Onion (Allium cepa L.) is one of the important horticultural crops cultivated throughout the world. It is known for its distinct flavor and pungency, which are attributed to a variety of sulfur compounds. Onions are beneficial in terms of offering health benefits due to the presence of flavonoids, anthocyanins, fructooligosaccharides, and organosulfur compounds (Goldman, 2011; Bisen and Emerald, 2016; Putnik et al., 2019). Since domestication, productivity, quality, and resilience to biotic and abiotic stresses have been improved in onions using several conventional breeding practices. However, conventional approaches have not produced substantial progress in the genetic improvement of onions and have been hampered by the biennial life cycle, cross-pollinated nature, and severe inbreeding depression (McCallum, 2007; Sudha et al., 2019; Lee et al., 2021). Despite the fact that onions have a large genetic variability, crop improvement has lagged behind other monocot crops (McCallum, 2007; Varshney et al., 2012). The burden of the current global challenges involving food security, increasing human population and changes in climate experienced worldwide lies in the need to improve horticultural crops with high yield and enhanced adaptation to the environment, for which conventional breeding is unlikely to meet the demand (Brewster, 2008; Van den Broeck and Maertens, 2016). To overcome these obstacles, advanced and innovative molecular approaches should be strategically employed to strengthen onion genetic improvement programs.

Several genome editing tools, such as transcription activator-like effector nucleases (TALENs), zinc-finger nucleases (ZFNs), and meganucleases (MNs), have facilitated targeted manipulation of a desired gene in crop plants. However, these methods are expensive, labor-intensive, and require intricate procedures to achieve successful editing (Razzaq et al., 2019). In contrast, CRISPR/Cas9 has emerged as a promising and versatile approach for precise and efficient plant genome editing. It offers an appealing advantage due to its user-friendly design, cost-effectiveness, and adaptability, making it an attractive tool for researchers in the field (Razzaq et al., 2019). The advent of CRISPR/Cas9-targeted genome editing has paved the way toward incorporating beneficial genetic changes for addressing limitations affecting production and utilization. CRISPR/Cas9 genome editing, as a popular genome engineering tool useful for crop improvement, has drawn the attention of breeders and stands out as an essential tool to generate plants with the ability to cope with challenges. The CRISPR/Cas9 system has been applied to a large number of plant species. Still, there are several horticultural crops where the applicability of this technology has not been demonstrated (Koltun et al., 2018), and onion is one among them. A reproducible system for the design, construction, and application of Cas9/guide RNA (gRNA) must be established and evaluated in onions in order to assess the potential of the CRISPR/Cas9 system. The strategy to knock out the Phytoene desaturase (PDS) gene has been widely used to validate the practicability of the CRISPR/Cas9 system in several plants, including cassava (Manihot esculenta) (Odipio et al., 2017), banana cv. Rasthali (Musa acuminata cv. Rasthali) (Kaur et al., 2018), and Cavendish banana (Musa acuminata subgroup Cavendish) (Naim et al., 2018). Only recently has the onion genome sequence become available (Finkers et al., 2021), thereby enabling the exploration of the potential for CRISPR/Cas9-based genome editing in applied research for this crop.

In the present study, for the first time, we report the successful establishment of a CRISPR/Cas9-based genome editing system by targeting the gene coding for PDS of the Indian short-day onion. The insights presented here lay the groundwork for future investigations into gene function and trait development in this important vegetable crop, leveraging the capabilities of the CRISPR/Cas9-based gene editing system.

The CRISPR/Cas9 technology has so far been used in numerous plant species for basic research and trait development. Apart from the evaluation of gene functions, the application of CRISPR/Cas9 technology has great potential in onion for manipulation of a metabolic pathway to alter flavor profile, flowering behavior, quality, and traits of agronomic importance for increasing yield potential. Though onion is one of the most important vegetable crops throughout the world, the study for CRISPR/Cas9-mediated gene editing in onion has not been conducted yet. To the best of our knowledge, the present study is the first report on CRISPR/Cas9-mediated genome editing in onion.

The PDS gene is involved in the carotenoid biosynthesis pathway, and its loss-of-function mutation leads to the development of the photobleached or albino phenotype, which can be easily and quickly detected by naked eye and does not require any biochemical, physiological, or microscopic assays. Therefore, PDS has been widely used as a common visual marker to validate the practicability of genome editing by CRISPR/Cas9 in several plant species, viz. populus (Populus angustifolia) (Fan et al., 2015), apple (Malus prunifoli) (Nishitani et al., 2016), grape (Vitis vinifera) (Nakajima et al., 2017), cassava (Manihot esculenta) (Odipio et al., 2017), banana cv. Rasthali (Musa acuminata cv. Rasthali) (Kaur et al., 2018), cavendish banana (Musa acuminata subgroup Cavendish) (Naim et al., 2018), melon (Cucumis melo) (Hooghvorst et al., 2019), strawberry (Fragaria ananassa) (Wilson et al., 2019), lettuce (Lactuca sativa) (Henderson et al., 2020), poplar (Populus tomentosa Carr) (Fan et al., 2015; Wang et al., 2020), and yam (Dioscorea rotunda) (Syombua et al., 2021). Thus, we targeted the AcPDS gene to evaluate the editing efficiency induced by the CRISPR/Cas9 system.

A detailed analysis of the targeted gene and its putative functions is a prerequisite for any genome editing studies. We retrieved the AcPDS sequence from the draft onion genome based on a homology search. As the B. Super cultivar was used for the genome editing study, the gene was additionally isolated from the Indian onion cv. B. Super and afterward employed for designing and constructing targets for genome editing studies. PDS has been reported as a single copy gene in many crop species like melon (Hooghvorst et al., 2019), cassava (Odipio et al., 2017), and cavendish banana (Naim et al., 2018). Likewise, BLASTn analysis also showed a single copy of the PDS gene in the draft genome sequence of onion. Furthermore, in-silico analysis of the AcPDS protein sequence from B. Super cultivar showed that it contained the characteristic signatures of the conserved amino-oxidase domain known from other crops (Li et al., 2013b; Kaur et al., 2017). Therefore, we targeted the 5′ region of a conserved amino-oxidase domain of AcPDS to ensure functional knockdown upon mutation. Moreover, we selected two gRNA targets in the amino-oxidase domain, i.e., in exon 3 and exon 4, to improve our chances of success and to evaluate any large deletions in addition to InDels, as they were used in other crops (Odipio et al., 2017; Ntui et al., 2020; Lu and Tian, 2022).

Onion is recalcitrant to transformation, has very low transformation efficiency (i.e., 0.2% to 2.5%), and transformation is highly genotype-dependent (Eady et al., 2000; Zheng et al., 2001; Eady et al., 2003; Zheng et al., 2005; Kamata et al., 2011; Naini et al., 2019; Manape et al., 2022). There are two reported techniques for the genetic transformation of onions: i.e., biolistic transformation and Agrobacterium-mediated transformation. A biolistic method is challenging to carry out within crop plants and requires expensive consumables such as gold particles or specialized machinery (e.g., gene gun). On the other hand, Agrobacterium-mediated is inexpensive, does not require any specialized equipment, and has been standardized in our laboratory for the onion cv. B. Super with 1% transformation efficiency (Manape et al., 2022). Therefore, we employed the same protocol to study the CRISPR/Cas9-based genome editing in the same onion cultivar and regenerated six Cas9 PCR-positive shoots out of 617 co-cultivated calli with a comparable transformation efficiency of 0.97%.

In general, putatively transformed calli are selected on growth medium by using either negative agents, namely, kanamycin, hygromycin B, and spectinomycin, or positive ones like mannose (Bibi et al., 2013; Rastogi et al., 2018). Negative selection can inhibit and delay shoot growth, cause chlorosis in shoots, and produce pale green phenotypes, as observed in crops like sugarcane (Rastogi et al., 2018) and cotton (Bibi et al., 2013). The pale green phenotypes were also observed due to Cas9-induced InDels in the PDS gene of the Cavendish banana (Naim et al., 2018) and hybrid poplar (Wang et al., 2020). This necessitates the screening of pale green shoots to differentiate between non-transformed and transformed shoots, which is a laborious and time-consuming process. In the onion variety B. Super, embryogenic calli induce shoots within 3 weeks on a shooting medium devoid of hygromycin B, whereas transformed calli take 6–8 weeks in the presence of hygromycin B, occasionally producing shoots with chlorotic effects (unpublished data). However, using a positive selection method requires optimization of the selection agent concentration, which is also labor- and time-intensive. Given that the regeneration protocol was already standardized for onion cv. B. Super, we transferred co-cultivated calli directly to a shooting medium without hygromycin B. This approach eliminated confusion over pale green shoots and expedited the regeneration protocol for quick validation of CRISPR/Cas9-mediated genome editing. Our study successfully regenerated 21 shoots with varying degrees of albino or chlorophyll loss within 7–8 weeks using Agrobacterium-mediated co-transformation of embryogenic calli of onion with two independent AcPDS targeting CRISPR/Cas9 constructs. This is significantly faster than the 22–24 weeks required by the transformation protocol that we had developed earlier (Manape et al., 2022). Of these 21 shoots, two were complete albinos, four were chimeric, and 15 were pale green. Similar phenotypes have been reported in PDS knockout experiments in various plant species (Pan et al., 2016; Naim et al., 2018; Hooghvorst et al., 2019; Henderson et al., 2020; Wang et al., 2020; Lu and Tian, 2022).

To ascertain genome editing in onion, PCR analysis was performed with a flanking primer set in six Cas9-positive regenerated shoots, and multiple clones of each were sequenced on Sanger’s platform to investigate mutations within and around one or both target sites. Sanger sequencing results revealed InDels and substitutions in and around the target sites of both sgRNAs in the albino phenotypes. A similar type of mutation has also been reported in other plants, including Arabidopsis (Arabidopsis thaliana) (Feng et al., 2014), apple (Malus prunifoli) (Nishitani et al., 2016), and Rasthali banana (Musa acuminata cv. Rasthali) (Kaur et al., 2018). The InDels recorded here could have resulted from a nonhomologous end-joining repair (NHEJ) pathway following the integration of T-DNA (Fan et al., 2015; Nishitani et al., 2016).

InDels are the most common mutations in CRISPR/Cas9-mediated gene editing, induced by NHEJ. The evaluation of the Sanger sequencing data, using multiple sequence alignment and the Synthego ICE Analysis tool, was found to be consistent with the results from NGS data. Both sets of data indicated similar InDels in the vicinity of the target sgRNAs. NGS data showed a high level (63.4% in exon 3 sgRNA) to a low level (1.1% in exon 4 sgRNA) of InDel frequencies. This clearly demonstrated a significant difference in the editing efficiency of exon 3 and exon 4 sgRNAs. The GC content of the target sgRNA sequence and secondary structure of targeted sgRNAs affect the editing efficiency of Cas9 (Wang et al., 2014; Zhang et al., 2014; Ma et al., 2015). Both the sgRNAs had 50% GC content. Therefore, we analyzed the secondary structures of both target sgRNAs. Analyzing the secondary structure revealed that exon 3 target sgRNA had pairings of only 2 to 4 bp (i.e., continuous pairing less than 5 bp) in the target sgRNA sequence. However, exon 4 target sgRNA had a typical stem-loop structure with a pairing of continuous 7 and 4 bp of the target sgRNA sequence ( Supplementary Figure S7 ). Secondary structures of target-sgRNAs with more than continuous 6 bp pairing of the target sgRNA sequence with stem-loop structures had significantly low editing efficiency, possibly due to inhibition of binding of the sgRNA to the target strand attributed to the formation of stem-loop structures (Ma et al., 2015).

Along with InDels, we also found a high frequency (54.84%) of substitutions in the exon 3 sgRNA sequence. Similar findings on the high level of substitutions mediated by CRISPR/Cas9 have been made by other studies. For example, in cotton (Gossypium hirsutum L.) protoplasts, only substitutions were found (Chen et al., 2017); in melon protoplasts and plants, 91% of mutations were substitutions (Hooghvorst et al., 2019); in cassava plants, substitutions occurred more frequently than InDel mutations (Odipio et al., 2017); in soybean (Glycine max L.) protoplasts and plants, a high level of substitutions were reported in the Cas9-edited events (Sun et al., 2015); and in rice (Oryza sativa L.), 25%–45% of the examined mutations were substitutions (Macovei et al., 2018). Observation of nucleotide substitution downstream of the target region implies activation of the homologous recombination (HR) pathway to repair double-strand breaks created in AcPDS. Simultaneous triggering of HR and NHEJ repair pathways proceeding integration of CRISPR/Cas9 gene T-DNA was documented earlier in apple (Malus prunifoli), Arabidopsis (Arabidopsis thaliana), tobacco (N. benthamiana), cassava (Manihot esculenta), and cotton (Gossypium hirsutum) (Li et al., 2013a; Nishitani et al., 2016; Gao et al., 2017; Odipio et al., 2017).

Four regenerated chimeric shoots showed the presence of the Cas9 gene, but no editing was detected in and around the target sites when Sanger sequencing results were analyzed. However, NGS data revealed additional InDels and substitutions, albeit at very low frequencies. As we sequenced only a few clones from each chimeric shoot using the Sanger platform, it is likely that we were not able to detect these very low frequencies of InDels and substitutions, which potentially could have been induced by transient expression of the Cas9 gene and sgRNA from the transformed construct. Bleached patches of tissue, similar to our observations, were reported in a study on yam, and they were attributed to the transient expression of Cas9 and PDS-targeted sgRNA (Syombua et al., 2021). While this is a possibility in our study, another plausible explanation could stem from the use of a non-selective medium leading to the regeneration of chimeric transformants.

In our study, out of the six Cas9-positive shoots identified, two demonstrated CRISPR/Cas9-mediated genome editing in the AcPDS gene, yielding an editing efficiency of 33.33%. Mutation efficiency in monocotyledons using 35S promoter-driven Cas9 constructs is 1.8%–81% (Basu et al., 2023). Therefore, mutation efficiency is moderate and can be improved in future studies. We used the pRGEB31 vector, which contains the cauliflower mosaic virus (CaMV35S) and Oryza sativa U3 promoter for driving the expression of Cas9 and sgRNAs, respectively. Optimal CRISPR/Cas9 function requires high transcription levels of both the sgRNA and Cas9 proteins (Mao et al., 2016). Editing efficiency is significantly influenced by the choice of promoter driving Cas9, with endogenous constitutive promoters yielding high transgene expression and, consequently, higher mutation efficiencies (Hassan et al., 2021). Similarly, the endogenous U6 promoter has been shown to enhance sgRNA expression and thus the overall CRISPR/Cas9-mediated editing efficiency by up to 87% (Yan et al., 2015; Zhang et al., 2018). As the promoters used in our study were not endogenous, it may have contributed to the less effective performance of the CRISPR/Cas9 system. Therefore, the utilization of endogenous promoters for Cas9 and sgRNA expression could be a more effective strategy for boosting CRISPR/Cas9 efficiency in onions.

We estimated the chlorophyll content in the regenerated albino, chimeric, pale green, and non-transformed control shoots to investigate the impact of AcPDS mutations. We observed a significant reduction in chlorophyll content in albino and chimeric shoots. InDels produced frameshift mutations and ectopic stop codons in the AcPDS gene of albino shoots. These alterations led to the early termination of AcPDS protein, leading to albinism; consistent with findings in Rasthali banana (Kaur et al., 2018). The variegation and reduced chlorophyll content in chimeric phenotypes could be the result of a transient expression of Cas9-mediated alterations in AcPDS. Similar observations were reported in yam, where bleached patches on shoots were linked to Agrobacterium-mediated transient expression of Cas9 and PDS sgRNA (Syombua et al., 2021). The nonsignificant reduction in chlorophyll content in the 15 pale green calli may be attributed to somaclonal variation induced by frequent culturing on the shooting medium. Further detailed analysis is required to confirm this hypothesis. In summary, the substantial decrease in chlorophyll content in albino shoots indicates a functional knockout of the AcPDS gene. This disruption likely affects both the chlorophyll and gibberellic acid biosynthesis pathways. The slow growth rate observed in albino plants may be due to the disruption of gibberellic acid biosynthesis, which hinders their growth and development. This is supported by a study in Arabidopsis, where the application of exogenous GA₃ partially reversed the dwarf phenotypes of AcPDS mutants (Qin et al., 2007).

In conclusion, the CRISPR/Cas9-mediated AcPDS knockout showcased in this study produced a readily observable mutant phenotype within a rapid timeframe of 8-weeks. This strategy provides a valuable tool for accelerating the evaluation and optimization of CRISPR/Cas9-mediated genome editing in onions, ultimately contributing to the improvement of their traits. While the observed editing efficiency in our study was moderate, the use of an endogenous promoter-driven Cas9 and sgRNA could potentially enhance it in future applications.

The data presented in the study are deposited in the NCBI database repository, accession number PRJNA976681 (http://www.ncbi.nlm.nih.gov/bioproject/976681). The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

PM, TKM, and SA conceived and designed the experiments. PM and TKM conducted all the experiments. PM, TKM, and SA drafted the manuscript. SA and VS carried out the data analysis and manuscript editing. SA supervised the entire work. All authors contributed to the article and approved the submitted version.