Efficient Haploid Induction Via Egg Cell Expression of Dandelion PARTHENOGENESIS in Foxtail Millet (setaria Italica)

Foxtail millet (Setaria italica) is one of the most significant multigrain crops in arid and semi-arid regions worldwide. Given its relatively small genome size, strong environmental adaptability, high C4 light efficiency and remarkable nutritional value, foxtail millet is emerging as a rapidly developing C4 model crop for research on drought resistance, nutrition efficiency, and photosynthesis (Peng and Zhang, 2021). Doubled haploid (DH) technology has the potential to revolutionize crop breeding efficiency by rapidly fixing recombinant haplotypes within just two generations compared to traditional breeding methods (Jacquier et al., 2020). However, the limited research on haploid induction systems in foxtail millet hampers the widespread application of DH technology in this crop. Here, we report the successful generation of haploid plants through seed by egg cell expression of dandelion PARTHENOGENESIS (PAR), providing the possibility for further advancements in DH technology utilization in foxtail millet breeding programs. Dandelion (Taraxacum officinale) is a well-known apomictic plant that predominantly reproduces through parthenogenesis. The recent discovery of the PAR gene has provided insights into the genetic mechanism underlying parthenogenesis in dandelions. This gene encodes a K2-2 zinc finger domain protein with an EAR (ethylene-responsive element-binding factor-associated amphiphilic repression) motif, exerting a dominant effect on apomixis in dandelion. Furthermore, when introducing the dandelion PAR gene into the sexual crop lettuce (Lactuca sativa) under the control of an egg cell-specific Arabidopsis EC1 promoter, it demonstrates the ability to trigger parthenogenesis and generate haploid embryo-like structures without fertilization (Underwood et al., 2022). However, PAR-induced parthenogenesis has, thus far, only been observed in dicot plants. It remains unknown whether PAR also plays a role in inducing haploids in monocots. To this end, we constructed a pAtDD45::ToPAR vector, expressing the Taraxacum PAR gene under the egg cell-specific Arabidopsis DD45 promoter, and introduced it into the foxtail millet landrace accession Ci846 using Agrobacterium-mediated genetic transformation (Figure 1a). Subsequently, 11 independent transgenic plants that heterologously express ToPAR, referred to as ToPAR-HE hereafter were generated and grown in controlled greenhouse conditions. Genotyping analysis confirmed the presence of heterologously expressed cassettes in all 11 T0 plants (Figure 1b). In comparison to the Ci846 wild type (plant height: 104 ± 11.7 cm; panicle length: 15.4 ± 1.9 cm), these transgenic plants exhibited a broader range of variation in both plant height (ranging from 53 cm to 169 cm) and panicle length (ranging from 8 cm to 23 cm) (Figure 1c,d, Figure S1). Additionally, we observed a significant decrease in the seed setting rate (ranging from 6.6% to 39.4%) of all transgenic lines compared to that of the Ci846 wild type (80.6 ± 6%) (Figure 1d). Given the dominant nature of PAR, all 11 ToPAR-HE lines were evaluated for parthenogenesis occurrence. T1 progeny originating from selfing of the 11 events (ToPAR-HE #1–ToPAR-HE #11), along with Ci846 wild-type plants, were grown, and ploidy levels were determined by flow cytometry. No haploid was found among 130 Ci846 wild-type plants, whereas among the T1 progeny of the 11 ToPAR-HE lines, six lines (ToPAR-HE #2, #4, #5, #8, #9 and #11) showed haploid induction, whereas the remaining transformants produced exclusively diploids. The average haploid induction rate (HIR) ranged from 1.4% to 10.2% (Figure 1d,e, Figure S2). All haploid S. italica plants exhibited typically haploid characteristics, such as short stature and smaller size in all organs, including panicles, grains and anthers, relative to the wild type and complete loss of fertility (Figure 1f). Furthermore, we investigated the phenotypes of both negative and positive plants in the T1 generation of ToPAR-HE #5, concurrently assessing the haploid induction rate (Figure S3). Our results indicate that the phenotype of the transgenic T0 generation can be stably inherited in the T1 generation. Overall, our findings demonstrate the feasibility of achieving haploid induction in the monocot crop foxtail millet through the heterologous expression of dandelion PAR in egg cells. In a previous study, Cheng et al. utilized the CRISPR/Cas9 system to disrupt the SiMTL gene, establishing a foxtail millet haploid induction line with a maximum HIR of 3.5% (Cheng et al., 2021). In our research, we implemented a novel strategy involving the heterologous expression of dandelion PAR in egg cells to effectively induce haploid production in foxtail millet. Notably, our approach yielded a significantly higher maximum HIR of 10.2%, surpassing the SiMTL knockout method. Additionally, up to 25.6% of embryo-like structures were observed in dicotyledonous lettuce following emasculation using a similar approach to the heterologous expression of dandelion PAR (Underwood et al., 2022). These findings indicate that the utilization of dandelion PAR-induced haploid production may represent a conservative and promising strategy for both monocotyledonous and dicotyledonous crops, demonstrating great potential for deploying DH technology in foxtail millet breeding programs. Furthermore, we have included a schematic diagram illustrating doubled haploid breeding through the heterologous expression of ToPAR in Figure S4. The observed variation in dandelion PAR-induced haploid production frequency in our study may potentially stem from positional effects of transgene insertion. Native egg cell-specific promoters or optimized transgenes might potentially achieve a more consistent and higher level of HIR in foxtail millet. Moreover, recent research has unveiled that integrating ToPAR-induced parthenogenesis with Mitosis instead of Meiosis (MiMe) enables the simultaneous production of high-fertility and high-frequency synthetic apomixis in hybrid rice (Song et al., 2024). This discovery provides valuable theoretical guidance for realizing efficient synthetic apomixis in foxtail millet. This work was supported by the National Natural Science Foundation of China (32188102, 32025028, 32101721), the National Key R&D Program of China (2022YFF1003304), the Key R&D Program of Zhejiang Province (2021C02063-6) and Hainan Seed Industry Laboratory (B23CJ0208). No conflict of interest is declared. K. W., Q. Q. and X. W. managed the project; Y. H., Y. L. and Y. X. performed the experiments and analysed the data; Y. H. wrote the manuscript; K. W., Q. Q. and X. W. revised the manuscript. The data that supports the findings of this study are available in the supplementary material of this article. Figure S1 The plant morphology and panicles of the remaining ToPAR heterologous expression T0 plants that can produce haploids. Figure S2 Flow cytometry of the remaining 15 haploids generated by the ToPAR heterologous expression lines. Figure S3 Comparison of phenotypes between negative (Neg) and positive (Pos) plants in the T1 generation of ToPAR-HE #5. Figure S4 Schematic diagram of doubled haploid breeding via the heterologous expression of ToPAR. Table S1 Primers used in this study. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.