Development of mutants with varying flowering times by targeted editing of multiple SVP gene copies in Brassica napus L.

Sunny Ahmar, Yungu Zhai, Huibin Huang, Kaidi Yu, Muhammad Hafeez Ullah Khan, Muhammad Shahid,Rana Abdul Samad, Shahid Ullah Khan, Olalekan Amoo, Chuchuan Fan, Yongming Zhou

National Key Laboratory of Crop Genetic Improvement, Huazhong Agricultural University, Wuhan 430070, Hubei, China

Keywords:Brassica napus Flowering time BnaSVP Gene editing

ABSTRACT Manipulation of flowering time to develop cultivars with desired maturity dates is fundamental in plant breeding.It is desirable to generate polyploid rapeseed(Brassica napus L.)germplasm with varying flowering time controlled by a few genes.In the present study,BnaSVP,a rapeseed homolog of the Arabidopsis SVP(Short Vegetative Phase)gene,was characterized and a set of mutants was developed using a CRISPR/Cas9-based gene-editing tool.A single construct targeting multiple sites was successfully applied to precisely mutate four copies of BnaSVP.The induced mutations in these copies were stably transmitted to subsequent generations.Homozygous mutants with loss-of-function alleles and free transgenic elements were generated across the four BnaSVP homologs.All mutant T1 lines tested in two environments (summer and winter growing seasons) showed early-flowering phenotypes.The decrease in flowering time was correlated with the number of mutated BnaSVP alleles.The quadruple mutants showed the shortest flowering time, with a mean decrease of 40.6%-50.7% in length relative to the wild type under the two growth conditions.Our study demonstrates the quantitative involvement of BnaSVP copies in the regulation of flowering time and provides valuable resources for rapeseed breeding.

1.Introduction

Rapeseed (Brassica napusL.) is used as a source of edible oil for humans, protein-rich feed for animals, and raw materials for industrial processes worldwide [1-3].Rapeseed originated from the spontaneous hybridization ofB.oleraceaandB.rapa[4].It is closely related to the model plantArabidopsis thaliana, permitting the exploitation of the knowledge of developmental and biochemical pathways generated in that model species [5].

Stabilization of flowering time and flowering period would support efficient mechanized harvesting of rapeseed [6,7].Flowering time is controlled by a complex genetic network that integrates endogenous signals and environmental stimuli including vernalization, temperature, photoperiod, and gibberellin, as identified inB.napusandArabidopsisstudies [8-12].

More than 300 flowering time genes have been identified inArabidopsis,including key regulators of pathways controlling flowering time [13].For example,FRIGIDA(FRI),FLOWERING LOCUS C(FLC) andSHORT VEGETATIVE PHASE(SVP) control vernalization response, whileCONSTANS(CO) controls photoperiodic flowering[14,15].Autonomous-pathway genes includingFLOWERING LOCUS D(FLD),FLOWERING LOCUS CA(FCA), andFLOWERING LOCUS KH DOMAIN(FLK)inhibitFLCand promote floral transition[12].Based on the ABCDE model for floral patterning [16,17],SVPandAGAMOUS-LIKE 24(AGL24) genes act withAPETALA1(AP1) to ensure precise floral patterning by suppressing B-classPISTILLATA(PI)andAPETALA3(AP3),C-classAGAMOUS(AG),and E-classSEPALLATA1(SEP1),SEP2,SEP3andSEP4genes in the early stages of flower organ development [18].

TheSVPgene acts in plant in response to ambient-temperature fluctuations.Loss ofSVPfunction results in insensitivity to variation in ambient temperature, andSVPmediates the temperaturedependent functions ofFCAandFVEin the thermosensory pathway inArabidopsis[19].SVPhomologs promote the floral transition in tomato (Solanum lycopersicumL.) and bell pepper (Capsicum annuumL.) [20-22].InAntirrhinumand barley, the role ofSVP-likegenes is identical to that ofSVPandAGL24[23,24].In barley,overexpression of theSVP-likegene leads to an increase in spike branches [24].These findings suggest that theSVPgene may be necessary for the genetic network involved in development of various inflorescence structures.

In recent years,the CRISPR/Cas9 system has been used to modify the genomes of many crops including rapeseed,to generate targeted genetic mutation for the improvement of agronomic traits such as multilocular silique, pod shattering resistance, leaf shape,plant height and architecture,disease resistance,fatty acid composition,and yellow seed[25-29].To better undertand the functions ofSVPin polyploidy rapeseed, we used the CRISPR/Cas9 geneediting system to generate knockout mutants in multipleBnaSVPcopies to develope a combinatorial set of mutants at the four loci of theBnaSVPgene.We investigated the genomic edting efficiency at different targeting sites of the gene and changes in flowering time with the mutants.

2.Material and methods

2.1.Plant materials

J9707, a semi-winterB.napuspure line with high regeneration capacity under tissue-culture conditions that was developed in our laboratory[26],was used as a receptor inAgrobacterium-mediated transformation.

2.2.Phylogenetic tree construction and identification of protein conserved motif

The amino acid sequences of SVP fromB.napus,A.thaliana,B.rapa, andB.oleraceawere used to generate a phylogenetic tree with Mega 7 software (https://www.megasoftware.net/) by the neighbor-joining method.Conserved motifs were identified with MEME (http://meme-suite.org/tools/meme) with default settings[30].

2.3.CRISPR/Cas9 vector construction and plant transformation

The binary multiplex genome targeting vector pYLCRISPR/Cas9 was used for construct assembly following Ma et al.[31].Selection of single-guide RNA (sgRNA) in the target gene, CRISPR/Cas9 construct assembly, andA.tumefaciens-mediated hypocotyl transformation was conducted inB.napusfollowing Yang et al.[26].The oligo primers used for preparing the sgRNA vectors are listed in Table S1.

2.4.Identification of transgenic mutants and potential off-target mutations

The presence of T-DNA was determined by PCR using the sitespecific primer PB-L/BnaSVPT3-R (Table S1).The mutations were detected in T0transgenic plants using the high-throughput tracking of mutations (Hi-TOM) method described by Zhai et al.[27].Barcode primers,site-specific primers,and index primers are listed in Table S2.

Potential off-target sites were identified with CRISPR-P (http://cbi.hzau.edu.cn/cgi-bin/CRISPR).Specific primers flanking each potential off-target site were designed and used for PCR amplification (Table S2).DNA from T0plants carrying the editing construct was used as a template with wild-type (WT) DNA as a control.DNA library construction, PCR amplification, sequencing, and data analysis followed Huang et al.[32].

2.5.Field experiments and phenotyping

The T0plants were grown in the greenhouse under controlled photoperiod and temperature (16 h/8 h of light/dark at 22 °C) in 2018.The T1lines together with WT were grown in isolated fields in the winter season of 2018-2019 Hubei province and the summer season of 2019 in Gansu province.The homozygous mutants of T2lines were grown in the winter season of 2019-2020 in the isolated field in Hubei province.Field management followed standard breeding practices.

Flowering time was recorded for both T1and T2generations for 10 plants of each homozygous mutant line and the WT from the date of first flower opening.Student’st-test was used to compare the means of the mutant lines and WT.

3.Results

3.1.Molecular cloning and characterization of SVP homologs in B.napus

Four copies ofBnaSVPwere identified in the A and C subgenomes ofB.napus:BnaA04.SVP(BnaA04g12990D),BnaA09.SVP(BnaA09g42480D),BnaC04.SVP(BnaC04g35060D), andBnaC08.SVP(BnaC08g34920D).The genomic DNAs of aboveBnaSVPcopies were cloned from theB.napuspure line J9707,and corresponding coding sequences were deduced based on the Darmor-bzh reference genome [33].

Sequence alignments of DNA and protein revealed high similarity between the four copies of theBnaSVPgene (Figs.S1, S2).They showed 93.5% similarity at the nucleotide level and 97.1% at the amino acid level, suggesting that they encode enzymes with similar functions.

The MEME conserved-motif search showed that motif 1 was highly conserved in the four copies; motif 2, motif 3, and motif 5 were highly conserved in three copies; and motif 4 and motif 6 were conserved in two copies (Fig.1A, B).

Furthermore, the phylogenetic tree showed that the fourBnaSVPcopies were highly similar to their respective homologs inB.rapa,B.oleracea, andA.thaliana(Figs.1C, S3).

3.2.CRISPR/Cas9 vector construction to knock out gene BnaSVP

Among the four copies ofBnaSVP, three copies (BnaA09.SVP,BnaC04.SVPandBnaC08.SVP) contained eight exons in their open reading frame, while one copy (BnaA04.SVP) contained seven exons, with the conserved region similar to that ofArabidopsis(Fig.2A).Four sgRNAs (S1-S4) were designed with CRISPR-P [34]to create targeted mutations in the copies ofBnaSVP.S1 was designed to target the fourth exons ofBnaA09.SVP, BnaC04.SVP and BnaC08.SVP,and the third exon ofBnaA04.SVP;two sgRNAs(S2 and S4) were designed to target the first exons of three copies

(BnaA09.SVP, BnaC04.SVP, andBnaC08.SVP), and S3 targeted the first exon of theBnaA04.SVPcopy (Fig.2A).ACRISPR/Cas9 vector harboring these four sgRNAs was created (Fig.2B).

Fig.1.Conserved motifs in BnaSVP and phylogenetic relationships.(A) Distribution of amino-acid motifs in BnaSVP protein sequences.Motif 1 is present in all copies in BnaSVP, and the remaining motifs are present in two or three copies of the gene.(B) Motifs conserved among BnaSVP proteins.(C) Phylogenetic tree showing the protein sequence relationship among SVP homologs from Brassica napus(BnaA04g12990D,BnaA09g42480D,BnaC04g35060D,and BnaC08g34920D),A.thaliana(AT2G22560),B.rapa(Bra038510, Bra030227), and B.oleracea (Bol031757).

3.3.Creation of CRISPR/Cas9-targeted mutations in BnaSVP

To identify putative mutations in the expected locations, all T0transgenic plants were tested with copy-specific primers for Hi-TOM sequencing of the target sites (Fig.S3).Of 67 transgenic plants, 30 (45%) contained targeted mutations, including 24(36%) homozygous, 24 (36%) heterozygous, 16 (24%) bi-allelic and 6(9%)chimeric mutations at the target sites(Table S3).Among them were two quadruple, six triple, seven double, and 15 single mutants (Table S3).

The four target sites showed variable editing efficiency,with the highest at S2(36%).Lower mutagenesis efficiency was observed at S1 (28.%) and S4 (12%), and no mutation was detected at S3(Table S3), suggesting that proper selection of sgRNAs is needed for effective gene editing.The S1 site was able to target theBnaA04.SVPandBnaA09.SVPcopies (40% on average) but not theBnaC04.SVPandBnaC08.SVPcopies, although it precisely matched allBnaSVPcopies(Fig.2A;Table S3).Similarly,a marked bias of targeted editing at S2 was observed in the A subgenome(67%)in contrast to that in the C subgenome (28%).

3.4.Isolation of stably edited mutants without T-DNA elements in T1 and T2 generations

To develop stable mutant lines, 10 independent T0mutants ofBnaSVPwere self-pollinated to generate T1and T2progeny.Hi-Tom sequencing confirmed the presence of targeted mutations at the target sites of the progeny (Table S4).Variable numbers and types of homozygous/bi-allelic mutants with various types of allelic combinations were detected in allBnaSVPcopies, includingBnaA09.SVPsingle mutants (AA1aa2CC1CC2),BnaA09.SVP/BnaC04.SVP(AA1aa2cc1CC2) andBnaA09.SVP/BnaC08.SVP(AA1aa2CC1cc2)double mutants,BnaA04.SVP/BnaA09.SVP/BnaC08.SVP(aa1aa2CC1cc2)triple mutants,andBnaSVP(aa1aa2cc1cc2)quadruple mutants(Fig.3; Table S4).Most of the homozygous/biallelic mutations in the target sites were predicted to cause frameshifts, resulting in nonfunctional proteins (Figs.3, S6).The mean transmission rates of allelic mutations observed in the T0-to-T1and T1-to-T2generations were 64% and 66%, respectively (Table S5).

To identify mutants with targeted modification but without transgenes in theB.napusgenome, PCR analysis of the T1and T2plants was performed using the site-specific primer PB-L/BnaSVPT3-R.A variety ofBnaSVPmutant types were obtained,includingsingle, double, triple, and quadruple homozygous mutants with T-DNA-free elements in the T1and T2generations(Figs.4, S4; Table S4).

Fig.2.Gene model of BnaSVP with the target sequence and schematic diagram of a binary plasmid vector.(A)The annotated SVP gene structure is shown on the black line.BnaA09g42480D,BnaC04g35060D,and BnaC08g34920D have eight exons,marked as squares separated by seven introns,and BnaA04g12990D consists of seven exons separated by six introns.In the gene model,the vertical dotted line indicates the target site and the arrow indicates the sgRNA direction.The target sequence sgRNAs are represented by protospacer adjacent motif (PAM) sites and marked in red in the wild-type sequences.(B) Schematic diagram of the binary plasmid vector used for BnaSVP gene editing.Hygromycin-resistance cassette is composed of the hygromycin phosphotransferase coding sequence and the 35S promoter and terminator of cauliflower mosaic virus;Cas9 expression cassette is composed of the Cas9 coding sequence and the 35S promoter and terminator of cauliflower mosaic virus; four sgRNA (S1-S4) were driven by respectively AtU3d,AtU3b,AtU6-1,and AtU6-29 promoters of Arabidopsis thaliana.PB-L/BnaSVPT3-R is a site-specific primer pair for PCR detection of the presence of T-DNA.

Fig.3.CRISPR/Cas9-induced homozygous mutants of BnaSVP in T1 generation.PAMs are underlined and in red,and nucleotide indels are marked in red with details labeled at right; ‘‘a(chǎn)a#” and ‘‘cc#” represent therespective mutated alleles of the target gene in the A and C genomes.‘‘a(chǎn)a1aa2cc1cc2”, ‘‘a(chǎn)a1aa2CC1cc2”, ‘‘AA1aa2cc1CC2”,‘‘AA1aa2CC1cc2”, and ‘‘AA1aa2CC1CC2” represent homozygous mutations of the target gene in the quadruple copies, BnaA04.SVP/BnaA09.SVP/BnaC08.SVP triple copies,BnaA09.SVP/BnaC04.SVP double copies,BnaA09.SVP/BnaC08.SVP double copies,and BnaA09.SVP single copy,respectively.‘‘-”and‘‘+”denote respectively deletion and insertion of a specified number of nucleotides.

Fig.4.CRISPR/Cas9-induced germline transmission at the S1 and S2 target sites of SVP from the T0 to the T2 generation.Insertions and deletions by CRISPR/Cas9 are shown with red font and red hyphens, respectively.aa# and cc# show mutant-allele numbers.‘‘-‘‘ and ”+‘‘ indicate respectively deletion and insertion of the indicated number of nucleotides.

3.5.Off-target editing in T0 transgenic plants

Potential off-target sites in theB.napusgenome were searched for,based on similarity to the four sgRNAs,using the CRISPR-P program to investigate whether off-target mutations occurred.There were 8, 5, 0, and 4 possible off-target sites for S1, S2, S3, and S4,respectively (Table S6).Sequencing of the PCR products of these 17 potential off-target sites from the T0mutant plants revealed no mutations (Tables S6 and S7), indicating that the designed sgRNAs did not create off-target mutations in the genome of screened plants.

3.6.Targeted mutations in BnaSVP promoted early flowering in B.napus

In comparison with the WT plants (86.5 days), all the mutants showed accelerated flowering, ranging from 51.4 to 75.2 days under the summer growing condition (Fig.5A).The quadruple mutant showed the most significant change, with the flowering time shortened by about 51.4 days on average compared to the WT, whereas theBnaA09.SVP(AA1aa2CC1CC2) single mutant showed the least change (11.3 days earlier) relative to the WT.The other mutants with different allelic combinations at theseBnaSVPcopies showed a consistent but various alteration in flowering time.The changes in flowering dates showed good correlation with number of mutatedBnaSVPalleles(Fig.5A).

Under winter growing conditions, the flowering times of all mutant lines ranged from 71.9 to 104.9 days, earlier (P< 0.05)than that of the WT (Fig.5B, C).The changes in flowering time of the mutant lines with different allelic combinations showed a similar trend to those observed in the summer-growing condition, with some variation due to difference in environments(Fig.5B, C).

Fig.5.Phenotypic performance of BnaSVP mutants.(A) Statistical analysis of days to flowering of wild type, single, double, triple, and quadruple mutants under winter growth conditions in Hubei province.(B)Statistical analysis of days to flowering of wild type,single,double,triple,and quadruple mutants under summer growth conditions in Gansu province.(C)Quadruple mutants(left)flowered earlier than wild-type plants(right).The image was acquired in an isolated research field at Huazhong Agricultural University, Hubei on day 71 after sowing in the winter season of 2019-2020.Values are means±SD (n ≥10 plants per genotype).Lowercase letters indicate significant differences at P = 0.01.‘‘a(chǎn)a1aa2cc1cc2”, ‘‘a(chǎn)a1aa2CC1cc2”, ‘‘AA1aa2cc1CC2”, ‘‘AA1aa2CC1cc2”, and ‘‘AA1aa2CC1CC2” represent homozygous mutations of the target gene in quadruple copies, BnaA04.SVP/BnaA09.SVP/BnaC08.SVP triple copies, BnaA09.SVP/BnaC04.SVP double copies, BnaA09.SVP/BnaC08.SVP double copies, and BnaA09.SVP single copy, respectively.

4.Discussion

The CRISPR/Cas9 system has shown great potential for promoting gene function research in plants,because it can be easily modified to produce functional mutations of a single or multiple gene clusters with unknown functions[35,36].In this study,we applied the system to effectively knock out multiple copies ofBnaSVPand compared the editing efficiency.Among the four copies,BnaA09.SVP(AA2) andBnaA04.SVP(AA1) showed the most abundant mutants while fewer mutants was obtained forBnaC04.SVP(CC1)andBnaC08.SVP(CC2)were observed.To estimae the editing accuracy,we examined potential off-targets in the T0mutant plants and found mutations nowhere rather than the targeted sites, suggesting that with well-designed specific sgRNA, the risk to produce unwanted off-target changes is low.

As these two children would possibly inherit the kingdom, it was natural that their people should take a great interest in them, and it fell out that all the tranquil10 and peace-loving citizens desired that Placida should one day be their Queen, while the rash and quarrelsome hoped great things for Vivien

Various factors affect the efficacy of plant sgRNA,such as different expression levels of Cas9 and sgRNA,GC%content,target environment, and secondary target structure of sgRNA [31,37,38].In this study, the mutagenic efficiency ranged from 0 to 36% across the four sgRNAs (Table S8).Among the four sgRNAs (S1-S4) used,S1 and S2 showed the highest mutation rates, indicating that not all the promoters used for the sgRNA are equally efficient for driving genome editing in rapeseed.Yang et al.[26] used the same CRISPR/Cas9 system with 10 different sgRNAs and similar promoters inB.napusto target three genes:BnCLV1(two copies),BnCLV2(two copies), andBnCLV3(two copies) driven by promoters of AtU3d, AtU3b, AtU6-1, or AtU6-29.Zhai et al.[27,39] targeted three genes,BnTT8(two copies),BnIND(two copies), andBnALC(two copies) with 12 different sgRNAs driven by these same promoters (Table S8).Overall, our findings are consistent with those of our previous studies, indicating that the editing efficiency of AtU3d and AtU6-29 is higher than that of other promoters.

TheSVPgenes are well known as controllers of flowering time inArabidopsis.In a previous study [40], the expression of theSVPgene decreased to an undetectable point before the formation of sepals during flower development.TheArabidopsis svpmutant showed an early-flowering phenotype without marked changes in other characteristics.Thesvpmutant showed rapid vegetative growth.Genetic study [41] further showed that mutation of theSVPgene inhibited the late flowering caused by the overexpression of theFLM/MAF1gene, and ansvp flmdouble mutant showed behavior similar to that of its single mutants inArabidopsis.Overexpression ofMtSVPa1andMtSVPa2,Arabidopsis SVPhomologs fromMedicagospp., caused floral defects and delayed flowering inArabidopsis[42].However, there is limited information about the functions ofSVPin other species,especially polyploid crops like rapeseed, in which multiple copies of the gene are present in two subgenomes.

To characterize the phenotypes ofBnaSVPmutants, homozygous mutant lines of the T1and T2generations were grown in two different environments.In comparison with WT plants, all mutant lines showed dramatic advances in flowering time.The flowering time of the mutants ranged from 71.9 to 104.9 days,much earlier than that of the WT.Such phenotypes were similar to that of theArabidopsis svpmutant [40].Also, an additive effect on flowering time was observed in mutants with different allelic combinations.The quadruple mutants showed the shortest flowering time, with a mean decrease of 40.6%-50.7% in the number of days to flowering during the summer and winter growth conditions in comparison with the WT.Our results provide strong evidence thatBnaSVPplays a key role in the control of flowering time in rapeseed.It is uncertain whether this gene could alter the vernalization requirement in rapeseed, because the receptor(J9707) is a semi-winter type ofB.napusand can flower without strict vernalization.Still, the mutants developed in the present study may be useful for early-maturity breeding in rapeseed.

We have demonstrated that targeted mutations using the CRISPR/Cas9 gene-editing system can be effectively applied to polyploid species.The development of plants simultaneously mutated in multiple copies ofBnaSVPwill not only allow the analysis of gene dosage and function in this allotetraploid species but also provide valuable germplasms with expected flowering time by selection of optimal allelic combinations in rapeseed breeding programs.

CRediT authorship contribution statement

Yongming Zhou and Chuchuan Fanconceived the study;Sunny Ahmar, Yungu Zhai, and Chuchuan Fandesigned the experiment;Sunny Ahmar, Yungu Zhai, and Huibin Huangperformed the experiments;Kaidi Yuperformed data analysis;Rana Abdul Samad, and Muhammad Hafeez Ullah Khanhelped in plant transformation;Sunny Ahmar, Yungu Zhai, and Chuchuan Fanwrote the paper;Sunny Ahmar, Olalekan Amoo, Shahid Ullah Khan,and Muhammad Shahidhelped in the revision of this paper;Yongming Zhou and Chuchuan Fansupervised the study.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The study was supported by the National Key Research and Development Program of China (2017YFE0104800) and the National Natural Science Foundation of China (31671725).

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2021.03.023.