ZHOU Chun-yun ,XlONG Hong-chun,Ll Yu-ting,GUO Hui-jun,XlE Yong-dun,ZHAO Lin-shu,GU Jia-yu ,ZHAO Shi-rong ,DlNG Yu-ping,SONG Xi-yun LlU Lu-xiang

1 School of Life Sciences,Qingdao Agricultural University,Qingdao 266109,P.R.China

2 Institute of Crop Sciences,Chinese Academy of Agricultural Sciences/National Engineering Laboratory for Crop Molecular Breeding/National Center of Space Mutagenesis for Crop Improvement,Beijing 100081,P.R.China

Abstract Low stature in wheat is closely associated with lodging resistance,and this impacts harvest index and grain yield.The discovery of novel dwarfing or semi-dwarfing genes can have great significance for dwarf wheat breeding.In this study,we identified an EMS-induced dwarf wheat mutant JE0124 from the elite cultivar Jing411.JE0124 possesses increased stem strength and a 33% reduction in plant height compared with wild type.Gibberellic acid (GA) treatment analysis suggested that JE0124 was GA-sensitive.Analysis of the frequency distribution of plant height in four F2 populations derived from crosses between JE0124 and the relatively taller varieties Nongda 5181 and WT indicated that the dwarfism phenotype was quantitatively inherited.We used two F2 populations and 312 individuals from the reciprocal cross of Nongda 5181 and JE0124 to map the quantitative trait locus (QTL) for reduced height to a 0.85-cM interval on chromosome 2DL.The mapping was done by using a combination of 660K SNP array-based bulked segregant analysis (BSA) and genetic linkage analysis,with logarithm of odds (LOD) scores of 5.34 and 5.78,respectively.Additionally,this QTL accounted for 8.27-8.52% of the variation in the phenotype.The dwarf mutant JE0124 and the newly discovered dwarfing gene on chromosome 2DL in this study will enrich genetic resources for dwarf wheat breeding.

Keywords:reduced height gene,wheat,QTL,BSA,molecular marker

1.lntroduction

Wheat (Triticum aestivumL.) yield can significantly impact global food security (Appelset al.2018).Plant height is one of the key agronomic traits in wheat and it greatly influences grain yield and quality (Casebowet al.2016).The introduction of dwarfing or semi-dwarfing genes into wheat during the ‘Green Revolution' resulted in substantially increased yields due to improved lodgingresistance and harvest index (Pinthus and Levy 1983；Penget al.1999).

More than 20 reduced height(Rht) genes have been identified in wheat (Tianet al.2017；Moet al.2018),and according to the gibberellic acid (GA) response of theseRhtgenes,they have been classified into GA-sensitive or -insensitive categories.The ‘Green Revolution' genesRht-B1b(Rht1),Rht-D1b(Rht2),and their homologous genesRht-B1c(Rht3),Rht-D1c(Rht10),Rht-B1e(Rht11),andRht-B1p(Rht17) were determined to be GA-insensitive(Divashuket al.2012b；Tanet al.2013；Bazhenovet al.2015).Spontaneous base substitutions inRht-B1bandRht-D1bresulted in premature stop codons in DELLA domains,and these mutations led to GA-insensitive dwarfism.Rht-B1ccaused extreme dwarfism due to a 2-kbVejuretrotransposon insertion inRht-B1athatresulted in DELLA domain disruption (Wuet al.2011；Wenet al.2013).Additionally,two copies ofRht-D1bin theRht-D1callele contributed to a 3-fold reduction in height (Liet al.2012).

OnlyRht-D1band its homologous genes were cloned,while the rest of theRhtgenes were located on chromosomes with molecular markers (Borneret al.1997；Gasperiniet al.2012；Chenet al.2015).Specifically,Rht14,Rht16,Rht18,andRht25were found to be located on 6AS (Haqueet al.2011；Grantet al.2018；Moet al.2018)；Rht7was located on 2AS (Chaudhry 1973)；andRht4,Rht5,Rht8,Rht9,Rht12,Rht13,Rht22,Rht23,andRht24were located on 2BL,3BS,2DS,7BS,5AL,7BS,7AS,5DL,and 6AL,respectively (Elliset al.2005；Asplundet al.2012；Chenet al.2015；Vikheet al.2017).Recently,fine mapping of GA-responsiveRht12,Rht18andRht25dwarfing genes has been undertaken (Fordet al.2018；Moet al.2018；Sunet al.2019).The geneRht12was mapped to an 11.21-Mb region on 5AL and included a 10.73-Mb fragment deletion,which resulted in reduction of plant height through activation ofTaGA2ox8(Sunet al.2019).Rht18semi-dwarfism was caused by the increased expression ofGA2oxA9and reduced GA content (Fordet al.2018).Additionally,Rht25was mapped to a 4.3-Mb interval on 6AS and was found to impact heading time and spike development (Moet al.2018).

Although many dwarfing or semi-dwarfing genes were identified,onlyRht-B1b,Rht-D1b,Rht8,andRht9have been extensively utilized for wheat breeding (Yanget al.2006；Wojciechowskiet al.2009；Asplundet al.2012).TheRht-B1bandRht-D1bgenes in Norin 10 originated from the Japanese semi-dwarf wheat variety Daruma,and numerous semi-dwarf wheat varieties were bred from the Norin 10 variety (Galeet al.1985；Hedden 2003).TheRht8gene was tightly linked to thephotoperiod 1(Ppd1) gene which had the advantage of providing earlier flowering in winter wheat and was widely utilized (Divashuket al.2012a).Many dwarfing genes affect the early stages of plant development,such as coleoptile length,early vigor and root length (Elliset al.2004).The genesRht5,Rht12,Rht13,andRht18decreased thousand-grain weight as a tradeoff for reducing height (Daouraet al.2013,2014；Yanget al.2015).The identification of novel dwarfing genes has great significance in the field of wheat breeding.

Lodging has caused widespread economic loss by affecting wheat production and quality and reducing mechanical harvesting efficiency (Berryet al.2003；Islamet al.2007).Previous research suggests that lodging has led to an 8.3% reduction in the production of wheat (Tripathiet al.2004) and approximately 30% annual yield reduction in maize (Xueet al.2016).Dwarfing genes,however,have facilitated the improvement of stem mechanical strength,lowered the center of gravity and increased the resistance to lodging (Tanget al.2009).In this study,we identified an ethyl methanesulfonate (EMS)-induced dwarf wheat mutantJE0124with significant stem strength,and the dwarfing gene was mapped to a 0.85-cM interval on the long arm of chromosome 2D by 660K SNP array-based BSA and genetic linkage analysis.The discovery of this dwarfing gene located on 2DL will enrich genetic resources for dwarf wheat breeding.

2.Materials and methods

2.1.Plant materials and growth measurement

Plants ofJE0124,which was derived from the Chinese winter wheat cultivar Jing411,were compared with wild type(WT) wheat plants growing in field plots at Zhongpuchang Experimental Station of the Institute of Crop Sciences,Chinese Academy of Agricultural Sciences (ICS-CAAS),Beijing,China between 2016 and 2018.The plots received normal fertilization and irrigation.Plant height was measured at maturity in each of the three years.The stem strength from 32 plants of each phenotype in 2018 was determined at 10 cm from the bottom of the stem using a digital push-pull tester.Seedling heights were measured in eight replications of each phenotype on March 30,April 8,April 15,April 21,April 27,May 5,and May 13,2017.The thousand-grain weight,number of kernels per spike and spike length of WT andJE0124were measured in plants collected from the plots in 2016 at Zhongpuchang and Changping Experimental Stations of the ICS-CAAS.Internode length was determined in 2017 in eight replications during the grain-filling stage when plant height was stable.

In addition,four F2segregating populations were constructed by reciprocal crosses ofJE0124with WT plants and the relatively taller variety Nongda 5181.Plants of the crosses were grown at the Zhongpuchang Experimental Station and heights were determined in F1plants in 2017 and 2018 and F2individuals in 2018.

2.2.Cytological and histological analysis

Peduncle samples from WT andJE0124plants collected from plots after the flowering stage were cut into 5 mm fragments.These samples were immediately fixed in a solution of FAA (formalin,acetic acid and 70% ethanol ratio 1:1:18,v/v) for at least 24 h.The segments were next dehydrated with different dilutions of ethanol (70% for 2 h,85% for 2 h,95% for 2 h,100% for 1.5 h,and 100% for 1.5 h).The dehydrated segments were then successively immersed into a mixture of xylene and ethanol with varying concentrations (2:3,3:2,v/v).The transparent samples were then embedded in paraffin and longitudinally cut into 16-μm thick sections with a rotary microtome.The sections were printed onto glass slides and stained with safranine and fast green staining solution.Finally,the stem cells in the paraffin sections were observed using an Olympus light microscope (Olympus,Japan).The cell length and numbers were measured and calculated using Olympus DP2-BSW Software.

2.3.GA treatment

GA treatment was conducted as previously reported (Xionget al.2018) in a growth chamber with 200-300 μmol L-1m-2s-1light at 21°C.Briefly,dry seeds of WT andJE0124mutant were germinated in water for 1-2 days,and then grown in half-concentration of the recommended nutrient solution with or without 1 mg L-1GA3(Sigma,St.Louis,MO,USA).After growth for 8 days,the length of 7-8 seedlings in each treatment was measured.

2.4.DNA extraction and quality assessment

DNA was extracted from 0.1 g of young leaves of plants collected from the plots in 2018 using the DNA-quick Plant System Kit according to the manufacturer's protocol(Tiangen Biotech,Beijing,China).The quality of DNA was tested using 1% agarose gel stained with GelRed.The DNA concentration was measured using Nanodrop One (Thermo scientific,USA).

2.5.Bulked segregant analysis (BSA) by using wheat 660K SNP array

Extremely tall or dwarf individuals in the F2populations that were derived from the cross between Nongda 5181 andJE0124were used for the establishment of BSA bulks.According to the phenotype distribution in the two populations,40 extremely tall and dwarf individuals were selected to pool from each population.The Nongda 5181 andJE0124parents were sampled from 10 plants.DNA samples of 1 μg from each individual were mixed to construct tall and dwarf pools.The two pools along with parent DNA samples were genotyped using the Axiom?wheat 660K SNP array (Thermo) and the genotyping was performed by China Golden Marker (Beijing) Biotech Co.,Ltd.(CGMB,http://www.cgmb.com.cn/).High quality genotyping data were obtained by filtering with the Dish QC threshold of ＞0.82 and the Call-Rate threshold of ＞94.The poly-high-resolution SNPs were selected for further genotyping analysis,and the SNP linked to the phenotype were selected with the following steps:homozygote SNP from Nongda 5181 andJE0124were selected；probe sets were selected with the same genotypings betweenJE0124and the dwarfing bulk,and different genotypings betweenJE0124and Nongda 5181；the candidate regions associated with dwarf phenotypes were limited to the physical position which enriched the selected probes.

2.6.Development of molecular markers by kompetitive allele-specific PCR (KASP) assays

According to the SNPs betweenJE0124and Nongda 5181 from 660K SNP array genotyping,neighboring or included in the mapped region by BSA,we tried to develop molecular markers based on KASP assays.The online primer design pipeline PolyMarker (http://polymarker.tgac.ac.uk/)was used to design specific primers.KASP assays were performed in a 384-well Applied Biosystems Thermal Cycler(Thermo Fisher Scientific) according to the protocol from LGC Genomics (https://www.lgcgroup.com/).The PCR was performed with the following cycling conditions:94°C for 15 min,nine cycles of 94°C for 20 s,touchdown starting at 65°C for 60 s (decreasing 0.6°C per cycle),30 cycles of 94°C for 20 s,and 57°C for 1 min.End-point fluorescence data were screened using the microplate reader FLUOstar Omega SNP (BMG LABTECH,Germany) and analyzed by the Klustering Caller Software.The KASP markers were tested with the two parents and then used for identification of genotypes of F2individuals.

2.7.Linkage map construction

For QTL mapping of the dwarfing gene,each of the 312 F2individuals from the reciprocal cross ofJE0124with ND5181 were used.Based on the genotypes obtained from the KASP assays in the F2populations,the linkage map was constructed using QTL IciMapping 4.0.The threshold of the logarithm of odds (LOD) score was set to 3.0 and the genetic distances between linked markers were assessed according to the combination frequencies using the Kosambi function.

3.Results

3.1.Phenotypic variations of the dwarf mutant JE0124

During 2016-2018,the mean plant height ofJE0124ranged from 60-65 cm,while the mean plant height of WT ranged from 90-95 cm (Fig.1-A and B).Importantly,the stem strength inJE0124was significantly greater than that of WT(Fig.1-C),which indicated this mutant had lodging-resistant potential in the field.Plant height measurements from 30 March,the beginning of the jointing stage,to 13 May,when plant height was relatively stable showed thatJE0124was significantly shorter than WT during all of the investigated stages and the difference in plant height between the phenotypes was greater starting from 21 April,the end of jointing stage (Fig.1-D).In addition to reduced plant height,theJE0124mutant also showed compact plant architecture and delayed heading and maturation time by about 1 week compared to WT.For the yield components,theJE0124mutant had a significantly lower thousand-grain weight than WT at both study locations (Fig.1-E).In contrast,the number of kernels per spike was similar inJE0124and WT (Fig.1-F),and the spike length was slightly reduced inJE0124(Fig.1-G).

3.2.lnternode characteristics and GA sensitiveness

The stems oftheJE0124mutant contained four internodes while WT had five.Additionally,the length of each internode inJE0124was shorter than that of WT (Fig.2-A and B).The cell length of the second internode at the jointing stage ofJE0124was significantly shorter than that of WT (Fig.2-C and D).The cell length distribution showed that the number of shorter cells were greater inJE0124than that in WT(Fig.2-E).

Measurements of seedling coleoptile length to determine if the dwarfing gene inJE0124was GA-sensitive or-insensitive,showed that the coleoptile or seedling length of both WT andJE0124with GA treatment was significantly greater than that of the coleoptile or seedling length without GA treatment (Fig.2-F).In WT,coleoptile length was increased by 59% while inJE0124coleoptile length was increased by 47% of that.It is suggested that the dwarfing gene included inJE0124was GA sensitive.

3.3.Genetic analysis of reduced height gene in JE0124

Fig.1 The phenotype comparison of wild type (WT) and JE0124.A,phenotypes of WT and JE0124.Bar=10 cm.B,mean plant height of WT and JE0124 in 2016,2017 and 2018.C,stem strength of WT and JE0124.D,plant height of WT and JE0124 phenotypes measured on seven dates in 2017.E,thousand-grain weight.F,kernels per spike.G,spike length of WT and JE0124 determined at Zhongpuchang and Changping experimental stations.Bars mean SD (n=8).＊,P＜0.05；＊＊,P＜0.01.

Fig.2 The internode length and longitudinal stem section measurements for wild type (WT) and JE0124.A,the internode phenotype.B,internode length comparison of WT and JE0124.The x-axis 1st to 5th indicates the first to fifth internodes measured starting from the internodes below the ear.C,the stem cell of WT and JE0124.Bar=100 μm.D,the stem cell length comparison of WT and JE0124.E,the distribution of cell numbers with specific length.F,the seedling length of WT and JE0124 in response to gibberellic acid (GA) treatment.Data are means of eight replicates±SD.＊,P＜0.05；＊＊,P＜0.01.

Plant height measurements of F1progeny of crosses betweenJE0124and the relatively taller varieties Nongda 5181 and WT showed that all of the F1progeny were significantly taller thanJE0124(Fig.3-A).Additionally,the heights of 282 individuals of the WT/JE0124F2population,317 individuals of theJE0124/WT F2population,312 individuals of the Nongda 5181/JE0124F2population,and 312 individuals of theJE0124/Nongda 5181 F2population were measured.Frequency distribution analysis suggested that the height measurements fitted the normal distribution in these four populations (Fig.3-B-E).

3.4.Mapping QTL using 660K SNP array-based BSA

The wheat 660K SNP array is a powerful tool for genotype investigations in wheat plants；the array contained 660 009 SNPs,and 44.83% of them were poly high resolution SNPs.We combined the bulked segregant analysis (BSA) method and the 660K SNP array to identify the major QTL for reduced plant height in the two F2populationsfromJE0124and Nongda 5181 parents.A total of 71 114 SNPs were identified between Nongda 5181 andJE0124.By comparison with the SNPs in the extremely tall and short pools (and the parents),we found that a large number of SNPs associated with plant height were enriched on chromosome 2D in the two populations (Fig.4-A and B),suggesting a major QTL was located on chromosome 2D.According to the distribution of SNPs on chromosome 2D,we detected two QTLs namedqPH-2D.1andqPH-2D.2for plant height and one region of SNPs ranging from 550 to 640 Mb (qPH-2D.2) had never been reported as containing theRhtgene and was commonly identified in the two populations (Fig.4-C and D).Therefore,we focused on the region in chromosome 2DL to validate and limit the mapped region.

Fig.3 The plant height of F1 and F2 populations.A,height comparison of parents and F1 plants at maturity.JE0124/wild type(WT) indicates the cross of JE0124 as the female parent and WT as the male parent；WT/JE0124 indicates the cross of WT as the female parent and JE0124 as the male parent；Nongda 5181/JE0124 indicates the cross of Nongda 5181 as the female parent and JE0124 as the male parent；JE0124/Nongda 5181 indicates the cross of JE0124 as the female parent and Nongda 5181 as the male parent.B-E,frequency distribution of plant height in F2 populations from WT/JE0124 (B),JE0124/WT (C),Nongda 5181/JE0124 (D),and JE0124/Nongda 5181 (E).Data are means of eight replicates±SD.Different letters indicate the significant difference between each group at P＜0.05.

3.5.Genetic linkage map analysis of the Rht gene on chromosome 2DL

Based on the identified SNPs betweenJE0124and Nongda 5181 from the 660K SNP array,48 KASP markers on chromosome 2DL were designed and verified in the two parents.A total of 10 KASP markers were successfully developed and used for genetic map construction in chromosome 2DL.In the Nongda 5181/JE0124F2population,the QTLqPH-2D.2with a LOD score of 5.34 was detected at a distance of 0.85 cM between markers ph-3 and ph-4 using IciMapping 4.0 (Fig.5-A).Additionally,the same QTL was located between the above two markers with a LOD score of 5.78,and this QTL was also detected in theJE0124/Nongda 5181 F2population (Fig.5-B).Consistent with this result,the detected region in the 660K SNP arraybased BSA analysis contained the mapped region between markers ph-3 and ph-4.According to the Chinese Spring reference genome sequence (http://www.wheatgenome.org/),we found that the region between markers ph-3 and ph-4 included a physical distance of 15 Mb and contained 166 genes (Appendix A).

Fig.4 The distribution of SNPs across 21 chromosomes identified by 660K SNP array-based bulked segregant analysis (BSA).A,Nongda 5181/JE0124 F2 population and (B) JE0124/Nongda 5181 F2 population.The distribution of SNPs on chromosome 2D in (C) Nongda 5181/JE0124 F2 population and (D) JE0124/Nongda 5181 F2 population.

Fig.5 Location of the detected QTL on chromosome 2DL in (A) Nongda 5181/JE0124 F2 population and (B) JE0124/Nongda 5181 F2 population.

4.Discussion

Dwarfing or semi-dwarfing genes used for wheat breeding practices are of great importance for increasing yield due to the improvements in lodging resistance and harvest index (Foulkeset al.2011；Rebetzkeet al.2011；Duet al.2018).Development of novel dwarfing lines is the basis for enrichment of genetic resources in dwarf wheat breeding.In this study,we identified a dwarf wheat mutantJE0124with a 33% reduction in plant height compared to WT (Fig.1-A and B).It has been suggested that the reduction effect of geneRht-B1bwas 23% and that ofRht8was 7% (Rebetzkeet al.2012).It is possible that the reduction effects inJE0124from two or more dwarfing genes were higher than that of previously reportedRhtgenes.A major QTL on chromosome 2DL was identified using two populations for BSA analysis (Fig.4),and further genetic mapping by KASP markers on 2DL also indicated that theRhtgene on 2DL play important roles in reduction of height.It has been suggested that the GA-sensitive dwarfing genes could be used for seedling vigour improvement under unfavourable growing conditions in the breeding program (Rebetzkeet al.1999).GA treatment suggested that this dwarfing gene is GA-sensitive (Fig.2-F) and would be a promising candidate for application in agricultural practice.

Previous studies have suggested that onlyRht8was located on chromosome 2D (Tianet al.2017),andRht8was mapped to a 1.29-cM interval on the short arm of chromosome 2D.By linked molecular marker analysis ofRht8,we found that the parent Nongda 5181 contained theRht8gene but WT andJE0124did not (Appendix B).Additionally,the peak region detected on the short arm of chromosome 2D (Fig.4-C and D) was near to theRht8mapped region (Gasperiniet al.2012).Therefore,the detected peak region on 2DS was due to the difference in genetic background ofRht8between Nongda 5181 andJE0124.This result indicated that the dwarfing gene located on the long arm of chromosome 2D between KASP markers ph-3 and ph-4 (Fig.5) was a newly identified dwarfing gene.The QTL between markers ph-3 and ph-4 with a LOD score of more than 5 was detected using both F2populations (Fig.5) and indicated the dwarfing gene was found in this region.In addition,this QTL explained 8.27-8.52% of phenotype variation and suggested there were also other dwarfing genes affecting height reduction inJE0124.Enlarged F2populations for fine mapping of this dwarfing gene are needed for further studies.

The dwarf wheat mutantJE0124with strong stem mechanical strength reported in this study (Fig.1-A-C) has potential in breeding for lodging resistance.Lodging is a frequent agricultural production problem in wheat (Berryet al.2006) that can lead to extensive damage to the canopy and can cause major losses to harvestable yield (Berryet al.2004).In addition,lodging in cereal crops significantly decreases photosynthesis and photoassimilation that limits grain growth and lowers yield (Kashiwagi and Ishimaru 2004).Lodging in high plant density conditions occurs due to the growth of individuals that are thinner and taller (Xuet al.2017) which has a negative effect on stem strength(Zhanget al.2014).Lodging was found to result in about a 26-kg ha-1loss in rice yield in southern India (Duwayri and Nguyen 1999) and a 22% yield reduction in soybean (Noor and Caviness 1980).Resolutions have been proposed to alleviate lodging by using reduced height genes and plant growth regulators (Berryet al.2004).In crop breeding,plant growth regulators have been extensively used,but are expensive and require a significant amount of labor for application (Berryet al.2006).The genetic dissection of the dwarfism mechanism could allow wheat producers to make full use of reduced height genes to improve crop yields.Although theJE0124mutant decreased thousandgrain weight compared with WT (Fig.1-E),the higher stem strength which would give improved lodging resistance would provide an advantage for further utilization.Some negative effects may occur with the introduction ofRht-B1bandRht-D1b,which are linked to shorter coleoptile length and reduced early seedling vigor (Rebetzkeet al.2001).Taken together,the dwarf wheat mutantJE0124would seem to be a promising option for breeders to use in developing lodging resistance in wheat cultivars.

5.Conclusion

In this study,we identified a GA-sensitive dwarf wheat mutant with high stem strength and mapped a major QTL for reduced height to an interval of 0.85 cM on the long arm of chromosome 2D.The newly identified dwarfing gene in the present study enriches the genetic resources for dwarf wheat breeding,and the dwarf wheat mutantJE0124provides an important basis for agricultural practice in lodging resistance.

Acknowledgements

This work was financially supported by the National Key Research and Development Program of China(2016YFD0102100 and 2016YFD0101802),the National Natural Science Foundation of China (31801346) and the earmarked fund for China Agriculture Research System(CARS-03).

Appendicesassociated with this paper can be available on http://www.ChinaAgriSci.com/V2/En/appendix.htm