phenotype almost rarely observed in mutants associated with lipid metabolism. We speculate that the defect in ddp1 anther dehiscence may be due to the reduced mechanical pressure from aborted pollen grains or may be caused by defective endothecium lignification. Previous research has shown that anther dehiscence is closely related to auxin signaling, jasmonic acid (JA) biosynthesis and endothecium lignification. Excessive indole-3-acetic acid (IAA) accumulation generally leads to decreased JA content and abnormal endothecium lignification, thereby causing anther indehiscence (Cecchetti et al., 2013). Moreover, it is reported that cuticular wax is closely related to IAA and JA. Zhu et al (2014) reported that expression of EsWAX1 from Eutrema salsugineum was slightly increased by IAA treatment, thus promoting accumulation of cuticular wax. Recently, a study demonstrated that there was a negative relationship between total cuticular wax content and JA concentration in maize, the mutation of ZmGL8, encoding a β-ketoacyl-CoA reductase, reduced wax accumulation and activated a JA-dependent pathway (Liu et al. 2024). These results reveal that IAA increases cuticular wax accumulation, thereby reducing JA levels and leading to anther indehiscence. Consistent with this, we observed that in the ddp1 mutants, the total wax content increased by 34%±0.08% compared to WT, a significant increase in expression was observed for six auxin signaling related genes and three JA signaling suppressor genes, whereas two genes critical for the production of JA showed a dramatic decrease in expression levels. These findings imply that DDP1 may regulate anther dehiscence by modulating auxin and JA signaling pathways. However, the exact mechanism by which DDP1 regulates these signaling pathways and endothecium lignification remains unclear. Therefore, additional research is required for fully elucidating the function of DDP1 in auxin signaling, JA biosynthesis and endothecium lignification.

Author contribution statement B. Qin, L. Shang, X. He, and Y. Xu conceived and designed the experiments; Y. Xu, and S. Zhou performed most of the experiments and analyzed the data; J. He, W. Tan, J. Tian, W. Zhao, and J. Wei participated in the phenotype measurement and field experiments; B. Qin and Y. Xu wrote the manuscript; S. Zhou, J. Tian, W. Zhao, J. Wei, and R. Li revised the manuscript. All authors read and approved the final manuscript.

Funding This work was supported by the National Natural Science Foundation of China (32372023 and 31971809), Centrally Guided Local Science and Technology Development Fund (ZY23055028), the Guangxi Natural Science Foundation (2019GXNSFDA185009 and 2024GXNSFGA010003), State Key Laboratory for Conservation and Utilization of Subtropical Agro-bioresources (SKLCUSA-b202306), Innovation Project of Guangxi Graduate Education (YCBZ2022016), the Project of Bama County for Talents in Science and Technology (No.20220009).

Conflict of interest The authors declare that they have no conflict of interest.

Ariizumi T, Toriyama K (2011) Genetic regulation of sporopollenin synthesis and pollen exine development. Ann Rev Plant Biol 62:437–460. https://doi.org/10.1146/annurev-arplant-0
Blacklock B, Jaworski J (2006) Substrate specificity of Arabidopsis 3-ketoacyl-CoA synthases. Biochem Biophys Res Commun 346(2):583–590. https://doi.org/10.1016/j.bbrc.2006.05.162
Buschhaus C, Jetter R (2011) Composition differences between epicuticular and intracuticular wax substructures: how do plants seal their epidermal surfaces? J Exp Bot 62(3):841–853. https://doi.org/10.1093/jxb/erq366
Campbell A, Stenback K, Flyckt K, Hoang T, Perera M, Nikolau B (2019) A single-cell platform for reconstituting and characterizing fatty acid elongase component enzymes. PLoS ONE 14(3):e0213620. https://doi.org/10.1371/journal.pone.0213620
Caroli M, Lenucci M, Sansebastiano GP, Dalessandro G, Lorenzo G, Piro G (2011) Protein trafficking to the cell wall occurs through mechanisms distinguishable from default sorting in tobacco. Plant J 65(2):295–308. https://doi.org/10.1111/j.1365-313X.2010.04421.x
Cecchetti V, Altamura M, Brunetti P, Petrocelli V, Falasca G, Ljung K, Costantino P, Cardarelli M (2013) Auxin controls Arabidopsis anther dehiscence by regulating endothecium lignification and jasmonic acid biosynthesis. Plant J 74(3):411–422. https://doi.org/10.1111/tpj.12130
Chang Z, Chen Z, Yan W, Xie G, Lu J, Wang N, Lu Q, Yao N, Yang G, Xia J, Tang X (2016) An ABC transporter, OsABCG26, is required for anther cuticle and pollen exine formation and pollen-pistil interactions in rice. Plant Sci 253:21–30. https://doi.org/10.1016/j.plantsci.2016.09.006
Chen H, Zhang Z, Ni E, Lin J, Peng G, Huang J, Zhu L, Deng L, Yang F, Luo Q, Sun W, Liu Z, Zhuang C, Liu Y, Zhou H (2020) HMS1 interacts with HMS1I to regulate very-long-chain fatty acid biosynthesis and the humidity-sensitive genic male sterility in rice (Oryza sativa). New Phytol 225(5):2077–2093. https://doi.org/10.1111/nph.16288
Fiebig A, Mayfield J, Miley N, Chau S, Fischer R, Preuss D (2000) Alterations in CER6, a gene identical to CUT1, differentially affect long-chain lipid content on the surface of pollen and stems. Plant Cell 12(10):2001–2008. https://doi.org/10.1105/tpc.12.10.2001
Franke R, Hofer R, Briesen I, Emsermann M, Efremova N, Yephremov A, Schreiber L (2009) The DAISY gene from Arabidopsis encodes a fatty acid elongase condensing enzyme involved in the biosynthesis of aliphatic suberin in roots and the chalaza-micropyle region of seeds. Plant J 57(1):80–95. https://doi.org/10.1111/j.1365-313X.2008.03674.x
Gan L, Zhu S, Zhao Z, Liu L, Wang X, Zhang Z (2017) Wax CrystalSparse Leaf 4, encoding a β-ketoacyl-coenzyme A synthase 6, is involved in rice cuticular wax accumulation. Plant Cell Rep 36 (10):1655–1666. https://doi.org/10.1007/s00299-017-2181-5
Gao X, Zhang Y, Zhang H, Huang J (2022) A β-ketoacyl-CoA synthase OsCUT1 confers increased drought tolerance in rice. Rice Sci 29(4):353–362. https://doi.org/10.1016/j.rsci.2021.12.009
Haslam TM, Kunst L (2013) Extending the story of very-long-chain fatty elongation. Plant Sci 210:93–107. https://doi.org/10.1016/j.plantsci.2013.05.008
Haslam T, Haslam R, Thoraval D, Pascal S, Delude C, Domergue F, Fernández A, Beaudoin F, Napier J, Kunst L, Joubès J (2015) ECERIFERUM2-LIKE proteins have unique biochemical and physiological functions in very-longchain fatty acid elongation. Plant Physiol 167(3):682–692. https://doi.org/10.1104/pp.114.253195
Hegebarth D, Buschhaus C, Joubès J, Thoraval D, Bird D, Jetter R (2017) Arabidopsis ketoacyl-CoA synthase 16 (KCS16) forms C36/C38 acyl precursors for leaf trichome and pavement surface wax. Plant Cell Environ 40(9):1761–1776. https://doi.org/10.1111/pce.12981
Huai D, Xue X, Li Y, Wang P, Li J, Yan L, Chen Y, Wang X, Liu N, Kang Y, Wang Z, Huang Y, Jiang H, Lei Y, Liao B (2020) Genome-wide identification of peanut KCS genes reveals that AhKCS1 and AhKCS28 are involved in regulating VLCFA contents in seeds. Front Plant Sci 11:406. https://doi.org/10.3389/fpls.2020.00406
Huang H, Ayaz A, Zheng M, Yang X, Zaman W, Zhao H, Lü S (2022) Arabidopsis KCS5 and KCS6 play redundant roles in wax synthesis. Int J Mol Sci 23(8):4450. https://doi.org/10.3390/ijms 23084450
Huang H, Yang X, Zheng M, Chen Z, Yang Z, Wu P, Jenks M, Wang G, Feng T, Liu L, Yang P, Lv S, Zhao H (2023a) An ancestral role for 3-KETOACYL-COA SYNTHASE3 as a negative regulator of plant cuticular wax synthesis. Plant Cell 35(6):2251–2270. https://doi.org/10.1093/plcell/koad051
Huang H, Yang X, Zheng M, Lv S, Zhao H (2023b) Fine-tuning the activities of β-KETOACYL-COA SYNTHASE 3 (KCS3) and KCS12 in Arabidopsis is essential for maintaining cuticle integrity. J Exp Bot 74(21):6575–6587. https://doi.org/10.1093/jxb/erad337
Islam M, Du H, Ning J, Ye H, Xiong L (2009) Characterization of Glossy1-homologous genes in rice involved in leaf wax accumulation and drought resistance. Plant Mol Biol 70(4):443–456. https://doi.org/10.1007/s11103-009-9483-0
Jenks M, Joly R, Peters P, Rich P, Axtell J, Ashworth E (1994) Chemically induced cuticle mutation affecting epidermal conductance to water vapor and disease susceptibility in sorghum bicolor (L.) moench. Plant physiol 105(4):1239–1245. https://doi.org/10.1104/pp.105.4.1239
Joubès J, Raffaele S, Bourdenx B, Garcia C, Laroche-Traineau J, Moreau P, Domergue F, Lessire R (2008) The VLCFA elongase gene family in Arabidopsis thaliana: phylogenetic analysis, 3D modelling and expression profiling. Plant Mol Biol 67(5):547–566. https://doi.org/10.1007/s 11103-008-9339-z
Jung K, Han M, Lee D, Lee Y, Schreiber L, Franke R, Faust A, Yephremov A, Saedler H, Kim Y, Hwang I, An G (2006) Wax-deficient anther1 is involved in cuticle and wax production in rice anther walls and is required for pollen development. Plant Cell 18(11):3015–3032. https://doi.org/10.1105/tpc.106.042044
Kim J, Jung J, Lee S, Go Y, Kim H, Cahoon R, Markham J, Cahoon E, Suh M (2013) Arabidopsis 3-ketoacyl-coenzyme a synthase 9 is involved in the synthesis of tetracosanoic acids as precursors of cuticular waxes, suberins, sphingolipids, and phospholipids. Plant Physiol 162(2):567–580. https://doi.org/10.1104/pp.112.210450
Kim J, Lee S, Suh M (2021) Arabidopsis 3-ketoacyl-CoA synthase 4 is essential for root and pollen tube growth. J Plant Biol 64(2):155–165. https://doi.org/10.1007/s12374-020-09288-w
Lee S, Jung S, Go Y, Kim H, Kim J, Cho H, Park O, Suh M (2009) Two Arabidopsis 3-ketoacyl CoA synthase genes, KCS20 and KCS2/DAISY, are functionally redundant in cuticular wax and root suberin biosynthesis, but differentially controlled by osmotic stress. Plant J 60(3):462–475. https://doi.org/10.1111/j.1365-313X.2009.03973.x
Lee S, Suh M (2013) Recent advances in cuticular wax biosynthesis and its regulation in Arabidopsis. Mol Plant 6(2):246–249. https://doi.org/10.1093/mp/sss159
Li H, Pinot F, Sauveplane V, Werck-Reichhart D, Diehl P, Schreiber L, Franke R, Zhang P, Chen L, Gao Y, Liang W, Zhang D (2010) Cytochrome P450 family member CYP704B2 catalyzes the ω-hydroxylation of fatty acids and is required for anther cutin biosynthesis and pollen exine formation in rice. Plant Cell 22(1):173–190. https://doi.org/10.1105/tpc.109.070326
Liu J, Li L, Xiong Z, Robert C, Li B, He S, Chen W, Bi J, Zhai G, Guo S, Zhang H, Li J, Zhou S, Zhang X, Song C (2024) Trade-offs between the accumulation of cuticular wax and jasmonic acid-mediated herbivory resistance in maize. J Integr Plant Biol 66(1):143–159. https://doi.org/10.1111/jipb.13586
Ma K, Han J, Zhang Z, Li H, Zhao Y, Zhu Q, Xie Y, Liu Y, Chen L (2021) OsEDM2L mediates m⁶A of EAT1 transcript for proper alternative splicing and polyadenylation regulating rice tapetal degradation. J Integr Plant Biol 63(11):1982–1994. https://doi.org/10.1111/jipb.13167
Millar A, Clemens S, Zachgo S, Giblin E, Taylor D, Kunst L (1999) CUT1, An Arabidopsis gene required for cuticular wax biosynthesis and pollen fertility, encodes a very-long-chain fatty acid condensing enzyme. Plant Cell 11(5):825–838. https://doi.org/10.1105/tpc.11.5.825
Qin B, Lu T, Xu Y, Shen W, Liu F, Xie X, Li Y, Wang K, Li R (2021) Identification and characterization of An-4, a potential quantitative trait locus for awn development in rice. BMC Plant Biol 21:298. https://doi.org/10.1186/s12870-021-03055-w
Qin B, Tang D, Huang J, Li M, Wu X, Lu L, Wang K, Yu H, Chen J, Gu M, Cheng Z (2011) Rice OsGL1-1 is involved in leaf cuticular wax and cuticle membrane. Mol Plant 4(6):985–995. https://doi.org/10.1093/mp/ssr028
Qin P, Tu B, Wang Y, Lu C, Quilichini T, Li T, Wang H, Ma B, Li S (2013) ABCG15 encodes an ABC transporter protein, and is essential for post-meiotic anther and pollen exine development in rice. Plant Cell Physiol 54(1):138–154. https://doi.org/10.1093/pcp/pcs162
Shi J, Tan H, Yu X, Liu Y, Liang W, Ranathunge K, Franke R, Schreiber L, Wang Y, Kai G, Shanklin J, Ma H, Zhang D (2011) Defective pollen wall is required for anther and microspore development in rice and encodes a fatty acyl carrier protein reductase. Plant Cell 23(6):2225–2246. https://doi.org/10.1105/tpc.111.087528
Shi J, Cui M, Yang L, Kim YJ, Zhang D (2015) Genetic and biochemical mechanisms of pollen wall development. Trends Plant Sci 20(11):741–753. https://doi.org/10.1016/j.tplants.2015.07.010
Shim S, Mahong B, Lee S, Kongdin M, Lee C, Kim Y, Qu G, Zhang D, Ketudat Cairns J, Jeon J (2022) Rice β-glucosidase Os12BGlu38 is required for synthesis of intine cell wall and pollen fertility. J Exp Bot 73(3):784–800. https://doi.org/10.1093/jxb/erab439
Song Y, Tang Y, Liu L, Xu Y, Wang T (2022) The methyl-CpG-binding domain family member PEM1 is essential for Ubisch body formation and pollen exine development in rice. Plant J 111(5):1283–1295. https://doi.org/10.1111/tpj.15887
Todd J, Post-Beittenmiller D, Jaworski JG (1999) KCS1 encodes a fatty acid elongase 3-ketoacyl-CoA synthase affecting wax biosynthesis in Arabidopsis thaliana. Plant J 17(2):119–130. https://doi.org/10.1046/j.1365-313x.1999.00352.x
Tsuda K, Akiba T, Kimura F, Ishibashi M, Moriya C, Nakagawa K, Kurata N, Ito Y (2013) ONION2 fatty acid elongase is required for shoot development in rice. Plant Cell Physiol 54(2):209–217. https://doi.org/10.1093/pcp/pcs169
Wu Y, Guan Y, Wu Z, Yang K, Lv J, Converse R, Huang Y, Mao J, Zhao Y, Wang Z, Min H, Kan D, Zhang Y (2014) OsABCG15 encodes a membrane protein that plays an important role in anther cuticle and pollen exine formation in rice. Plant Cell Rep 33:1881–1899. https://doi.org/10.1007/s00299-014-1666-8
Xiao G, Wang K, Huang G, Zhu Y (2016) Genome-scale analysis of the cotton KCS gene family revealed a binary mode of action for gibberellin A regulated fiber growth. J Integr Plant Biol 58(6):577–589. https://doi.org/10.1111/jipb.12429
Xie X, Ma X, Zhu Q, Zeng D, Li G, Liu Y (2017) CRISPR-GE: A Convenient Software Toolkit for CRISPR-Based Genome Editing. Mol Plant 10:1246–1249. https://doi.org/10.1016/j.mol p.2017.06.004
Xu D, Shi J, Rautengarten C, Yang L, Qian X, Uzair M, Zhu L, Luo Q, An G, Waßmann F, Schreiber L, Heazlewood J, Scheller H, Hu J, Zhang D, Liang W (2017) Defective pollen wall 2 (DPW2) encodes an acyl transferase required for rice pollen development. Plant Physiol 173(1):240–255. https://doi.org/10.1104/pp.16.00095
Xue Z, Xu X, Zhou Y, Wang X, Zhang Y, Liu D, Zhao B, Duan L, Qi X (2018) Deficiency of a triterpene pathway results in humidity-sensitive genic male sterility in rice. Nat Commun 9(1):604. https://doi.org/10.1038/s41467-018-03048-8
Yang Z, Zhang Y, Sun L, Zhang P, Liu L, Yu P, Xuan D, Xiang X, Wu W, Cao L, Cheng S (2018) Identification of cyp703a3-3 and analysis of regulatory role of CYP703A3 in rice anther cuticle and pollen exine development. Gene 649:63–73. https://doi.org/10.1016/j.gene.2018.01.058
Yang L, Fang J, Wang J, Hui S, Zhou L, Xu B, Chen Y, Zhang Y, Lai C, Jiao G, Sheng Z, Wei X, Shao G, Xie L, Wang L, Chen Y, Zhao F, Hu S, Hu P, Tang S (2023) Genome-wide identification and expression analysis of 3-ketoacyl-CoA synthase gene family in rice (Oryza sativa L.) under cadmium stress. Front Plant Sci 14:1222288. https://doi.org/10.3389/fpls.2023.1222288
Yu D, Ranathunge K, Huang H, Pei Z, Franke R, Schreiber L, He C (2008) Wax Crystal-Sparse Leaf 1 encodes a β–ketoacyl CoA synthase involved in biosynthesis of cuticular waxes on rice leaf. Planta 228:675–685. https://doi.org/10.1007/s00425-008-0770-9
Zhang D, Liang W, Yin C, Zong J, Gu F, Zhang D (2010) OsC6, encoding a lipid transfer protein, is required for postmeiotic anther development in rice. Plant Physiol 154(1):149–162. https://doi.org/10.1104/pp.110.158865
Zhang D, Luo X, Zhu L (2011) Cytological analysis and genetic control of rice anther development. J Genet Genomics 38(9):379–390. https://doi.org/10.1016/j.jgg.2011.08.001
Zhang X, Wang Y, Wang X, Zhu Z, Zhang X, Jia L, Li Y, Tian W, Chen H, Zhu X, He G, Sang X (2022) A very-long-chain fatty acids synthesis gene, SD38, influences plant height by activating ethylene biosynthesis in rice. Plant J 112(4):1084–1097. https://doi.org/10.1111/tpj.15998
Zhou L, Ni E, Yang J, Zhou H, Liang H, Li J, Jiang D, Wang Z, Liu Z, Zhuang C (2013) Rice OsGL1-6 is involved in leaf cuticular wax accumulation and drought resistance. PLoS ONE 8(5):e65139. https://doi.org/10.1371/journal.pone.0065139
Zhou X, Li X, Xiang J, Gao G, Xu F, Liu A, Zhang X, Peng Y, Chen X, Wan X (2015) OsGL1-3 is involved in cuticular wax biosynthesis and tolerance to water deficit in rice. PLoS ONE 10(1):e116676. https://doi.org/10.1371/journal.pone.0116676
Zou T, Liu M, Xiao Q, Wang T, Chen D, Luo T, Yuan G, Li Q, Zhu J, Liang Y, Deng Q, Wang S, Zheng A, Wang L, Li P, Li S (2018) OsPKS2 is required for rice male fertility by participating in pollen wall formation. Plant Cell Rep 37(5):759–773. https://doi.org/10.1007/s00299-018-2265-x
Zhang Z, Lu Y, Feng J, Liu X, Zhang G (2004) Studies on the anther dehiscence in F₁ of hybrids between Taichung 65 and its F₁ pollen sterility near isogenic lines. J Trop Subtrop Bot 12(6):521–527. https://doi.org/10.3969/j.issn.1005-3395.2004.6.006
Zhu L, Guo J, Zhu J, Zhou C (2014) Enhanced expression of EsWAX1 improves drought tolerance with increased accumulation of cuticular wax and ascorbic acid in transgenic Arabidopsis. Plant Physiol Biochem 75:24–35. https://doi.org/10.1016/j.plaphy.2013.11.028

Download PDF

Reviewers agreed at journal
04 Aug, 2024
Reviewers invited by journal
03 Aug, 2024
Editor assigned by journal
30 Jul, 2024
First submitted to journal
29 Jul, 2024

You are reading this latest preprint version

A β-Ketoacyl-CoA Synthase encoded by DDP1 controls rice anther dehiscence and pollen fertility by maintaining lipid homeostasis in the tapetum

Status:

Version 1

Abstract

Figures

Key Message

Introduction

Materials and methods

Plant materials and growth conditions

Characterization of mutant phenotypes

Examination of anther dehiscence

Histochemical analysis

Wax components analysis

Protein sequence analysis

Subcellular localization

Results

Expression pattern and subcellular location of DDP1

Discussion

Declarations

References

Supplementary Files

Status:

Version 1