CRISPR/Cas9 mediated TaRPK1 root architecture gene mutagenesis confers enhanced wheat yield

2.1 Wheat germplasm and growing conditions

The seeds of the hexaploid bread wheat cultivar ‘‘Pakistan 2013″ were obtained from the National Institute of Genomics and Advanced Biotechnology (NIGAB), NARC, Islamabad, Pakistan. The control and transgenic (T0) plants were grown in glasshouse at NIGAB. Green-house conditions for day light period of 16 h with 25 ± 3 °C temperature and night period of 8 h with 15 ± 3 °C temperature were kept. Experiments were carried out on complete randomized design (CRD). Three readings were observed for agronomic parameters. The main steps involved in CRISPR-based targeted mutagenesis are summarized in Supplementary Figure S1.

2.2 Target selection and guide RNAs designing

The sequences of the RPK1 genes in the A, B, and D genomes of wheat were retrieved from “Ensemble Plant” through the BLAST of RPK1 sequences of Arabidopsis thaliana and Oryza sativa (Shi et al., 2014, Zou et al., 2014). The sequence of RPK1 with a significant percentage ID and low E-value was selected. The gRNAs were designed manually from exon region, and the possible off-target sites were identified using BLAST search (https://blast.ncbi.nlm.nih.gov/Blast.cgi). The CRISPR/Cas9 system provided by Dr. Bing Yang (Genetics, Development and Cell Biology department, Ames, Iowa State University, USA) was followed. The gRNA expression cassettes were synthesized, with BtgZI restriction site, TGTT, and AAAC at the 5́ ends of sense and antisense strand, respectively. Similarly, from the 2nd exon, another four oligos were designed with BsaI restriction site, GTGT, and AAAC at the 5́ ends of sense and antisense strand, respectively (Table 1).

2.3 CRISPR/Cas9 based vector construction

CRISPR/Cas9 system was used for TaRPK1 gene disruption, to unveil its role in wheat. The constructs were synthesized following Bing Yang’s method (Char et al., 2019).

2.4 Transformation and screening of edited wheat plants

The seeds of the Pakistan-2013 wheat cultivar were used following the method of in-planta transformation (Abdul et al., 2011). For in-planta transformation, the seeds were first soaked in 70 % v/v solution.Seeds were then washed with 50 % Clorox for 15 min and were allowed to completely dry. Sowing of seeds was done on Murashige and Skoog(MS) solid medium (Murashige and Skoog, 1962).

The overnight grown liquid cultures of A. tumefaciens strain LBA4404 in LB liquid media were centrifuged at 85000 rpm/10 min, then supplemented with 200 µM acetosyringone and mixed gently. For the three-day-old seedlings, the apical meristem part was pierced up to 1 mm depth, and agrobacterium inoculum was injected into it. After three days of co-cultivation, to remove excess bacteria. The seedlings were then shifted to the selection and regeneration medium containing cefotaxime 250μg/mL; BAP 1.5 mg/L, and hygromycin. The transformed seedlings were then placed in a greenhouse under a controlled environment (22/16 °C with day/night 16/8h photoperiod and 70 % humidity) till maturity.

For the detection of gRNA efficiency in wheat plants, a T7 endonuclease I assay was performed (Kim et al., 2018). To confirm the DNA double-strand break, PCR amplicons were gel purified, denatured, reannealed, and treated with T7EI. The PCR amplicons were sequenced to determine the mutation frequency.

2.5 Root architecture traits and morphological parameters determination

The root architecture traits including root length, depth, diameter, volume, surface area, and root orientation were observed in four-week-old T0 transgenic lines and wild-type plants with the help of RhizoVision software (Seethepalli et al., 2021). Morphological data measured included the number of effective tillers, plant height, spike length, and 1000-grain weight at the maturity stage. A minimum of three replicates from T0 transgenic lines and control plants were recorded for each trait.

2.6 Statistical analysis

All the recorded data was statistically analyzed by using Microsoft Excel 2019. Student t-test was performed to determine the significant differences among the samples (n = 3, *= P < 0.05, **=P < 0.01).

3.1 Selection of target homeolog of TaRPK1 gene

The RPK1 sequences in the A, B, and D sub-genomes of wheat retrieved by BLAST from “Ensemble Plant”, with the highest percentage ID and low E-value were located on chromosome 2. TraesCS2A02G176500, TraesCS2B02G202900, and TraesCS2D02G183900 wheat sequence Ensemble IDs showed the highest percentage of similarities 86 %, 85.5 %, and 94.1 %, respectively (Supplementary Table S1). The coding sequences were aligned and sgRNAs were synthesized from the conserved region. The sgRNA1 was designed from the conserved region of A and D genomes, sgRNA2 from A, B, and D genomes, and sgRNA3 from the conserved region of A and B genomes (Supplementary Table S2). These sgRNAs were synthesized from the first exon of the A and B genomes, and the second exon of the D genome (Fig. 1).

Close

Fig. 1

Illustration of sgRNA expression cassette. (A) The representation of TaRPK1-gRNA1 (Target 1) and TaRPK1-gRNA2 (Target 2) for constructing LR1. (B) Indication of TaRPK1-gRNA1 (Target 1) and TaRPK1-gRNA3 (Target 2) for LR2 construct. PAM sites were shown in red color and restriction endonucleases (BtgZI and BsaI) recognition sequences were underlined. (C) Schematic illustration of binary vector transformed into wheat. The constructs TaRPK1-gRNA consists of a cassette with Kanamycin resistant gene, flanked by attR1 and attR2 gateway recombination sequence, inserted into the pOs-Cas9 binary vector via LR recombination reaction. The expression cassette comprises gRNA pairs (TaRPK1-gRNA1 and TaRPK1-gRNA2/3) followed by Cas9 endonuclease regulated by Rice Ubiquitin promoter. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Export to PPT

3.2 CRISPR/Cas9 TaRPK1 vector and in-planta transformation

The binary vector CRISPR/Cas9 containing TaRPK1-sgRNAs were constructed accomplished by gateway recombination reaction (Supplementary Figure S2). The first construct with a combination of sgRNA1 and sgRNA2 is termed LR-1, while the LR-2 construct contains the sgRNA1 and sgRNA3 expression cassettes, respectively. In planta apical meristem transformation being an easy, cheap, and efficient method for transformation was adopted for transformation using CRISPR/cas9 targeted mutagenesis (Fig. 2). More than 25 independent lines were obtained by using this technique.

Close

Fig. 2

In planta apical meristem transformation CRISPR/Cas9 targeted mutagenesis. (A) Wheat seeds germination on MS medium. (B) Transfer of ex-plant into co-cultivation media. (C) Screening of plants on selection media. Plants with hygromycin resistance survived on selection media. (D) Acclimatization of transplanted seedlings in the growth room, by covering the seedlings with a transparent sheet. (E) Transformed plants shifted to pots in glass house and grown under controlled conditions.

Export to PPT

3.3 Mutation detection in T0 plants

The detection of the edit mutation type was performed after the transformation and regeneration of T0 plants. A total of 31 transgenic plants were obtained from screening of selection antibiotic, 16 of which were of LR-1 constructs and 15 of LR-2 constructs (Table 2). Mutation detection in these transgenic plants was performed by genomic DNA extraction from the leaf tissues of T0 plants (Supplementary Figure S3), followed by designing specific primers flanking target sites of A, B, and D genomes (Supplementary Table S3). The PCR products were digested with T7 endonuclease I (T7EI) assay followed by Sanger sequencing for mutation detection. T7 endonuclease I assay with PCR products did not reveal the expected pattern of digestion for the insertion/ deletion (InDel) mutation. However, Sanger sequencing of the PCR products confirmed seven CRISPR/Cas9-based mutated T0 lines of LR-1 constructs and six CRISPR/Cas9-based mutated T0 lines of LR-2 constructs, with a percentage mutation frequency of 43.75 %, and 40 % for LR-1 and LR-2 constructs, respectively. The two LR-1 and LR-2 constructs targeting the RPK1 gene achieved similar mutation frequency. Overall, these results showed a high mutation efficiency of 41.93 %.

3.4 Types of mutations caused by CRISPR/Cas9

The sequencing results showed two types of mutations, deletions, and insertions. The insertions of one bp, A, and T were detected at target sites (Figs. 3-6). The frequency of insertion was less as compared to that of deletion. Short deletions of one to nine bp were also observed. The deletions were mostly short, however longer deletions such as 12d, 17d, 19d, and 20d (Fig. 3A) were detected in T0-6, T0-10, T0-14, and T0-16 lines. The longer deletions were observed at target site 2 by sgRNA2 of LR-1 construct, showing high efficiency of sgRNA2 compared to sgRNA1 and sgRNA3 (Fig. 5A). The total deletion ratio was higher compared to total insertions and Cas9 cleavage sites usually 3 bp upstream of PAM when all the mutation events were summed. These findings indicate that our CRISPR/Cas9 system did work in the T0-edited plants.

Close

Fig. 3

CRISPR/Cas9 induced targeted mutagenesis of the TaRPK1 target gene by LR-1 construct within A, B, and D genomes of wheat T0 plants (A). PAM sites were shown in red, nucleotides in blue denote target site 1 (T1), and nucleotides in orange represent target site 2 (T2). The lowercase letter denotes insertion, dashes represent deletions and dots indicate nucleotides not shown. Spike architecture of wild type and edited transformed LR-1 T0 lines (B). Root architecture of wild type and edited transformed LR-1 T0 lines (C). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Export to PPT

Close

Fig. 4

Phenotypic comparison of LR-1 construct for Plant height (cm, A), Number of effective tillers (B), Spike length (cm, C), Seed weight (g, D), Root length (mm, E), Root depth (mm, F), Root diameter (mm, G), Root volume (mm³, H), Root surface area (mm², I), and Root orientation (J) of wild type (WT) and edited T0 lines. The data indicates the mean ± SE of three biological replicates. Student’s t-test was used (* = P < 0.05, and ** = P < 0.01).

Export to PPT

Close

Fig. 5

CRISPR/Cas9 induced targeted mutagenesis of the TaRPK1 target gene by LR-2 construct within A, B, and D genomes of wheat T0 plants (A). PAM sites were shown in red, nucleotides in blue denotes target site 1 (T1), and nucleotides in purple represent target site 2 (T2). The lowercase letter denote insertion, dashes represent deletions and dots indicates nucleotides not shown. Spike architecture of wild type and edited transformed LR-2 T0 lines (B). Root architecture of wild type and edited transformed LR-2 T0 lines (C). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Export to PPT

Close

Fig. 6

Phenotypic comparison of LR-2 construct for Plant height (cm, A), Number of effective tillers (B), Spike length (cm, C), Seed weight (g, D), Root length (mm, E), Root depth (mm, F), Root diameter (mm, G), Root volume (mm³, H), Root surface area (mm², I), and Root orientation (J) of wild type (WT) and edited T0 lines. The data indicates the mean ± SE of three biological replicates. Student’s t-test was used (* = P < 0.05, and ** = P < 0.01).

Export to PPT

3.5 Monoallelic and diallelic mutations in T0 generation

Given the complexity of possible results, from zero to three target sites and using 2 different constructs in complex sub-genomes of hexaploid bread wheat (Table 3). Results displayed higher monoallelic mutation compared to diallelic mutation. 37.5 % monoallelic mutation at target site 1 within the D genome by sgRNA1 was observed by the LR-1 construct (Table 3). The LR-2 constructs showed a higher monoallelic mutation frequency of 26.67 % at target sites 1 and 2 within the A, B, and D genomes (Table 3). The highest diallelic mutation of 25 % was observed at target site 2, by sgRNA2 of LR-1 construct within A genome of wheat. However, the 20 % higher diallelic mutation frequency was detected by sgRNA3 at target site 2 of the LR-2 construct within the A genome (Table 3).

3.6 Disruption of TaRPK1 effects morphological parameters and root system architectural traits

To see if TaRPK1 gene disruption effects the morphological parameters and root system architecture, we observed and measured these traits in edited lines along with control plants (Figs. 3-6). Alteration in spike architecture in the edited plants was observed (Fig. 3B). Plant height increased considerably in LR-1 edited T0 lines compared to WT plants (Fig. 4A). Plant height increased significantly in T0-16 and T0-6 lines. T0-16 had the most tillers number, followed by T0-6 > T0-10 > T0-1 LR-1 edited lines (Fig. 4B). Spike lengths were reduced significantly in genome-edited LR-1 lines (Fig. 4C). A significant reduction in spike length was observed in T0-16 lines. Data regarding 1000 grain weight was significantly increased in all of the T0-edited LR-1 lines with a maximum in T0-16 > T0-6 > T0-1 lines compared to the control plants (Fig. 4D).

We also detected LR-2 edited T0 wheat plants, which had changed spike architecture (Fig. 5B). The morphological traits such as the number of effective tillers and plant height were significantly increased, while the spike length was significantly reduced in LR-2 edited wheat T0-1, T0-4, and T0-6 plants (Fig. 6A-C). 1000 grain weight was significantly enhanced in all of T0 edited LR-2 wheat lines (Fig. 6D). All LR-2 T0 edited lines increased seed weight, with T0-1, T0-4, and T0-6 lines having the highest.

The root architectural traits showed significant effects in LR-1 genome-edited T0 lines (Fig. 3C). The root length, root depth, volume, and surface area were significantly increased (Fig. 4E-F and H-I), however, the root diameter and root angle were significantly reduced in genome-edited LR-1 (Fig. 4G and 4 J) T0 lines compared to the WT plants. T0-1, T0-6, T0-10, and T0-16 lines displayed very significant differences in comparison to the control plants. Similarly, in the other construct (LR-2), root architectural traits also displayed significant variations in genome-edited T0 lines (Fig. 5C).

Notably, editing of wheat TaRPK1 genes with specifically designed sgRNAs may have changed the open reading frame and amino acid sequence in transgenic edited lines relative to WT plants (Supplementary Figure S4). In comparison to WT plants, the reduced spike length, root diameter, and root angle and increased root length, depth, surface area, and volume increase the number of effective tillers and grain weight. Increased grain weight and effective tillers on modified lines offset the grain yield loss from lower spike length. This research shows that TaRPK1 editing improves root architecture and yield.