OsCKq1 Regulates Heading Date and Grain Weight in Rice in Response to Day Length

Abstract

Background

Photoperiod sensitivity is among the most important agronomic traits of rice, as it determines local and seasonal adaptability and plays pivotal roles in determining yield and other key agronomic characteristics. By controlling the photoperiod, early-maturing rice can be cultivated to shorten the breeding cycle, thereby reducing the risk of yield losses due to unpredictable climate change. Furthermore, early-maturing and high-yielding rice needs to be developed to ensure food security for a rapidly growing population. Early-maturing and high-yielding rice should be developed to fulfill these requirements. OsCKq1 encodes the casein kinase1 protein in rice. OsCKq1 is a gene that is activated by photophosphorylation when Ghd7, which suppresses flowering under long-day conditions, is activated.

Results

This study investigates how OsCKq1 affects heading in rice. OsCKq1-GE rice was analyzed the function of OsCKq1 was investigated by comparing the expression levels of genes related to flowering regulation. The heading date of OsCKq1-GE lines was earlier (by about 3 to 5 days) than that of Ilmi (a rice cultivar, Oryza sativa spp. japonica), and the grain length, grain width, 1,000-grain weight, and yield increased compared to Ilmi. Furthermore, the culm and panicle lengths of OsCKq1-GE lines were either equal to or longer than those of Ilmi.

Conclusions

Our research demonstrates that OsCKq1 plays a pivotal role in regulating rice yield and photoperiod sensitivity. Specifically, under long-day conditions, OsCKq1-GE rice exhibited reduced OsCKq1 mRNA levels alongside increased mRNA levels of Hd3a, Ehd1, and RFT1, genes known for promoting flowering, leading to earlier heading compared to Ilmi. Moreover, we observed an increase in seed size. These findings underscore OsCKq1 as a promising target for developing early-maturing and high-yielding rice cultivars, highlighting the potential of CRISPR/Cas9 technology in enhancing crop traits.

Introduction

Rice is an important crop that is used as a staple food by more than half of the world’s population (My et al. 2021). However, ongoing rapid and unpredictable changes in global climate have contributed to reductions in rice yields in recent years (Guo et al. 2021). These yield reductions come against a backdrop of continual rapid world population growth (Baker et al. 2022), thereby exacerbating existing food shortages, and thus creating serious social problems and threatening food security (Mottaleb et al. 2021). To avert such food shortages, it is necessary to breed rice cultivars that can guarantee consistent yields that are at least twice those obtained cultivating current varieties (Chandio et al. 2022). Moreover, given current trends in desertification and urban development, the area of land suitable for crop cultivation continues to contract (Bestelmeyer et al. 2015; van Vliet 2019). Consequently, it is essential to devise cultivation methods and breed cultivars that can produce maximum yields in a limited production area. One promising approach in this regard is the breeding and cultivation of early-maturing rice cultivars that are conducive to harvesting using double-cropping systems, which could be established as a stable basis for achieving yield security (Hu et al. 2015). Currently, however, rice double-cropping systems tend to be associated with low yields, owing to the comparatively short growing periods, and the quality of grain is often poor due to poor growth (Krishnan et al. 2011). Consequently, to achieve maximal effects when utilizing a limited area of cultivatable land, it is necessary to breed rice cultivars that can be grown stably during the contracted growth period necessary for rice double-cropping systems. Most of the rice cultivars that are currently double-cropped are early-maturing cultivars, and these are typically small-panicle types, which contrasts with the generally large-panicle late-maturing rice cultivars (Kawasaki 2019). Recent yield reductions have been associated with an increasing frequency of typhoons and abnormally high temperatures that coincide with the rice filling stage (Arshad et al. 2017), which have detrimental effects with respect to grain quality (Sreenivasulu et al. 2015). Taking into account the aforementioned considerations, it is thus necessary to breed early-maturing rice cultivars that are characterized by a short growing period, large panicles, and high yields (Kawasaki 2019).

During the domestication of rice from its wild rice, the plant has evolved a range of different morphological and physiological traits that are conducive to commercial cultivation (Vaughan et al. 2008). In particular, the flowering control network is one of the most elaborately evolved elements of wild rice (Itoh et al. 2018; Lu et al. 2012). Given photoperiod sensitivity is a key agronomic trait that contributes to determining yield and grain quality (Yun et al. 2016), it represents a particularly important target trait for breeders. Flowering time is controlled by a complex network of numerous genes (Liu et al. 2014). It is affected to a considerable extent by seasonal changes in factors such as photoperiod, autonomy, and vernalization (Bernier et al. 1993; Kwon et al. 2014). Photoperiodic flowering, among the most important biological systems regulating floral transition (Wang et al. 2020), is regulated by light signals and endogenous plant circadian rhythms (Osugi et al. 2011). In the photoperiodic pathway, genes associated with photoreceptors and circadian rhythms influence the expression of downstream genes to heading date (Lee and An 2015). In rice, Hd3a (Heading date 3a, interacts with 14-3-3 protein of the Gf14 family) and RFT1 (Rice Flowering Time 1, activate the Hd3a) activate flowering by inducing the expression of MADS-box genes, such as OsMADS14 and OsMADS15, in the floral meristem (Komiya et al. 2008, 2009). Photophosphorylation of Ghd7 induces OsCKq1 expression, which regulates photosensitivity. OsCKq1 is a gene involved in and responding to photoperiod sensitivity in the rice vegetative growth stage (Wang et al. 2018). Although OsCKq1 has a particularly high homology to Hd16 (Heading date 16), it has yet to be characterized. Hd16 encodes a CKI (casein kinase I)-type protein that contributes to fine-tuning the response to day length by regulating florigen gene expression within a limited photoperiodic range (Nemoto et al. 2018). Given that Hd16 does not appear to influence the expression pattern of clock-associated genes within the photoperiodic pathway, it thus contributes to the control of the heading date independently of circadian rhythms (Hori et al. 2013). CKI is a member of the conserved serine/threonine-specific casein kinases that play roles in a range of eukaryote signaling pathways (Hori et al. 2013). It has been established that the Hd16/EL1 protein phosphorylates Ghd7, a floral repressor (Zhang et al. 2019), and the phosphorylated Ghd7 promotes the upregulation of OsCKq1, thereby inhibiting heading date (Wang et al. 2018).

Currently, research related to the improvement of agronomic traits applied to CRISPR/Cas9 has been conducted (Rao and Wang 2021). In this research, the OsCKq1 screened through QTL mapping by Wang et al. in 2018 was bred the genome-edited plant using CRISPR/Cas9. Genome editing lines mediated by CRISPR/Cas9 are receiving the most attention because target mutations are highly accurate and effective (Zhang et al. 2015). The target mutation is inherited by the next generation, but the Cas9 used for genome editing is removed by separation (Ahmar et al. 2020). This technology has been applied as a powerful and general tool for genome engineering and gene function analysis and has become an innovative method for rice breeding (Rao and Wang 2021).

Therefore, in this research, we applied CRISPR/Cas9 to Ilmi (Oryza sativa ssp. japonica cv. Ilmi) for genome editing of the photoperiod sensitivity-related gene, OsCKq1. The characteristics of the traits related to the heading date and yield of the four genome editing lines generated by CRISPR/Cas9 were investigated. OsCKq1 genome editing lines had earlier heading date than Ilmi, grain length, grain width, and 1,000-grain weight increased, and culm length and panicle length were similar. Therefore, OsCKq1 provides new potential for rice breeding to increase yield by regulating the heading date. This indicates that as a consequence of heading date promotion by day length insensitivity rice, early-maturing-high-yielding type rice cultivars can minimize yield loss.

Results

Development of OsCKq1-GE Lines using the CRISPR/Cas9 System

Target gene editing was performed using the CRISPR RGEN Tools program, for which we designed three sgRNAs that included the domain region of the target gene (Fig. 1A-B). The sgRNAs were inserted into the BsaI restriction enzyme site within the pRGEB32 vector, with their expression therein being driven by the U3 promoter (Fig. 1C). The expression of Cas9 and the hygromycin (Hyg) were driven by the ubiquitin (Ubi) and 35 S promoters, respectively. Following the introduction of sgRNA::pRGEB32 plasmid vectors into Ilmi calli via Agrobacterium-mediated transformation (Fig. 1D), we obtained 39 regenerating plants (Fig. 2A-H), of which 24 survived transplantation into the soil.

Fig. 1

Schematic diagram of OsCKq1 genome editing (OsCKq1-GE) rice process in rice using CRISPR/Cas9 technology. A Three sgRNAs were designed targeting specific domains within the 828-nucleotide sequence of OsCKq1. B Three sgRNAs were selected based on criteria including a GC content of 40–60%, an out-of-frame score of 60 or higher, and 1-0-0 mismatches. C In the pRGEB32 vector, Cas9 is driven by the ubiquitin (Ubi) promoter, and sgRNA expression is controlled by the U3 promoter. The Hygromycin resistance gene serves as a selection marker under the control of the CaMV35S (35 S) promoter. LB and RB denote left border and right border sequences, respectively. D The recombinant pRGEB32::sgRNA vector was introduced into Ilmi-derived callus cells using the agrobacterium-mediated transformation method. T-DNA insertion in regenerated rice was confirmed, and GE lines were identified through genotyping of the target gene region

Fig. 2

Genome editing line selection and genotyping analysis process. A Callus derived from callus induction media for 21 days, followed by agrobacterium-mediated transformation. B Co-cultivation of agrobacterium with callus for 3 days. C Emergence of green spots after 14 days in regeneration media. D Shoot and root formation at green spots after 30 days. E Acclimatization of regenerated plants for 3 days. F Soil transplantation and harvesting of 1st GE seeds. G Advancement of generations in the field. H Efficiency of regeneration plant breeding using CRISPR/Cas9 technology. I PCR-based identification of T-DNA insertion in regeneration plants, specifying the amplification results for Cas9 across different generations (0th, 1st, 2nd, and 3rd ). J Genotyping analysis to identify genome-edited (GE) rice based on sequence alignment with Ilmi at the OsCKq1 region, highlighting specific insertions and homozygosity in OsCKq1-GE rice

Detection of Genome-edited Rice and Assay of Transgene-free Lines

To confirm T-DNA insertion in the regenerating plants (Fig. 2I), we performed PCR analysis to screen for expression of the selection marker Cas9. We accordingly detected T-DNA insertions in 13 of 16 regeneration plants using OsCKq1-sgRNA1, one in four plants using OsCKq1-sgRNA2, and one in four plants using OsCKq1-sgRNA3. Cas9 was amplified in all OsCKq1-GE 6, 7, and 14 lines of GE₁. These target region sequences were compared with the Ilmi sequence for sequence alignment, and we thus identified the OsCKq1-GE₀ 6, OsCKq1-GE₀ 7, and OsCKq1-GE₀ 14 lines in which a total of three indels were generated (Fig. 2J). Specifically, we detected insertions of a thymine (T) insertion in OsCKq1-GE₀ 6, a T and Guanine (G) insertion in OsCKq1-GE₀ 7, and a T insertion in OsCKq1-GE₀ 14. All of these OsCKq1-edited forms were homozygously genome-edited. The protein sequences of OsCKq1-GE₀ 6, OsCKq1-GE₀ 7, and OsCKq1-GE₀ 14 were compared with Ilmi (Additional file 1: Fig. S1). In Ilmi, the OsCKq1 gene at positions 61–63 bp consists of G, A, and G bases, translating to the amino acid glutamic acid (E). In OsCKq1-GE₀ 6, insertion of an A base at position 60 bp causes the 61–63 bp sequence to be translated as glycine (G), resulting in an altered amino acid sequence due to this insertion. In OsCKq1-GE₀ 7, insertions of C and A bases at positions 60 and 61 bp, respectively, lead to a change in the amino acid sequence from glutamic acid (E) to arginine (R). Additionally, in OsCKq1-GE₀ 14, an insertion of a T base at position 42 bp causes a change in the amino acid sequence from proline (P) to serine (S).

From the cultivation of OsCKq1-GE₀ line plants, we harvested grains for each panicle of GE₁ plants. For the OsCKq1-GE₀ 6 line, 22 panicles developed, each of which was harvested after labeling (OsCKq1-GE₀ 6 − 1 to OsCKq1-GE₀ 6–22); in the OsCKq1-GE₀ 7 line, 15 panicles developed (OsCKq1-GE₀ 7 − 1 to OsCKq1-GE₀ 7–15); in the OsCKq1-GE₀ 14 line, 16 panicles developed (OsCKq1-GE₀ 14 − 1 to OsCKq1-GE₀ 14–16). Each of these panicles was harvested and the GE₁ grains of each OsCKq1-GE line were sown. For each generation, as a result of confirming the amplification of Cas9, was not amplified in OsCKq1-GE 6-4-1, 6-13-1, 7-2-1, and 14-2-1 lines of GE₂, but Cas9 was amplified in all other lines. Similarly, Cas9 was not amplified in OsCKq1-GE 6-4-1-1, 6-13-1-1, 7-2-1-1, and 14-2-1-1 lines of GE₃, but Cas9 was amplified in all other lines. To establish the editing type of the target site in these T-DNA insertion lines, we amplified and sequenced the OsCKq1 target sequence.

Morphological Characters of OsCKq1-GE Rice

OsCKq1-GE lines were analyzed for differences in morphological traits compared with those of Ilmi (Fig. 3, Additional file 2: Table S1-3). In terms of heading date, we found that compared with the 126 days after sowing observed for Ilmi, OsCKq1-GE₀ 6 − 4, OsCKq1-GE₀ 7 − 2, and OsCKq1-GE₀ 14 − 2 headed earlier or at a similar time as Ilmi (Fig. 3A and D), with OsCKq1-GE₀ 6–13 heading significantly earlier, by 5 days, than Ilmi (p < 0.05). The culm and panicle lengths (cm) were statistically similar between Ilmi and the OsCKq1-GE lines (Fig. 3E), For the number of tillers, OsCKq1-GE₀ 6 − 4 did not differ significantly from Ilmi (Fig. 3G), whereas OsCKq1-GE₀ 6–13, OsCKq1-GE₀ 7 − 2, and OsCKq1-GE₀ 14 − 2 had significantly fewer tillers than Ilmi (p < 0.05). Regarding grain dimensions (Fig. 3B and J), OsCKq1-GE₀ 6 − 4 exhibited the longest grains and Ilmi the shortest, with corresponding, 1,000-grain weights showing that OsCKq1-GE₀ 6 − 4 had the heaviest and Ilmi the lightest grains, both differences being statistically significant (p < 0.01).

Fig. 3

Comparative analysis of phenotypes of OsCKq1-GE rice grown under long-day (LD) conditions. A Morphology of 4-month-old plants after sowing accelerated flowering in OsCKq1-GE rice compared to Ilmi by 3 to 5 days. B Comparative analysis of grain length. C Investigation of panicle architecture revealing variations in spikelet number and primary branches. D-J Comparative agronomic traits between OsCKq1-GE rice and Ilmi, including earlier heading (D), similar culm length (E), similar panicle length (F), variable tiller numbers (G), longer grain length (H), wider grain width (I), and heavier 1,000-grain weight (J). Scale bars: A = 10 cm, B = 2 mm, C = 2 cm. Statistical significance (p < 0.05) indicated by letters denoted by Duncan’s multiple range test (DMRT)

The yield was compared to Ilmi and OsCKq1-GE rice in each generation (Fig. 4). In GE₁ plants, OsCKq1-GE 6 − 4, 6–13, and 7 − 2 exhibited significantly higher yields compared to Ilmi (p < 0.05), while the yield of OsCKq1-GE 14 − 2 was statistically similar to Ilmi (p > 0.05). In GE₂ plants, OsCKq1-GE 6 − 4 had the highest yield (p < 0.05). The yields of OsCKq1-GE 6–13 and 7 − 2 were significantly higher than Ilmi (p < 0.05) and statistically similar to each other. The yield of OsCKq1-GE line 14 − 2 remained statistically comparable to Ilmi.

Fig. 4

Analysis of the yield each generation. In GE₁ plants, 6 − 4 (yield 6,698 kg/ha, moisture content 11.0%), 6–13 (6,906 kg/ha, 14.3%), and 7 − 2 (6,922 kg/ha, 13.7%) were equally clustered in the high yield group and Ilmi (5,341 kg/ha, 13.8%) and 14 − 2 (5,752 kg/ha, 14.0%) were equally clustered in the low yield group. In GE₂ plants, 6 − 4 (7,673 kg/ha, 17.6%) clustered in the high yield group, followed by 6–13 (6,793 kg/ha, 14.7%) and 7 − 2 (6,952 kg/ha, 14.5%). Ilmi (5,841 kg/ha, 16.2%) and 14 − 2 (5,996 kg/ha, 16.0%) clustered in the low yield group. Mean denoted by the same letter are not significantly different (p < 0.05) as evaluated by Duncan’s multiple range test (DMRT)

Relationship Between OsCKq1 and the Expression of Heading Date Regulatory Genes

To investigate the downstream genes regulated by OsCKq1, we performed quantitative RT-PCR analyses of the relative expression levels of major flowering-associated genes (Ghd7, Ehd1, RFT1, Hd3a, Hd16, Hd1, and OsGI) in Ilmi and OsCKq1-GE₂ lines grown under LD and SD conditions (Fig. 5). Statistical analysis using t-tests revealed significant reductions in the mRNA levels of OsCKq1 in OsCKq1-GE rice cultured under LD conditions compared to Ilmi (p < 0.05), whereas no significant changes in expression were observed under SD conditions. Similarly, there were no significant differences in the mRNA level of OsGI, Ghd7, Hd16, and Hd1 between Ilmi and OsCKq1-GE rice under both SD and LD conditions.

Fig. 5

Analysis of heading date-related gene expression levels in OsCKq1-GE₂ rice under short-day (SD) and long-day (LD) conditions in a growth chamber. RNA was isolated from leaves of 4-week-old plants, and qRT-PCR was used to compare transcription levels of flowering-related genes. Expression patterns of OsCKq1 and OsActin as a control were monitored over a 24-hour period with samples collected at 3-hour intervals. OsCKq1 expression in OsCKq1-GE₂ rice was minimal under LD conditions, contrasting with the fluctuating expression in Ilmi. Data represent mean values with error bars indicating standard deviation. Statistical significance indicated by asterisks (*p < 0.05, **p < 0.01). White and black bars denote periods of daylight and darkness, respectively

However, under LD conditions, OsCKq1-GE rice exhibited significantly higher mRNA levels of Ehd1, Hd3a and RFT1 compared to Ilmi (p < 0.01), while under SD conditions, the mRNA levels of these genes were similar between Ilmi and OsCKq1-GE rice. Additionally, Hd16 expression significantly increased under LD conditions (p < 0.05), whereas its expression was barely detectable under SD conditions with no significant difference between Ilmi and OsCKq1-GE rice. OsGI exhibited similar expression patterns under both photoperiods, with transcript levels increasing during the day, peaking at 18 h, and declining thereafter under LD conditions, and increasing during the day and declining after 12 h under SD conditions.

Coefficient of Variation Analysis for Characters in Rice

During the generation advancement of OsCKq1-GE lines in the field, we examined selected yield-related agronomic traits, namely, heading date, culm length, panicle length, and tiller number (Additional file 2: Table S1-3). for which we calculated CV (coefficient of variation) values (Fig. 6A). For OsCKq1-GE₀ plants, the CV values were 0.0% for heading date, 2.5% for culm length, 21.8% for panicle length, and 8.0% for tiller number. These results indicate minimal variability among plants for heading date and culm length, reflecting uniform growth in these traits in the 0th generation. The CV value for tiller number suggests moderate variability, while the panicle length showed a higher level of variability compared to other traits. In GE₁ plants, the CV values were 1.2% for heading date, 2.1% for culm length, 4.8% for panicle length, and 7.7% for tiller number. The panicle length exhibited low variability, which was reduced by approximately 78% compared to the 0th generation. The low variability in heading date, culm length, and panicle length suggests uniform growth in the current generation. For GE₂ plants, the CV values for heading date and panicle length were 1.4% and 3.8%, respectively, indicating minimal variability and reduced variation among plants for these traits. The CV values for culm length and tiller number were 8.4% and 7.0%, respectively, reflecting moderate variability. With respect to grain length, width, the ratio of grain length to width, and 1,000-grain weight, we obtained CV values of 1.0%, 2.8%, 3.7%, and 1.6%, respectively, for OsCKq1-GE₁ grains, and values of 0.8%, 2.5%, 1.4%, and 2.5%, respectively, for OsCKq1-GE₂ grains, and obtained CV values of 0.9%, 3.1%, 2.8%, and 4.9%, respectively, for OsCKq1-GE₃ grains. At the rice breeding stage, OsCKq1 is regulated by photoperiod sensitivity (Fig. 6B-C).

Fig. 6

Coefficient of variation (CV) analysis of plant and seed characteristics across generations in the field. A CV analysis for GE₀, GE₁, and GE₂ plants based on heading date (HD), culm length (CL), panicle length (PL), and a number of tillers (NT), as well as for GE₁, GE₂, and GE₃ seeds based on grain length (GL), grain width (GW), grain length-to-width ratio (RLW), grain number per panicle (GNP), and 1,000-grain weight (TGW). B Comparison of CV for agronomic traits, including heading date, between Ilmi and OsCKq1-GE rice in advanced generations. OsCKq1-GE rice exhibited an earlier heading date than Ilmi. C Diagram illustrating the role of OsCKq1 in photoperiod sensitivity, showing how OsCKq1 expression, via phosphorylation of Ghd7, suppresses the pathway from Ehd1 to Hd3a/RFT1. T Sen: Temperature sensitivity

Discussion

In this research, we used OsCKq1, identified by QTL mapping to play a role in photoperiod sensitivity, as a target for genome editing using the CRISPR/Cas9 system (Wang et al. 2018). By controlling the heading date of rice, we were able to realize a competitive rice-increasing yield based on the breeding of early-maturing and high-yielding panicle-type rice cultivars, the breeding of which will ensure yield stability.

Rice is a short-day plant, with long days promoting vegetative growth in most of the currently grown rice cultivars, in which flowering is induced in response to a transition from LD to SD conditions (Hayama et al. 2003). In this research, background used was Ilmi. Because Ilmi is a rice cultivar with excellent grain quality and is widely cultivated in Korea. Among OsCKq1-GE₀ plants developed by editing OsCKq1, we confirmed T-DNA insertion in 15 of a total of 24 regenerated OsCKq1-GE₀ plants (Fig. 2H). Based on sequence alignments with respect to the sgRNA target sites in these 15 T-DNA insertion lines, we identified indels in three of the OsCKq1-GE₀ lines (OsCKq1-GE₀ 6, OsCKq1-GE₀ 7, and OsCKq1-GE₀ 14), namely T, T and G, and T insertions, respectively (Fig. 2J). To apply the CRISPR/Cas9 technology to the short breeding cycle strategy and pedigree breeding method, OsCKq1-GE₁ 6 − 1, OsCKq1-GE₁ 6 − 2, up to a number of the last panicles in OsCKq1-GE₀ 6, 7, and 14 (Park et al. 2022). On analysis of the detection of the two target genes OsCKq1 and Cas9, in successive generations of the OsCKq1-GE lines, we initially detected the Cas9 in the OsCKq1-GE₀ lines; however, with generation progression to the OsCKq1-GE₁ and OsCKq1-GE₂ rice, not detected Cas9 in some lines (Fig. 2I). Additionally, in GE rice where Cas9 was not detected in the previous generation, it was also not detected in the subsequent generation. These findings support the research results indicating the silencing or elimination of Cas9 functionality in subsequent generations of plants (Park et al. 2022).

In this research, OsCKq1 mRNA was not detected in all OsCKq1-GE₀ plants because homogeneously OsCKq1-edited plants were selected and used in the GE₀ generation. These findings indicate that by rapidly and accurately editing a target gene, CRISPR/Cas9 can be applied as a powerful tool to efficiently breed new rice cultivars by shortening the breeding cycle. In breeding a new cultivar, it is crucial that the improved trait is fixed quickly and expressed stably in the natural environment. Therefore, this study examined the CV values of the improved traits and other major agronomic traits over three generations to estimate the fixation of the traits. Based on CV determinations performed for evaluated yield- and grain-related traits of each generation of OsCKq1-GE rice (Fig. 6A), we obtained CV values of less than 5% in both G₀ and G₁ plants, with values of 5% or less indicating very little variation, whereas CV values of 15% or less are considered to be indicative of a small degree of variation (Henryon et al. 2003). For example, we obtained a CV value of less than 5% for panicle length in G₁ plants, and CV values of 8.0% and 7.7% for tiller number in G₀ and G₁ plants, respectively. Similarly, for both G₁ and G₂ plants, we detected only very small variations (less than 5%) in the grain-related traits of OsCKq1-GE lines. Summarizing the results of our research, if CRISPR/Cas9 is applied to breeding, the target gene can be homozygous edited in G₀ plants, with targeted traits showing comparatively little variation in the G₀ and G₁ plants and G₁ and G₂ grains. These findings thus indicate the feasibility of breeding new cultivars by editing plants to achieve a shorter breeding cycle (Varshney et al. 2020; Xu et al. 2015).

For this research, we examined the main agronomic traits culm length, panicle length, number of tillers, and yield, as well as heading date, in gene-edited OsCKq1-GE lines (Fig. 3). Under LD conditions, the heading date of OsCKq1-GE₁ lines was established to be 3 days earlier than that recorded for Ilmi. Rice heading date is a complex trait controlled by diverse gene–environment interactions (Chen et al. 2018). Among the G₁ generation OsCKq1 genome-edited plants, we identified OsCKq1-GE 6 − 4, OsCKq1-GE 6–13, OsCKq1-GE 7 − 2, and OsCKq1-GE 14 − 2 as being characterized by an earlier heading than that in Ilmi. However, in the G₂ generation, OsCKq1-GE plants headed earlier than Ilmi (Additional file 2: Table S1). We suspect that these differential heading dates could be attributable to environmental influence and that regardless of the extent of OsCKq1 editing, the heading date will still be dependent to a large degree on environmental factors (Ye et al. 2018).

In response to editing a section of the OsCKq1 sequence using a system combining the sgRNA of the domain region of OsCKq1 and Cas9, we detected a reduction in the expression of the target OsCKq1 gene, a flowering inhibitory gene, as indicated by the early flowering phenotype of OsCKq1-GE₂ rice. To determine how OsCKq1 regulates photoperiod sensitivity, we compared the mRNA expression patterns of the rice flowering-related genes (Zheng et al. 2019) (Ghd7, RFT1, Ehd1, Hd1, Hd16, Hd3a, and OsGI) and OsCKq1 in OsCKq1-GE₂ lines and Ilmi (Fig. 5). Under both SD and LD conditions, there was little difference between Ilmi and OsCKq1-GE₂ rice with respect to the expression of OsGI, Ghd7, Hd1, and Hd16 (Zhou et al. 2021). OsCKq1 is synthesized in response to the phosphorylation of Ghd7 and lies downstream of Ghd7 (Hori et al. 2013; Zhang et al. 2019). OsGI has been established to influence the photoperiod sensitivity of rice and is differentially expressed according to the circadian rhythm (Lee et al. 2016). It is strongly and weakly expressed in response to short and long days respectively, thereby accounting for the observed photoperiodic sensitivity (Tsuji et al. 2011). OsGI lies upstream of Ghd7, which it differentially regulates according to photoperiodic conditions (Zheng et al. 2019). Hd1 inhibits and induces flowering under long- and short-day conditions, respectively (Izawa 2007), and by interacting with Ghd7, contributes to downregulating the transcription levels of Ehd1, a key floral inducer under LD photoperiods (Nemoto et al. 2016).

The full-length cDNA of OsCKq1 comprises 2,097 bp and encodes 698 amino acids, and based on sequence homology analysis, we established that the amino acid sequence of OsCKq1 is similar (82%) to that of rice Hd16. Hd16 has been established to interact with Hd1, Ghd7, RFT1, Ehd1, and Hd3a based on text mining analysis (Hori et al. 2013). As previously mentioned, OsCKq1 is synthesized in response to the photophosphorylation of Ghd7 (Fig. 6C), which has a protein homology with Hd1 (Zong et al. 2021). By regulating the expression of Hd1 and Ghd7, OsCKq1 inhibits heading date by inducing the downregulation of Hd3a and RFT1 under LD conditions, similarly to Hd16 when OsCKq1 is bound, and thus CRISPR/Cas9 editing of OsCKq1 has the effect of promoting heading date. According to Hori et al. 2013, Hd16, which encodes casein kinase 1, controls rice flowering time by modulating the day-length response. Hd16 near-isogenic lines had earlier days to flowering than parental lines (Nipponbare and Koshihikari) under long-day conditions. In this study, OsCKq1 genome editing rice was an earlier heading date than the Ilmi (background as OsCKq1 genome editing rice), which was the same as the results of Hori et al. 2013. The breeding stage of rice is divided into vegetative stage and reproductive stage. Vegetative stage is again divided into basic vegetative growth phase and photophase. The reproductive stage is divided into reproductive growth phase and grain filling phase. OsCKq1 is a gene that responds to the photoperiod sensitivity of rice photophase. The expression of OsCKq1 increased as Ghd7 was activated, and the expression of OsCKq1 suppressed the expression from Ehd1 to Hd3a/RFT1.

In this research, we found that targeted editing of OsCKq1 resulted in changes in rice heading under the long-day conditions, and also were changed the 1,000-grain weight, and yield, while other major agronomic traits did not change significantly statistically. According to Chen et al. 2022, OsCKI encodes casein kinase 1. OsCKI antisense transgenic rice presented low sensitivity to brassinosteroid (BR). BR regulates the expression of WG3 in rice. WG3 encodes a GRAS protein and is a gene involved in grain size. In wg3 plants (WG3 genome editing plants), grain width is increased, therefore increasing grain weight (g). When BR signaling is activated in rice, WG3 expression increases and regulates grain width. Based on these experiment results, we conclude that OsCKq1 genome editing rice has a lower sensitivity to BR signaling and does not activate WG3 expression, resulting in increased grain size.

Editing of OsCKq1 not only accelerated heading date but also promoted increases in yield and weight, thereby indicating the likelihood that OsCKq1 may be a gene associated with the regulation of photoperiod sensitivity and yield. These findings provide evidence that OsCKq1-edited plants can promote rice heading date under LD photoperiods, leading to an earlier harvest than that obtained with mid- to late-maturing rice cultivars, and could thus be a potential candidate gene for securing stable rice yield. OsCKq1-edited plants could accordingly contribute to solving the problems of yield reduction and poor growth, which are currently the most problematic factors hampering the efficient application of double-cropping systems. However, although early-maturing rice cultivars are being bred based on the editing of heading date inhibition-related genes, such early-maturing cultivars are characterized by a shorter period of vegetative growth than late-maturing cultivars, which may ultimately be reflected in reduced yields, contrary to the objective of securing stable rice yields. Nevertheless, except for heading date, the phenotypes of major agronomic traits in the OsCKq1-GE₂ lines were either similar or superior to those of the widely grown cultivar Ilmi, and these plants accordingly provided higher yields than those obtained for Ilmi. In this research, our primary goal was focused on the breeding with field test. However, gene cloning is necessary to fully understand gene function. Together, these findings establish that OsCKq1 is a critical gene involved in regulating heading date in response to day length.

Materials and Methods

Experimental Design

This research was conducted from 2019 to 2022. The genome editing line was cultivated at the Kyungpook National University Agricultural Laboratory (36°6′41.54″N, 128°38′26.17″E). Trials at this location were performed in a randomized complete block design with 3 replicates per year. For each line, 1 rows were transplanted to the field. Each row contains 25 plants, and the plant distance is 30 × 15 cm. All field tests were conducted in compliance with international guidelines and laws provided by the Rural Development Administration (RDA) in Korea. In addition, the rice population was cultivated according to local practice. For experimental materials, we chose the Ilmi cultivar. Ilmi is highly favored in Korea for its ease of cultivation and consistent yields, particularly in the country’s hot and humid climate. It is celebrated for its superior taste and texture, making it a favorite in Korea and abroad. Additionally, Ilmi rice is widely exported internationally. These characteristics make Ilmi an excellent candidate for experimental material. This study complied with the Convention on Trade in Endangered Species of Wild Fauna and Flora (https://www.cites.org/).

RNA-guided Engineered Nuclease Design

For genome editing using the CRISPR/Cas9 system, the guide RNA (gRNA) for the target gene was designed using CRISPR RNA-guided engineered nuclease (RGEN) Tools (Naito et al. 2015). As a proto-spacer adjacent motif sequence type, we selected 5ʹ-NGG-3ʹ, a site recognized by Cas9 derived from Streptococcus pyogenes. Among the single-guide RNA (sgRNA) lists obtained based on automatic analysis after inputting the target sequence, those containing the domain region and having an out-of-frame score of 65 or higher were initially selected (Naito et al. 2015). We finally selected three sgRNA sequences comprising GC contents of between 40% and 60% and mismatches of 1-0-0, which were those from which off-targets were removed in mismatches 0, 1, and 2. Mismatches 0, 1, and 2 indicate the degree of mismatch between the analyzed sgRNA sequence and target sequence of 0, 1, and 2 bp, respectively. The out-of-frame score is a score that is used to predict the possibility of frameshift occurring when an indel occurs in the target sequence.

CRISPR/Cas9 Vector Construction-based Genome Editing in Rice

Having synthesized the selected sgRNAs, we prepared 100 pmole solutions of the forward and reverse sgRNAs. Thereafter, 50 µL reaction mixtures were prepared containing 5 µL of T4 ligase buffer, 5 µL each of the forward and reverse sgRNAs, and 35 µL of distilled water (DW). After heating this working solution at 37 °C for 30 min and 94 °C for 5 min, the temperature was incrementally reduced to 30 °C in 1 °C steps at 28-s intervals, and thereafter treated at 10 °C for 10 min to synthesize a double-stranded sgRNA fragment. Having obtained this sgRNA, we added 3 µL of BsaI and 3 µL of 1× CutSmart buffer (NEB) to 1 µg pRGEB32 vector and DW to a final volume of 30 µL and heated this mixture for 1 h 30 min at 37 °C. For ligation of the double-stranded sgRNA and a BsaI-cut pRGEB32 vector, we used a mixture containing 2 µL of double-stranded gRNA, 1 µL of T4 ligase buffer, 1 µL T4 ligase, and 10 ng of the pRGEB32 vector, made to a final volume of 10 µL with DW, which was incubated at 16 °C for 3 h (Park et al. 2022). Having introduced this working solution into Escherichia coli JM109 competent cells (TAKARA, Tokyo, Japan), these cells were spread on solid Luria–Bertani (LB) medium containing kanamycin and incubated in the dark at 37 °C for 16 h. To confirm sgRNA insertion, plate colonies were used to inoculate a liquid LB medium containing kanamycin and incubated at 37 °C for 16 h. Thereafter, plasmid DNA was extracted and subsequently sequenced commercially by Solgent (Solgent Co. Ltd, Daejeon, Korea). Based on the sequencing results, a plasmid into which sgRNA was inserted was selected and used to transform Agrobacterium tumefaciens EHA105 (TAKARA, Tokyo, Japan).

Agrobacterium-mediated Genetic Transformation and Plant Growth

Agrobacterium-mediated rice transformation was performed using the procedure described by Nishimura et al. (2007). Grains of the Ilmi rice cultivar were sterilized by initially immersing in 70% ethanol for 30 s and then immersing in 1% sodium hypochlorite for 30 min, after which they were rinsed three times with DW and dried on sterile filter paper. The sterilized grains were placed in callus cultural soil medium (4.4 g/L MS medium, 30 g/L sucrose, 2.878 g/L proline, 300 mg/L casein hydrolysate, 100 mg/L myo-inositol, 3 ppm/L 2,4-D, and 4 g/L gelrite) and incubated in the dark for 3 weeks at 28 °C. Agrobacterium into which the pRGEB32 vector had been inserted were resuspended in DW containing acetosyringone to prepare a working solution of OD₆₀₀ = 0.8. Small, hard, bright calli were selected and cultured with shaking in the working solution for 30 s, after which they were dried on sterile filter paper and cultured on co-culture medium (4.4 g/L MS medium, 30 g/L sucrose, 2.878 g/L proline, 300 mg/L casein hydrolysate, 100 mg/L myo-inositol, 3 ppm/L 2,4-D, 100 mg/L acetosyringone, and 4 g/L gelrite) in the dark for 3 days. The Agrobacterium-inoculated calli were then washed with DW containing 25 µL Triton-X and 100 mg/mL cefotaxime for 20 min. Having removed moisture using sterile filter paper, the calli were incubated on a regeneration medium under illuminated conditions at 28 °C and sub-cultured at 2-week intervals. Calli characterized by emergent green spots were placed on regeneration medium (4.4 g/L MS, 30 g/L sucrose, 30 g/L sorbitol, 300 mg/L casein hydrolysate, 2 mg/L kinetin, 1 mg/L NAA, 100 mg/L myo-inositol, 50 mg/L hygromycin, and 4 g/L gelrite) to obtain transformed plants.

Field Test of Successive Generations of OsCKq1 Genome Editing Lines

OsCKq1-GE rice, 0th generation, were cultivated into Wagner pots (256 × 234 × 297 mm) at Kyungpook National University laboratory. For each OsCKq1-GE line, each panicle was assigned a line number, and each panicle was harvested. For seed germination, harvested seeds of the OsCKq1-GE₀ plants were sterilized using Spotak pesticide (HANKOOKSAMGONG, Seoul, South Korea) and soaked in the dark at 25 °C for 4 days. The germinated seeds were sown in seedling trays and incubated in the raising of seedling greenhouse for a month. In the field soil, 25 plants of each line were transplanted in a single row (Park et al. 2022) with a planting distance of 30 × 15 cm. When performing field tests, international guidelines and regulations provided by the Rural Development Administration (RDA) were followed, and management was carried out according to normal local practices. During the field test trials, the plants were fertilized with an N: P₂O₅:K₂O (9:4.5:5.7 kg/10a) fertilizer in accordance with the recommendation of the Agricultural Science and Technology Research Standards of the RDA. After transplanting into the field soil, agronomic traits, phenotype, genotype, and yield, were investigated during cultivation from the seedling stage to harvest.

Genotyping of Target Gene in Genome Editing Lines

To verify whether the target sequence had been edited in the genome-edited lines, we performed genotyping of the target gene in the regenerated plants. Genomic DNA was extracted from the leaves of the rice plants using a DNeasy Plant Mini Kit (Cat. 69104; QIAGEN, Hilden, Germany) in accordance with the manufacturer’s instructions. To select T-DNA insertion lines, we used the sequence of hygromycin, a selection marker of the CRISPR/Cas9 vector pRGEB32, to design primers for PCR analysis. The T-DNA insertion lines were amplified using the primer pair GE_check_F and GE_check_R (Additional file 2: Table S4) which can amplify the CRISPR/Cas9 target region, thereby facilitating the characterization of the editing type of the target sequence. The edited target regions were detected by aligning sequencing chromatograms of these PCR products with those obtained for Ilmi rice. We compared the genetic variations between untransformed Ilmi and genome-edited rice. The primary purpose of using untransformed Ilmi as the control was to compare the overall phenotype and genotype of the genome-edited lines directly to the original cultivar without any additional modifications. Because in practical breeding and genome editing applications, comparisons are typically made against the original non-transformed cultivar. By using untransformed Ilmi as the control, we reflect real applications.

Detection of Transgene-free Lines by Cas9 Selection

Transgene-free lines among the OsCKq1-GE₁ lines were identified based on an analysis of Cas9 expression. Total RNA was extracted from rice plants using an RNeasy Plant Mini Kit (Cat.74904; QIAGEN, Hilden, Germany) in accordance with the manufacturer’s instructions, and used for the synthesis of cDNA using a qPCRBIO cDNA Synthesis Kit (PCRBIOSYSTEMS, USA). The cDNA thus obtained was PCR-amplified using the primer pair Cas9_F and Cas9_R (Additional file 2: Table S4). PCR was performed using a C1000 thermocycler (BioRad, USA) under the following conditions: pre-denaturation at 94 °C for 5 min; followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 30 s, with a final extension step at 72 °C for 5 min. The PCR products were stored at 4 °C and subsequently subjected to electrophoresis on 0.8% agarose gels, Based on which, we selected lines in which OsCKq1 and Cas9 were not expressed.

LD/SD Treatment for Expression Level Analysis of OsCKq1

To compare the expression levels of OsCKq1 and known flowering regulatory genes, OsCKq1-GE rice were treated with LD and SD. Six plants per line were randomly selected and transplanted into pots when the GE rice growing in field soil were at the maximum tillering stage. One plant was transplanted per pots (256 × 234 × 297 mm), and three plants per line were cultured in chambers. Three plants were treated with LD (14 h light, 30 °C/10 h dark, 25 °C, 60% humidity) and the other three plants were treated with SD (10 h light, 30 °C/14 h dark, 25 °C, 60% humidity). There were three plants per line in the chamber, for a total of 15 pots. The pots were spaced apart such that the leaves of the plants were not touching each other. RNA was extracted from leaves when heading occurred or was about to occur in LD. To compare the expression levels of OsCKq1 and flowering regulatory genes in OsCKq1-GE rice at LD/SD conditions, qRT-PCR was performed. After heading, the plants were analyzed to phenotype and transplanted from the chamber back into field soil to allow for sufficient ripening of the seeds.

Investigation of Rice Agronomic Traits

OsCKq1-GE rice was cultivated in the field and various agronomic traits were investigated from the plants. Agronomic traits such as heading date (days after sowing), plant height (cm), culm length (cm), panicle length (cm), number of tillers, and grain weight per plant (g) of OsCKq1-GE lines were investigated. The heading date is a measure of the number of days from sowing to the time of heading. Plant height was measured from the ground surface to the top of the uppermost leaf; culm length was measured from the ground surface to the panicle node; panicle length was measured from the neck to the tip of the panicle. For all data, mean and standard deviation values were obtained from 10 replicates per line.

Quantification and Statistical Analysis

For all data evaluated in this study, mean and standard deviation values were calculated in at least three replicates over 3 years and statistically analyzed using SPSS statistical software (IMMSPSS Statistics, version 22; IBMSPSS Statistics, version 26, Redmond, WC, USA). The data are presented as mean values ± SD and were analyzed using the t-test. The data was used to calculate significant differences (p < 0.05) using ordinary one-way ANOVA. One‐way ANOVA was used to determine statistically significant differences between the mean values of the over 3 groups (p < 0.05), followed by Duncan’s Multiple Range Test (DMRT).

Conclusions

In this research, the function of OsCKq1 screened through QTL mapping was confirmed using the CRISPR/Cas9 system. OsCKq1 regulates rice heading date in response to photosensitivity. By applying CRISPR/Cas9 technology, the functional defect of OsCKq1 upregulates the expression of the heading date promoting gene Ehd1, leading to an early heading date, and as a result, grain length, grain width, and 1,000-grain weight increased, resulting in yield has increased than Ilmi. These results indicate the accuracy of QTL mapping, and also, suggest that OsCKq1 plays an important role in inducing photosensitivity in rice and is related to seed agricultural characteristics. OsCKq1 can be used as a potential gene for early-maturing and high-yielding type rice cultivation and is very useful for improving yield in rice breeding. Furthermore, our research of OsCKq1 provides an excellent example of the advantage of CRISPR/Cas9 technology. First, genome-editing rice with an early flowering phenotype was cultivated quickly and precisely. And homozygous edited lines can be obtained from the G₀ generation. Finally, the exogenous genetic elements (Cas9-sgRNA) can be isolated from the G₁ generation. These results demonstrate that the CRISPR/Cas9 technology is a fast, feasible, and reliable breeding technology that can be used for the breeding of rice cultivars with agriculturally important traits.