Haploid induction in sweet potato by activating the <scp>AP2</scp>/<scp>ERF</scp> family transcription factor <scp><i>IbBBM</i></scp>

Sweet potato (Ipomoea batatas L.) is a major root crop in the world. As an asexual crop, sweet potato breeding and basic research face a number of challenges due to the highly heterozygous genome. Doubled haploid (DH) breeding is one of the most efficient means of improving crop germplasm. However, there is no reliable haploid induction system for sweet potato. BABY BOOM (BBM) encodes an AINTEGUMENTA-LIKE (AIL) APETALA2/ethylene-responsive factor (AP2/ERF) domain transcription factor (Boutilier et al., 2002). It is a key regulator of plant cell totipotency and induces asexual embryo formation when ectopically expressed (Chen et al., 2022). This property has been widely exploited to improve plant transformation and clonal propagation. Nevertheless, the ability of the BBM homologue to induce parthenogenesis and produce viable haploid plants in sweet potato remains to be investigated. To identify a candidate for a sweet potato BBM homologue, a tblastn analysis was performed against the sweet potato genome (Wu et al., 2024; Yang et al., 2017; Yoon et al., 2022) using the Arabidopsis BBM protein sequence (AT5G17430) as a query. The most likely candidates were then used for a phylogenetic study. Phylogenetic analysis revealed that candidate Ibat.Tzn_v2.02EG002520.1 (termed IbBBM) is the closest homologue of AtBBM and BnBBMs (Figures 1a and S1). Due to the highly heterogeneous nature of the sweet potato genome (2n = B1B1B2B2B2B2 = 6x = 90), sequence variations among alleles of the same gene are common, making it difficult to choose which allele to clone and overexpress. CRISPR activation (CRISPRa), based on deactivated Cas (dCas) proteins coupled to activator domains, offers a promising alternative to the conventional approach of gene overexpression (Konermann et al., 2015). Although CRISPRa technologies are now available in plants, the applications of transcriptional activation in dicotyledonous plants, especially sweet potato, are still scarce. Firstly, we developed a CRISPR-dCas9-6×TAL-2×VP64 system capable of inducing target gene activation in sweet potato (Figure 1b). Next, we designed a single-guide RNA (sgRNA) specific for a conserved region upstream of the IbBBM start codon (ATG) and introduced the construct into the sweet potato variety Xushu27 via Agrobacterium-mediated transformation (Figure 1c; Figure S2). Three IbBBM-CRISPRa lines were transgenic positive and showed high levels of IbBBM expression (Figure 1d,e). Compared to the wild type (WT), the transgenic positive lines did not show any obvious changes in flowering time, vine length, and stem base thickness, but they produced more seeds (Figure 1f,g; Figure S3). A total of 49 seeds were harvested from the IbBBM-CRISPRa transgenic-positive lines, whereas only 13 were harvested from the WT and the negative lines (Figure 1g; Table S1). Sweet potato is self-incompatible, so more seeds produced in transgenic plants may be the result of induced parthenogenesis following IbBBM activation. Furthermore, the harvested seeds were germinated to assess the presence of haploids in the progeny. All seed-germinated plants except two (#4–7 and #6–10) showed typical hexaploid plant phenotypes. Compared to the hexaploid control (Mock), the main stems of plants #4–7 and #6–10 became slender and weak (Figure 1h). Previous studies have shown that genome size is positively related to guard cell size (Beaulieu et al., 2008). Microscopic observations of leaf epidermal cells revealed that the guard cell size was reduced in #4–7 and #6–10 (19.4 ± 1.3 μm) compared to Mock (27.6 ± 0.8 μm) (Figure 1i; Figure S4). Flow cytometry was then performed on all seed-germinated plants to confirm their ploidy level. As expected, the signal intensity values in #4–7 and #6–10 were approximately half those of the Mock (Figure 1j; Figure S5a). Furthermore, it was further confirmed that both #4–7 and #6–10 were haploid using the known hexaploid sweet potato variety Xushu 18 as a reference (Yoon et al., 2022) (Figure S5b,c). By contrast, no haploid plants were found among the germinated seeds produced by WT and the negative lines. Here, we propose a model for the induction of sweet potato haploid formation through the activation of IbBBM (Figure 1k). As an asexual crop, hybridization and selection remain the primary methods of sweet potato breeding. However, the varieties currently used in sweet potato production practices are mainly derived from inter-varietal hybrids. With the development and innovation of plant haploid induction technologies, haploid breeding will be widely used in the varietal improvement of sweet potato. Sweet potatoes can naturally flower and set seed south of latitude 23° N latitude but generally do not flower in northern China. In this study, we selected a variety, Xushu27, which flowers naturally in Beijing, and used constitutive activation of IbBBM. More locations, varieties, and tissue-specific activation methods will be tested in future studies. Altogether, these results provide a favourable reference for inducing efficient parthenogenesis and forming viable haploid plants in sweet potato. We acknowledge with apologies many relevant studies that could not be discussed or cited due to space limitations. This work was supported by the earmarked fund for CARS-10-Sweetpotato, the Beijing Food Crops Innovation Consortium Program (BAIC02-2025), and the 2115 Talent Development Program of China Agricultural University. The authors have declared no conflict of interest. S.G. conceived the idea and designed the experiments. W.S. performed most of the experiments and analyses. S.G. constructed the vector with contributions from J.Z. and Y.Z. Y.Z. created the transgenic materials and contributed to the field experiments. H.Z., H.Z., N.Z., S.H. and Q.L. contributed to the acquisition of funding and supervised the project. W.S. drafted the original manuscript. S.G. revised and finalized the manuscript. All authors read and approved the final version of the manuscript. All data in this study are available in this article or supplementary information. Figure S1 Multiple protein sequence alignment of AtBBM, IbBBM and IbPLT, with conserved amino acids shaded in different colours. Figure S2 Sanger sequencing chromatogram of the CRISPRa target site of the IbBBM promoter in wild-type Xushu27. Figure S3 Inflorescences, flowers and seeds of negative lines. Figure S4 Bar graph showing the length of stomatal guard cells in Mock and haploid plants. Figure S5 Flow-cytometric DNA histograms for ploidy determination. Table S1 Analysis of the number of days to flowering, vine length, stem diameter at the base and the total number of capsules, seeds and haploids in the negative lines. Table S2 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.