Broadening the genetic base of Brassica juncea by introducing genomic components from B.rapa and B.nigra via digenomic allohexaploid bridging

Jiaqin Mi,Jin Liu,Fang Yu,Yangui Chn,Jiayi Ming,Zhiyong Xiong,Fngqun Yu,Jiana Li,Wi Qian,*

a College of Agronomy and Biotechnology,Southwest University,Chongqing 400716,China

b Academy of Agricultural Sciences,Southwest University,Chongqing 400716,China

c State Cultivation Base of Crop Stress Biology for Southern Mountainous Land of Southwest University,Chongqing 400715,China

d Key Laboratory of Herbage and Endemic Crop Biotechnology,School of Life Sciences,Inner Mongolia University,Hohhot 010070,Inner Mongolia,China

e Agriculture and Agri-Food Canada,Saskatoon Research and Development Center,107 Science Place,Saskatoon,SK S7N 0X2,Canada

Keywords:Brassica juncea Diploid progenitor species Exotic introgression Genetic diversity Digenomic hexaploid

ABSTRACT A narrow genetic base has hindered improvement of Brassica juncea (AjAjBjBj).In this study,large-scale genomic components were introduced from diploid ancestor species into modern B.juncea using a digenomic hexaploid strategy.The hexaploids AjAjArArBjBj and AjAjBjBjBnBn were first developed from B.juncea×B.rapa(ArAr)and B.juncea×B.nigra(BnBn),and then crossed with dozens of B.nigra and B.rapa,respectively.Both types of hexaploid showed high pollen fertility and moderate seed set throughout the S1 to S3 generations,and could be crossed with diploid progenitor species under field conditions,in particular for the combination of AjAjBjBjBnBn × B.rapa.Thirty-type and 31-type B.juncea resources were generated,of which thetype showed higher fertility.Of these new-type B.juncea resources,97 individual plants were genotyped with 42 simple sequence repeat markers,together with 16 current B.juncea accessions and 30 hexaploid plants.Based on 180 polymorphic loci,the new-type B.juncea resources and current B.juncea were separated clearly into distinct groups,with large genetic distance between the new-type B.juncea resources and current B.juncea.Our study provides a novel approach to introducing large-scale genomic components from diploid ancestor species into B.juncea.

1.Introduction

The Brassica allotetraploid species B.juncea,which originated from spontaneous interspecific hybridization between B.nigra(BB,2n=16)and B.rapa(AA,2n=20)[1,2],is cultivated worldwide as vegetable,oilseed,and condiment crops [3,4].Owing to its unique and favorable traits,in particular drought tolerance and shattering resistance [5,6],B.juncea offers enormous cultivation potential in semi-arid areas.However,its narrow genetic base has hindered its genetic improvement [1,4,7,8].

In contrast,the two diploid ancestors of B.juncea harbor large genetic variation [9–12].B.rapa has a long domestication history(thousands of years) and is cultivated for vegetable and oilseed uses.According to its growth habit,B.rapa can be classified into spring,winter,and semi-winter types.B.nigra is widely distributed in Europe,Asia,and North Africa,and harbors many useful genes and traits,such as heat tolerance,drought tolerance,and disease resistance (to blackleg and clubroot) [13–15].Broadening the genetic base of B.juncea by introducing genetic components from its diploid progenitor species is a promising strategy [16–19].Moreover,the exotic introgression may produce interspecific or intersubgenomic heterosis for allotetraploid species [20,21].

Artificial resynthesis of B.juncea from B.rapa and B.nigra is the conventional way to introduce genetic components from diploid progenitor species into B.juncea.However,the procedure is complicated (embryo rescue is needed to overcome the hybridization barrier between B.rapa and B.nigra),labor-intensive,and inefficient.Direct interspecific hybridization between B.juncea and diploid progenitor species is an alternative choice,but B.juncea and B.nigra are cross-incompatible,and it is time-consuming to recover the tetraploid from the triploid F1hybrid derived from the cross between B.juncea and diploid species.

Fig.1.Crossing scheme for developing new-type B.juncea with introgression of genomic components from B.nigra and B.rapa.

Trigenomic hexaploid bridging materials have been created and used to transfer genetic components from diploid species to allotetraploid species[22–27].The trigenomic allohexaploid AjAjBjBjCoCoderived from B.juncea × B.oleracea (CoCo) was used as a bridging material to transfer ‘‘tour” cytoplasmic male sterility to allotetraploid Brassica species [22].Yellow-seeded B.napus was developed from a cross between AjAjBjBjCoCowith a yellow-seeded B.oleracea line [27].A new-type B.juncea developed from B.juncea × ArArBcBcCcCcnot only broadened the genetic diversity of B.juncea,but also presented potential for intersubgenomic heterosis [26].An AnArCnCc-type B.napus was developed by crossing a natural B.napus with ArArCcCc-type B.napus derived from ArArBcBc-CcCc[31].An ArArCcCc-type B.napus was crossed with hundreds of ArArBcBcCcCcplants and subjected to extensive recurrent selection,generating a novel B.napus gene pool with rich genetic diversity and phenotypic variation [32,33].The trigenomic hexaploid strategy provides a route to constructing novel allotetraploid resources,although it is laborious and time-consuming to eliminate the undesired genome from the trigenomic hexaploid(such as by eliminating the C genome to develop a new-type B.juncea and cleaning the B genome to create a new-type B.napus).

Recently,AnArCnCo-type B.napus resources with wide genetic diversity were easily generated by crossing digenomic hexaploid AnAnCnCnCoCo(derived from B.napus×B.oleracea)with 142 B.rapa lines[28].Moreover,the hexaploid AnAnCnCnCoCowas successfully uesed to introduce the sclerotinia stem rot resistance from B.oleracea to B.napus[29,30].It is unknown whether the digenomic hexaploid strategy also applies to the creation of new-type B.juncea.In the present study,new-type B.juncea resources are generated from crosses between a wide range of diploid progenitors and digenomic hexaploids to estimate the potential of the digenomic hexaploid strategy in broadening the genetic base of current B.juncea.The stability of digenomic hexaploids and the crossability of hexaploid with diploid parents are investigated.

2.Materials and methods

2.1.Breeding strategy and plant materials

A breeding diagram for developing new-type B.juncea is presented in Fig.1.-and-type B.juncea were developed by crossing digenomic hexaploid AjAjArArBjBj(derived from B.juncea × B.rapa) and AjAjBjBjBnBn(derived from B.juncea × B.nigra) with B.nigra and B.rapa,respectively.

To generate digenomic hexaploids,three B.juncea accessions obtained from Southwest University (SWU) in China,including two oil-type accessions and one vegetable-type accession,were randomly pollinated by two B.rapa accessions (one semi-winter type and one spring type)and one B.nigra accession.Ovary culture was performed for 70 hybrid ovaries per cross combination at 8 days after pollination (DAP) and followed by chromosome doubling following Wen et al.[34].The plants from S1to S3generation of hexaploid AjAjArArBjBjwith pollen fertility ＞90%were pollinated with a panel of B.nigra accessions collected from the U.S.National Plant Germplasm System (NPGS) to develop-type B.juncea,while the plants from the S1to S3generation of hexaploid AjAjBjBjBnBnwith pollen fertility ＞90%were crossed with dozens of B.rapa accessions collected from NPGS (spring and winter types)and SWU (semi-winter types) to generate-type B.juncea.Selected plants in the S1,S2,and S3generations of hexaploid AjAjArArBjBjwere randomly pollinated with respectively 4,35,and 24 B.nigra accessions,while AjAjBjBjBnBnplants in S1,S2and S3generation were pollinated with 4,44,and 39 B.rapa accessions.Sixteen B.juncea accessions (including vegetable and oil types,referred to as current B.juncea) obtained from the Centre for Genetic Resources in the Netherlands and SWU were used for genetic comparison.Full details of all plant materials are presented in Table S1.All materials were planted at intervals of 30 cm between rows and 25 cm within rows in an experimental field of SWU,with conventional management.

2.2.Cytological analysis

The ovaries from young buds in 1–2 mm were treated with 8-hydroxyquinoline for 4 h,fixed with Carnoy’s solution (Vethanol:Vaceticacid=3:1) for 24 h,and then stored in 70% ethanol at 4 °C for chromosome counting.The young buds were fixed directly in Carnoy’s solution for meiotic study.Chromosome number,karyotype,and behavior were observed via fluorescence in situ hybridization (FISH) following Xiong and Pires [35,36] with a few modifications.In detail,A-genome specific probes CentBr I and CentBr II were developed following Xiong et al.[36],and the Bgenome specific probe pBNBH35 was developed following Schelfhout et al.[37].The probes were labeled with fluorescein-12-dUTP(Flu-,green),Cy3-dCTP (Cy3-,green),or Cy5-dUTP (Cy5-,red).Chromosomes were counterstained with 0.2% hydrochloride(DAPI) solution and examined with a fluorescent microscope(Nikon Eclipse 80i,Nikon Corporation,Tokyo,Japan).Plants were karyotyped with a combination of two probe mixtures (mixtures 1 and 2):FISH was first performed with mixture 1,which consisted of Cy3-5S rDNA,Flu-5S rDNA,Flu-pBrSTR rDNA,Cy5-26S rDNA,and Cy3-KBrH092N24,and then chromosomes were reprobed after stripping with mixture 2,which consisted of Cy3-pBNBH35,Flu-Cent Br II,and Cy5-Cent Br I.The chromosomes of A and B genomes were identified following Xiong and Pires[35]and Man et al.[38],respectively.

2.3.Fertility assessment

Pollen tubes were observed following Yao et al.[39]:pistils of the females were collected at 6 h after pollination and fixed in Carnoy’s solution for 24 h,then treated with 8 mol L-1NaOH for 8 h under dark,and finally stained with 0.1% aniline blue solution before observation under the fluorescence microscope.Pollen fertility(%)was assessed in 10 randomly chosen plants of each hexaploid line and three plants of each new-type B.juncea.Pollen grains from three flowers of each plant were stained with 1% acetocarmine and observed immediately under a microscope.Over 200 pollen grains were observed in each microscopic field and the proportion of stainable pollen grains was calculated.Seed set(seeds per pod) was recorded for five plants per line,with 30 continuous siliques sampled from the main inflorescence of each plant.Pod setting proportion (%) of hybrids was recorded before harvest as the percentage of hybrid pods in all pollinated flowers.

2.4.Genetic diversity evaluation

Genomic DNA was extracted using the CTAB method.Fingerprints of genotypes were developed with 42 simple sequence repeat (SSR) primer combinations designed randomly from the A genome (22 SSRs) and the B genome (20 SSRs) (Table S2).Polymerase chain reaction (PCR) was performed using the touchdown methodology of Wu et al.[40].The products were separated on 10% polyacrylamide gel and stained with silver nitrate solution.The bands were described by absence (0) or presence (1).The matrix of genetic distance (GD) among accessions was calculated following Nei and Li [41],and the values in the GD matrix among genotypes were subjected to principal component analysis (PCA)in NTSYS-PC 2.1 (http://www.exetersoftware.com/cat/ntsyspc/ntsyspc.html) [42].

2.5.Data analysis

Statistical analyses including analysis of variance (ANOVA),Fand t-tests,Duncan’s multiple comparison,and Pearson’s simple correlation analysis were performed using SAS version 8 (http://www.sas.com/) [43].

3.Results

3.1.Characterization of hexaploids

Following the crossing scheme in Fig.1,three AjAjArArBjBjand one AjAjBjBjBnBngenotypes were developed,presenting vigorous plants and hypertrophic organs in comparison with B.juncea(Fig.2A,B).Nine AjAjArArBjBjindividuals with pollen fertility ＞90% were found to carry 56 chromosomes,of which 40 belonged to the A genome and 16 to the B genome(Fig.2C).Plants with high pollen fertility and large flower organs were selected from the S0to S3generations of the hexaploids AjAjArArBjBjand AjAjBjBjBnBnfor subsequent study.

The S0generation of AjAjArArBjBjand AjAjBjBjBnBnshowed high pollen fertility(86.1%–93.0%)and moderate seed set(2.6–3.3 seeds per pod)(Fig.3A),values markedly higher than those of the triploid AjArBjand AjBjBn(pollen fertility ≤10.5%,seed set ≤0.1 seeds per pod)(Table S3).No difference(P ＞0.05)was detected across the S0to S3generations in both pollen fertility and seed set,while only small variation was found among genotypes (Fig.3A;Table S3).The mean pollen fertility of the hexaploids from S0to S3was 90.2%,with 3.1 and 3.0 seeds/pod by self-and open-pollination,respectively.

3.2.Crossability of hexaploids with diploid parents

Pollen grains of diploid parents germinated and extended well on the stigmas of the hexaploids(Fig.2D),and a few produced normal embryos (Fig.2E).The hybrid pod setting ratio did not differ(P=0.96) between AjAjBjBjBnBnand AjAjArArBjBjhexaploids,but differed (P ＜0.01) across generations (Fig.3B;Table S3).In the S1generation,the hybrid pod setting proportion was 40.0% in AjAjBj-BjBnBn× B.rapa and 44.8% in AjAjArArBjBj× B.nigra,increasing to respectively 73.2% and 65.4% in the S3 generation.The seed set in the AjAjBjBjBnBn× B.rapa hybrid was increased (P ＜0.01) from 0.05 seeds/pod in S1to 2.11 seeds/pod in the S3generation,whereas that in AjAjArArBjBj× B.nigra hybrid remained as low as respectively 0.03,0.20,and 0.07 seeds/pod in the S1to S3generations (Fig.3B;Table S3).These results indicate higher crossability of hexaploid AjAjBjBjBnBnwith B.rapa than of AjAjArArBjBjwith B.nigra.

3.3.Characteristics of new-type B.juncea resources

A total of 61 new-type B.juncea lines were obtained from hexaploid and diploid parents,including 30 AjArBnBjlines derived from AjAjArArBjBj× B.nigra and 31lines derived from AjAjBjBj-BnBn× B.rapa.The hybrids presented rich phenotypic variation and inherited morphological characteristics from the parental species,such as purple stem or small flowers from B.nigra,or dark green leaf and late flowering time from winter-type B.rapa(Fig.2F–L).Five plants with pollen fertility ＞88% were confirmed to carry 36 chromosomes (Fig.2M),implying the feasibility of selecting B.juncea-like individuals based on pollen fertility.Seed set was correlated with the pollen fertility of new-type B.juncea resources (rself-pollination=0.69,P ＜ 0.01;ropen-pollination=0.66,P ＜0.01).The-type B.juncea showed higher pollen fertility(79.9%) and seed set (4.5 seeds per pod) than thetype(60.1% and 1.9 seeds/pod,respectively) (P ＜ 0.01) (Fig.3C;Table S3).Furthermore,30.0% of thelines showed pollen fertility ＞88%and 46.6%showed self-pollinated seed set ＞5 seeds per pod,whereas the proportions were much lower in thetype(9.7%and 12.9%,correspondingly)(Fig.3C).Thus,thetype resources showed higher fertility than thetype.

Fig.2.Morphological and cytological characterizations of hexaploids and new-type B.juncea resources.(A)Morphology of hexaploid plants in S0 generation.Scale bar,10 cm.(B) The size of buds,flowers and pods of AjAjArArBjBj,AjAr Bj and B.nigra (BnBn).Scale bar,1 cm.(C) Euploid AjAjArArBjBj with 56 chromosomes.Chromosomes were counterstained with DAPI(blue);A-subgenome chromosomes are labeled white and green with probes CentBr I and CentBr II,respectively,while B-subgenome chromosomes are labeled red with the probe pBNBH35.Scale bar,10 μm.(D) Growth of B.nigra pollen on the stigma of hexaploid at 6 h after pollination.(E) Embryos in hybrid pod of AjAjArArBjBj×B.nigra at 15 days after pollination (DAP).Scale bar,5 mm.(F–L)Morphological characteristics of new-type B.juncea resources.Images show the variation in leaf shape and plant color,as well as the purple nodes and small flowers inherited from parental B.nigra.Scale bars,10 cm.(M) The AjBnBj-type B.juncea with 36 chromosomes.A-subgenome chromosomes are labeled in white and green and B-subgenome chromosomes in red.Scale bar,10 μm.

To investigate the extent of exotically introgressed genomic components in the new-type B.juncea resources,16 current B.juncea accessions,30 hexaploid plants,and 97 new-type B.juncea individual plants were genotyped with 42 SSR markers,resulting in 180 polymorphic loci.The mean GDs in new-type B.juncea resourceswere around twofold higher than that within current B.junceaand in hexaploids(Table 1),indicating wide genetic variation in the new-type B.juncea resources.The GDs between the new-type B.juncea resources and current B.junceawere over fourfold greater than the GD within current B.juncea(Table 1),suggesting a large genetic difference between the new-type B.juncea resources and current B.juncea.This inference was further supported by the PCA (Fig.4),in which the new-type B.juncea resources were clearly separated from current B.juncea.These results suggested that the genetic components of the new-type B.juncea resources,whether introgressed from B.rapa or B.nigra,were highly divergent from those of current B.juncea.

Table 1 Genetic distance (GD) within and among hexaploids,new-type B.juncea resources,and current B.juncea.

4.Discussion

In this study,we introduced a novel digenomic hexaploid strategy for developing new-type B.juncea resources.Digenomic hexaploid AjAjArArBjBjor AjAjBjBjBnBnwas first developed from a few combinations of B.juncea × B.rapa or B.juncea × B.nigra (with modest labor for embryo rescue),and then crossed with dozens of B.nigra or B.rapa,respectively.In comparison with the conventional strategies,the cross of diploid hexaploid bridging materials with diploid species can avoid the interspecific hybridization barrier and generate AABB-type hybrids easily under field conditions.In comparison with the trigenomic hexaploid strategy and direct hybridization between B.juncea and diploid progenitor species,the digenomic hexaploid strategy yields euploid AABB-type plants as early as the F1generation from hexaploid and diploid parents.The digenomic hexaploid strategy does not involve the C genome,saving much time.Thus,the strategy is convenient and efficient for introducing large-scale genomic components from diploid parental species into current B.juncea.

Fig.3.Characteristics of the diploid hexaploid and new-type B.juncea resources.(A)Fertility of hexaploids in three generations.Bars and circles represent seed set and pollen fertility respectively.‘‘Self” and ‘‘Open” below the X-axis indicate the ways of pollination.(B) The crossability of hexaploid with diploid ancestral species.The mean and median values are indicated by the‘‘×”and the line within each box.The number of cross combinations in each generation is shown in each box.(C)The fertility of new-type B.juncea resources.

Fig.4.Principal component analysis among new-type B.juncea resources and current B.juncea.The 45 AjArBnBj genotypes,52 AjArBnBj genotypes,and 16 current B.juncea accessions are indicated by squares,circles,and pluses,respectively.

Genomic instability caused by chromosomal rearrangements and unbalanced gametes has been reported repeatedly in artificially synthesized Brassica allohexaploids [24,26,28,44,45].For the same reason,the maintenance of hexaploidy from selfpollinated generations of digenomic hexaploids seems difficult or impossible.For example,in our previous cytological study [38],12.4% of randomly sampled AjAjArArBjBjS1plants were euploids,whereas no euploid was found in the subsequent S2and S3generations owing to a high frequency of irregular chromosome pairing and abnormal chromosome segregation.The genomic instability of allohexaploids will inevitably cause high segregation for chromosome number and phenotypic traits among progeny,limiting the potential of the hexaploid as a new crop.However,aneuploid digenomic hexaploids may also generate euploid gametes.In our previous study [38],13.8% of 47 pollen mother cells of the AAAABB S1generation (including aneuploid plants) presented euploid AABtype segregation at first anaphase,permitting the formation of AABB-type offspring after the cross with B.nigra.However,because it is difficult to identify the chromosome number and the chromosomal segregation for all plants in each generation,we selected hexaploid plants with large flowers and high pollen fertility to develop progeny and to cross with the diploid parents,given the known positive correlation between genome stability and seed set/pollen fertility of synthesized allohexaploids [24,26,28].The fertility of selfed progeny of hexaploids was stable across S0to S3generations (with 2–4 seeds per pod and pollen fertility around 90%),and these hexaploid plants were crossable with diploid progenitor species under field conditions,finally yielded B.juncea-type hybrids.Thus,although genomic instability limits the use of the digenomic hexaploid as a crop,it is possible to use hexaploid plants from different generations (particularly early generations)as bridging materials to create new-type B.juncea resources[20,46].Considering the decrease in the proportion of euploid gametes with generations,asexual propagation of true hexaploid plants in earlier generations (such as S0and S1) will ensure relatively high efficiency in developing B.juncea-type hybrids [38,45].

With digenomic hexaploid bridging,a panel of new-type B.juncea resources was generated,showing high genetic distance from current B.juncea.This result suggests a genetic base of the newly synthesized B.juncea resources differing from that of current B.juncea.Although these materials presented varied fertility,B.juncea-type plants(AABB genome,36 chromosomes)with high fertility were found in both theresources.Considering the varying proportions of euploid gametes from hexaploids across generations,some of these new-type B.juncea resources must be B.juncea-like plants with a few additional or missing chromosomes.However,these plants also have the potential to yield B.juncea-type offspring from hybridization with natural B.juncea,because B.juncea-like plants may also generate euploid gametes.For this reason,both B.juncea-type plants and B.juncea-like resources are of high potential for the breeding of B.juncea.The cross between these plants and current B.juncea will broaden the genetic base of B.juncea,introduce favorable traits from diploid species into current B.juncea [46],and possibly generate intersubgenomic heterosis[26].Of course,unfavorable traits will also be introduced into the recipient B.juncea.In this case,recurrent selection (using the recipient B.juncea or other superior B.juncea accessions as the recurrent parent)may help to eliminate undesired traits.

In summary,we have demonstrated an effective and convenient approach to widening the genetic base of current B.juncea,by introducing large-scale genomic components from diploid parental species into B.juncea using a digenomic hexaploid bridge that carries only the A and B subgenomes.The novel germplasm created in this study is a valuable resource for the breeding of B.juncea.

CRediT authorship contribution statement

Jiaqin Mei:Data curation,Formal analysis,Visualization,Writing -original draft.Jin Liu:Investigation,Data curation,Visualization,Writing -original draft.Fang Yue:Investigation,Data curation,Formal analysis.Yangui Chen:Investigation,Data curation,Formal analysis.Jiayi Ming:Investigation,Data curation,Formal analysis.Zhiyong Xiong:Supervision,Writing -review &editing.Fengqun Yu:Supervision,Writing-review&editing.Jiana Li:Supervision,Writing-review&editing.Wei Qian:Conceptualization,Project administration,Supervision,Writing -review &editing,Funding acquisition.

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 authors thank Dr.Qinfei Li of Southwest University and Ms.Shuling Man of Inner Mongolia University for technical assistance.This study was financially supported by the National Key Research and Development Program of China (2018YFE0108000),the Natural Science Foundation of Chongqing (cstc2019jcyj-zdxmX0012),and the Fundamental Research Funds for the Central Universities(XDJK2018B022,XDJK2018AA004).

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2021.09.008.