Parimal Vikhe,Suhasini Venkatesan,Ajit Chavan,Shubhada Tamhankar,Ravindra Patil*

Genetics and Plant Breeding Group,Agharkar Research Institute,G.G.Agarkar Road,Pune 411004,India Savitribai Phule Pune University,Ganesh Khind,Pune 411007,India

Keywords:Coleoptile length Rht14 Semidwarf wheat Triticum durum

A B S T R A C T Short coleoptiles associated with GA-insensitive Rht-1 alleles in wheat reduces yield due to poor seedling establishment under dry,or stubble-retained conditions.Hence there is a need for alternative dwarfing genes for wheat improvement programs.GA-sensitive dwarfing gene Rht14 confers semidwarf stature in wheat while retaining longer coleoptiles and early seedling vigor.Two RIL populations were used to identify the map position of Rht14 and to estimate its effect on plant height,coleoptile length,seedling shoot length,spike length and internode length.Rht14 on chromosome 6A was mapped in the genomic region 383-422 Mbp flanked by GA2oxA9 and wmc753 in a Bijaga Yellow/Castelporziano RIL population.Recombination events between Rht14 and GA2oxA9 in the RIL population indicated that Rht14 might not be allelic to GA2oxA9.The conserved DNA sequence of GA2oxA9 and its flanking region in Castelporziano also suggested that the point of mutation responsible for the Rht14 allele must be a few Mbp away from GA2oxA9.The dwarfing effects of Rht14 on plant height,internode length and seedling vigor were compared with those of Rht-B1b in an HI 8498/Castelporziano RIL population.Both genes significantly reduced plant height and internode length.Rht-B1b conferred a significant reduction in coleoptile length and seedling shoot length,whereas Rht14 reduced plant height,but not coleoptile and seedling shoot length.Therefore,Rht14 can be a used as an alternative to Rht-B1b for development of cultivars suitable for deeper sowing in dry environments and in conditions of conservation agriculture where crop residues are retained.

1.Introduction

Deployment of semidwarfing genes to increase harvest index and prevent lodging is a well-proven strategy to capture higher grain yield in major cereal crops,such as rice and wheat.The Green Revolution involved the use of semidwarf cultivars carrying the Semidwarf-1(sd-1)gene in rice.Height reduction in genotypes with this allele is due to loss of function in the GA biosynthetic gene GA20ox2 with a consequent increase in harvest index and retention of seedling vigor in water-sown rice[1,2].The ubiquitous GA-insensitive dwarfing alleles Rht-B1b and Rht-D1b in wheat carry truncated Rht1 proteins caused by mutations in the conserved N-terminal gibberellin-signaling domain.These Rht-1 alleles confer GA-insensitivity and impose their effects on plant stature by abolishing the interaction between Rht-1 protein,which is a repressor of plant growth,and the gibberellin(GA)receptor,GID1,thus preventing proteolytic degradation of Rht-1[3].Wheat cultivars carrying Rht-B1b and Rht-D1b produce optimum yields under high-fertility irrigated conditions,but are not well adapted for deeper sowing conditions in dryland agriculture due to shorter coleoptiles and low early vigor,often expressed as reduced seedling emergence[4,5].Moreover,crop stands also remain poor where previous crop residues pose a barrier for seedling emergence due to short coleoptiles associated with the Rht-1 semidwarfing alleles[6].Therefore,there is a need to include alternative dwarfing genes in wheat improvement programs.

Twenty four semidwarfing genes are known to modulate plant height in Triticum species[7-12].Eighteen are sensitive to exogenous gibberellic acid(GA)and plants carrying these GA-sensitive genes,unlike Rht-1 semidwarfing alleles,have unaltered early vigor and coleoptile length.GA-sensitive dwarfing genes retain longer coleoptiles,early vigor,and grain yield[13];and have greater biomass,harvest index,grain and spike number[14,15]as well as increased spikelet fertility[16].All of these traits are crucial for optimal crop stands and high yield,particularly in dry environments where surface moisture is often limited.Longer coleoptiles and early vigor also facilitate seedling emergence where previous crop residues are retained[6].Therefore,GA-sensitive dwarfing genes could be a better alternative to GA-insensitive Rht-1 alleles in dry environments and under stubble-retained conditions.

GA-sensitive dwarfing gene Rht14 in Castelporziano,a short-culm mutant of durum cultivar Cappelli,was generated by X-ray irradiation and thermal neutrons.Castelporziano had a yield and agronomic advantage over Cappelli due to short culms and resistance to lodging[17];and it and other dwarf mutants developed from Cappelli were utilized as parents in European durum breeding programs[7,18].Rht14 was located on chromosome 6A and was predicted to be allelic to GA-sensitive dwarfing genes Rht16 and Rht18[19].More recently the same region was found to carry GA-sensitive dwarfing gene Rht24 which is present in common wheat[10,11].Rht18 was earlier mapped near centromere of chromosome 6A and was linked with GA2-oxidase encoding gene GA2oxA9[20].The association of GA2oxA9 with Rht18 was also recently confirmed by Ford et al.[21].

In spite of its agronomic importance,very little information is available about map position of Rht14 as well as its effect on early seedling vigor and plant stature.In the present study,the effects of Rht14 on coleoptile length,seedling shoot length,plant height,internode length and spike length were compared with those of Rht-B1b in a RIL population.The map position of Rht14 was compared with those of other dwarfing genes on chromosome 6A.

2.Material and methods

2.1.Plant material

Seeds of cultivars Castelporziano(Rht14),Bijaga Yellow(Tall)and HI 8498(Rht-B1b)were obtained from the germplasm collection maintained at the Agharkar Research Institute at Pune,and durum cultivar Cappelli PI264949 was obtained from the National Small Grain Collection,US Department of Agriculture,USA.Castelporziano was crossed with Bijaga Yellow and HI 8498 and two sets of recombinant inbred lines(RIL)were generated.The Bijaga Yellow/Castelporziano population(n=233)was used for mapping the RHT14 locus,whereas the HI8498/Castelporziano population(n=144)was used to compare effects of Rht14 and Rht-B1b on plant height,and lengths of internodes,spikes,coleoptiles and seedling shoots.Ninety accessions of durum and bread wheat were used to examine the allelic frequencies of markers linked to Rht14.

2.2.Phenotype analysis

The Bijaga Yellow/Castelporziano RIL population was grown in crop seasons of 2013-2014,2014-2015,and 2017-2018,and the HI 8498/Castelporziano RIL population was grown in the five consecutive seasons 2013-2014 to 2017-2018.Both populations were sown in replicated trials in the first fortnight of November at the experimental farm of Agharkar Research Institute,Pune(18°31′N,73°55′E).Plots were irrigated 42 days after sowing when most lines were at the booting stage.Plant height was measured at physiological maturity as length of the main tiller from the soil surface to the top of the ear(excluding awns).RILs from the HI 8498/Castelporziano population segregating for Rht14 and Rht-B1b were evaluated for internode length,coleoptile length and seedling shoot length to compare the effects of both genes.Three main culms from each line from the HI 8498/Castelporziano RIL population were harvested at physiological maturity for measurement of spike and internode lengths.Seedlings were grown on germination paper[22]with nutrient solution[23].Coleoptile and early seedling leaf lengths of 10 seedlings per line were measured at Zadoks growth stage 11(first leaf emerged).Sensitivity of parental genotypes to exogenous gibberellic acid(GA3)was assayed by growing seedlings in nutrient solution containing 100 μmol L-1GA3.

2.3.Mapping RHT14

Seventy four SSR markers fromchromosome 6Awere tested for polymorphism between parentalgenotypes Castelporziano and Bijaga Yellow.Polymorphic markers were used to generate genotypic data for 233 RILs and a genetic linkage map of chromosome 6A was generated by MapMaker 3.0[24].Rht24-linked marker TaAP2[10]was used to compare the map positions of RHT24 with RHT14.Similarly,GA2oxA9 linked with RHT18 was also mapped using a co-dominant marker reported earlier[20].Since RHT-B1 and RHT14 were segregating in the HI 8498/Castelporziano population,a gene-specific marker for Rht-B1b[25]and closely linked marker to Rht14(barc118)were used for genotyping and single marker regression analysis was carried out to estimate the individual effects of Rht-B1b and Rht14 on plantheightas well as,internode,spike,coleoptile and seedling shoot lengths.

2.4.Sequence comparison of GA2oxA9

Considering the chromosomal location of GA2oxA9 and its association with Rht14 and Rht18,the nucleotide sequences of GA2oxA9 were obtained for Castelporziano(GenBank accession no.MH688158)and its wild-type parent Cappelli(GenBank accession no.MH688159).Deduced peptide sequences were compared with recently reported peptide sequences of GA2ox s in bread wheat[26]and durum[20,21]by phylogenetic analysis.Peptide sequences of various GA2ox s were obtained by BLAST search of the Rice Genome Annotation Database(http://ricegaas.dna.affrc.go.jp/),NCBI database(http://www.ncbi.nlm.nih.gov/BLAST/)and Phytozome Plant Genome Resource(http://phytozome.jgi.doe.gov/).The peptide sequences were aligned using MUSCLE[27].Positions containing gaps were eliminated and a phylogenetic tree was constructed using MEGA6[28]by the neighbor-joining method[29].Bootstrapping was performed with 1000 replicates and the evolutionary distances were computed using the Poisson correction method.TaGA1ox-B1 was used as an outlier for rooting the phylogenetic tree.

2.5.Statistical analysis

All the phenotype values reported are arithmetic means and standard deviations calculated using SPSS 11.0.Single marker regression analysis was carried out using the linear regression modelin SPSSpackage version 11.0.The differences in phenotype due to GA3treatment and dwarfing genes were tested for significance by Student's t-testassuming two tailed distributions.Genotypic(σ2g)and phenotypic variances(σ2p)were estimated as σ2g=(MSv-MSe)/r and σ2p= σ2g+ σ2e,respectively,and were used further to calculate heritability(h2= σ2g/σ2p),where r is number of replications,MSvand MSeare mean sums of squares for genotype and residual error,respectively.

3.Results

3.1.Phenotype evaluation

Fig.1-Effect of exogenous GA3(100 μmol L-1)on(a)coleoptile length and(b)seedling shoot length in different genotypes at seedling stage(Z11).Standard deviations are represented by whiskers.

Details of phenotypic variation for plant height,internode length,spike length,coleoptile and seedling shoot length are given in Table 1.Heritability(0.83 to 0.97)of plant height was high in both populations.The Bijaga Yellow/Castelporziano RIL population showed a bimodal distribution in each of the over three years of testing(Fig.S1),suggesting that height was governed by single major gene.The population was clearly divided into two classes,a tall group with height of＞85 cm and a semidwarf group with height of＜85 cm.Genotypes falling into these groups were used as binary data to map Rht14 in the Bijaga Yellow/Castelporziano population.In HI 8498/Castelporziano population,heritability was high for internode length and moderate for spike length.In addition to the high heritability and genotypic variance for all traits,environment and genotype×environment interaction was also significant(Table 2).The GA-sensitivity assay using exogenous GA3(100 μmol L-1)gave a 24%increase in coleoptile length when applied to Castelporziano seedlings but no effect when applied to HI 8498 as expected(Fig.1).

3.2.Mapping RHT14 in the RIL population

Genotype data were generated for 233 RILs from the Bijaga Yellow/Castelporziano population using specific markers for TaAP2 and GA2oxA9 as well as 17 polymorphic SSR loci. Rht14 was mapped to an interval of 0.8 cM on chromosome 6A flanked by markers GA2oxA9 and wmc753 (Fig. 2). Regression analysis suggested that Rht14 reduced plant height by 27 cm(R2= 0.80; P ＜ 0.001) (Table 3). The allelic frequencies of markers wmc753, GA2oxA9 and S470865SSR4 closely linked to Rht14 were tested on a set of 90 diverse durum and common wheat accessions.The combination of the 338 bp allele of wmc753 and 204 bp allele of S470865SSR4 differentiated Castelporziano(Rht14)from all accessions except HI 8498 and Icaro(Table S1).The location of RHT14 was identified by aligning sequences of the flanking markers with the IWGSC wheat reference genome sequence v1.0.RHT14 was located in the genomic region between 383 Mbp and 422 Mbp.Two recombination events each were observed between Rht14 and the GA2oxA9-specific marker and between Rht14 and the TaAP2 marker.Since GA2oxA9 is completely associated with Rht18,identification of recombinants revealed that Rht14 cannot be allelic to Rht18.Phenotyping and genotyping errors were unlikely as the plant heights of both recombinants(69.2±3.6 cm and 77.6±4.2 cm)were consistent over three years and the genotype at GA2oxA9 was also confirmed twice for identified recombinants.

3.3.Effect of Rht14 on plant height,internode length and seedling vigor

Fig.2-Dwarfing genes mapped on chromosome 6A in wheat.a)Position of RHT14 on the genetic linkage map is derived from Bijaga Yellow/Castelporziano RILs.b)Comparison of the position of RHT14 with other dwarfing genes on the physical map derived from the IWGSC reference sequence v1.0.

Table 1-Plant height,internode length,spike length,coleoptile length and seedling shoot length observed in Bijaga Yellow/Castelporziano and HI 8498/Castelporziano RILs.

Table 2-Sources of variance in plant height,internode length and spike length for the Bijaga Yellow/Castelporziano and HI 8498/Castelporziano RIL populations.

The HI 8498/Castelporziano population segregating at the RHT14 and RHT-B1 loci was examined to compare effects of the semidwarfing alleles on plant height,internode length,spike length,coleoptile length and seedling shoot length.Regression analysis showed that the reductions in plant height due to Rht14(R2=0.29;P＜0.001)and Rht-B1b(R2=0.37;P＜0.001)were significant and that the magnitudes of reduction were similar at about 25%(Table 3).All internode lengths were significantly reduced by both alleles(P＜0.001).However,the patterns of variation were different;the phenotypic variation explained by Rht14 was higher for upper internodes compared to lower internodes,and the reverse was true for Rht-B1b.The results showed that reduction due to Rht14 was 28.1%at internode-1 and 21.9%at internode 4,whereas the comparative values for Rht-B1b were 21.8%at internode 1 and 34.3%at internode 4(Table 4).RILs carrying Rht-B1b showed a significantly shorter Internode-4 than RILs carrying Rht14.Coleoptile length and seedling shoot length were significantly reduced by Rht-B1b(P＜0.001),whereas Rht14 did not affect either trait.RILs carrying both the genes also showed significant reduction in coleoptile length and seedling shoot length,however this effect was mainly attributed to Rht-B1b.Distributions of plant height,coleoptile length,seedling shoot length,spike length and internode length for Tall,Rht14,Rht-B1b,and Rht14+Rht-B1b genotypes are depicted as box-and-whisker plots in Fig.3.

One of the boys in the class went to a junior executive in a nearby company and honored him for helping4 him with his career planning. He gave him a blue ribbon and put it on his shirt. Then he gave him two extra ribbons, and said, We re doing a class project on recognition, and we d like you to go out, find somebody to honor, give them a blue ribbon, then give them the extra blue ribbon so they can acknowledge a third person to keep this acknowledgment ceremony going. Then please report back to me and tell me what happened.

3.4.Sequence comparison of GA2oxA9

GA2oxA9 was reported to be associated with GA-sensitive dwarfing gene Rht18[20,21].Rht14 is also located in the same genomic region on chromosome 6A,and also linked with GA2oxA9,although 2 recombinant RILs were identified in this study.Therefore,GA2oxA9 was investigated for sequence variation in Castelporziano and Cappelli.A GA2oxA9 nucleotide sequence of about 3935 bp containing 3 exons,2 introns and partial UTR regions was derived.The gene sequences in Castelporziano,Cappelli and HI 8498 were identical and varied from that of Bijaga Yellow by a few SNP.These sequences were aligned with GA2oxA9 sequences from Icaro durum,T.aestivum,Oryza sativa,Brachypodium distachyon,Zea mays,Setaria italica,Sorghum bicolor,and Arabidopsis thaliana.The aligned amino acid sequences consisted of two distinctgroups.Group I contained GA2ox members TaGA2ox-A12,TaGA2ox-B13,wheat homoeologs of TaGA2ox6-11 and TaGA2ox6-9,which inactivate C20GA precursors.TdGA2ox-A9 from Castelporziano,HI 8498,Icaro and Bijaga Yellow were also in this cluster(Fig.4).Group I also included GA2ox1,GA2ox2,and GA2ox3 from Zea mays and Setaria italica.Group II contained members that inactivate C19GA precursors,including GA2ox1,GA2ox2,GA2ox3,GA2ox4,GA2ox7,GA2ox8,and GA2ox10 from Oryza sativa,Brachypodium distachyon and homoeologs in wheat.

Table 3-Regression analysis showing the effects of Rht14 and Rht-B1b on plant height,internode length and spike length in the Bijaga Yellow/Castelporziano and HI 8498/Castelporziano RIL populations.

4.Discussion

4.1.Rht14 reduces plant height without affecting seedling vigor

Semidwarf,lodging-resistant phenotypes in cereals played a major role in yield enhancement during the Green Revolution.Many spontaneous and induced semidwarf mutants have been selected in crop species,among which various mutants in wheat and rice have made significant economic impact.GA-sensitive semidwarfing gene Rht14 in Castelporziano was created by thermal neutron irradiation of durum cultivar Cappelli.Castelporziano had a 17%to 26%yield improvement due to lodging resistance imparted by Rht14.The gene was considered to have high breeding value[7,17]and was extensively used in durum improvement programs[18].In the present study a HI 8498/Castelporziano RIL population was used to compare effects of Rht14 and Rht-B1b on coleoptile and seedling shoot length,as well as internode length.Significant reductions in coleoptile and seedling shoot length were conferred by Rht-B1b,but those traits were unchanged in lines carrying Rht14.These results support earlier findings indicating that GA-sensitive semidwarfs could be alternatives to the Rht-1 alleles in dry environments and fields with retained stubble[4,6].

Both Rht14 and Rht-B1b had significant effects on internode length in the HI 8498/Castelporziano RIL population,however,the R2values indicated that the genes differed in the pattern of effect on internode length(Table 3).Internode-1 length was reduced by 28.1%due to Rht14 compared to 21.8%for Rht-B1b,whereas internode-4 length was reduced by 34.3%due to Rht-B1b compared to 21.9%for Rht14.Overall the plant height reductions for both genes were similar.RILs carrying Rht14 showed spike length equivalent to tall RILs.Surprisingly,RILs carrying Rht-B1b had longer spikes than the tall counterparts as well RILs carrying Rht14(Table 4).An earlier comparison of Castelporziano and Cappelli showed that,although Rht14 was associated with reduced spike length,other spike-related traits such as number of spikelets per spike and 1000-grain weight remained unchanged.Moreover,there was a marginal increase in number of grains per spike in Castelporziano compared with HI 8498[17].Although Castelporziano has been utilized as a donor of Rht14 in European breeding programs[18],there has been no effort to incorporate it in elite Indian materials.To the best of our knowledge,Rht-B1b and Rht-D1b are the predominant dwarfing genes used in Indian wheat cultivars [30-32].Markers wmc753 and S470865SSR4 closely linked to Rht14 had rare alleles of 338 bp and 204 bp,respectively,in Castelporziano.The markers should be useful in marker-assisted transfer of the gene in Indian elite tetraploid and hexaploid wheat backgrounds.Introgressions of GA-sensitive dwarfing genes such as Rht18 produced reductions in plant height,and increases in harvest index and grains per spike with no reduction in seedling vigor in hexaploid wheat[15].However,while transferring such genes to hexaploid wheat,attention should be given to selection of the most suitable adapted recurrent parent as the effect of the gene may vary with the genetic background of the recipient genotype[15].Knowledge on the genetics of GA-sensitive semi-dwarfing genes in wheat and their effects on coleoptile length and seedling vigor should also be useful for other cereals such as barley where longer coleoptiles could be useful[33],although barley is known to have stronger seedling vigor than wheat.

Fig.3-Distribution of plant height,coleoptile length,seedling shoot length,internode length and spike length in Rht14,Rht-B1b and double dwarf(Rht14+Rht-B1b)genotypes in the HI 8498/Castelporziano RILs.

Fig.4-Phylogenetic position of GA2oxA9 from Castelporziano among Gibberellin-2-oxidases from T.aestivum,Oryza sativa,Brachypodium distachyon,Zea mays,Setaria italica,Sorghum bicolor,and Arabidopsis thaliana.Gene sequences of GA2oxA9 in Castelporziano and Cappelli were identical.

4.2.Rht14 might not be allelic to Rht18

Rht14 was reported to be allelic to Rht16 and Rht18 and linked to SSR marker barc3 on chromosome 6AS[19].Although,the report provided important findings about chromosomal location of three dwarfing genes,the study was conducted with a moderately sized F2population without phenotyping in F3.In the present study,a more precise map position of Rht14was established by using more markers and phenotypic data recorded over three years in a population of 233 RIL.In our earlier study,Rht18 cosegregated with GA2oxA9 and mapped within an interval of 1.8 cM representing a genomic region between 412 and 445 Mbp on chromosome 6A[20].Rht18 semidwarfism in Icaro was due to increased expression in GA2oxA9 and decreased levels of bioactive gibberellins[21].It was hypothesized that Rht14 was allelic to Rht18 based on overlapping map positions,a test of allelism conducted on an F2population[19],and a similar expression profile of GA2oxA9

in Icaro and Castelporziano[21].However,two recombination events were identified between Rht14 and GA2oxA9/Rht18 in the Bijaga Yellow/Castelporziano RIL population.Thus the genes might not be allelic.In the present study,Rht14 was mapped in the genomic region spanning 383 to 422 Mbp,overlapping with the genomic region of Rht18.However,sequences of the ORF and flanking regions of GA2oxA9 showed no evidence of deletion or insertion that might be expected following thermal neutron mutagenesis.These sequences were identical in Castelporziano,Cappelli,Icaro and HI 8498 with just a few SNPs separating them from Bijaga Yellow.Similar results reported by Ford et al.[21]indicated that the coding and flanking regions of GA2oxA9 were highly conserved in Icaro(Rht18)and its wild-type parent Anhinga.Ford et al.identified a fast neutron radiation-specific deletion at 5 Mbp upstream of the predicted start site of GA2oxA9[21].These results suggest that the point of mutation that modulates plant height through differential expression of GA2oxA9 in Castelporziano could be a few Mbp from the actual GA2oxA9 allele;however,this needs to be tested with deep sequencing and fine mapping of the region.The same genomic region also harbors GA-sensitive dwarfing gene Rht24 in bread wheat[10,11].The TaAP2 marker linked to Rht24 also showed two recombinants relative to Rht14.However,TaAP2 is an Rht24-linked marker and not a genespecific marker;hence the recombination events between Rht14 and TaAP2 could not be treated as evidence of recombination between Rht14 and Rht24.

Table 4-Comparison of effects of Rht14 and Rht-B1b on plant height,internode length,seedling shoot length,and coleoptile length in the HI 8498/Castelporziano RIL population.

5.Conclusion

The GA-sensitive semidwarfing allele Rht14 conferred a height reduction comparable to the ubiquitously exploited dwarfing gene Rht-B1b.Unlike Rht-B1b,Rht14 does not affect coleoptile length and seedling shoot length.It can therefore be utilized as an alternative semidwarfing gene to Rht-B1b for deep sowing conditions or for fields with retained stubble.Putative recombination events between Rht14 and GA2oxA9/Rht18 indicated that Rht14 might not be allelic to Rht18.Considering the presence of at least four GA-sensitive Rht genes in the 383 to 450 Mbp region of chromosome 6A there is potential for diversification of semidwarfing genes in breeding programs,and particularly for overcoming the problem of short coleoptiles and poor seedling establishment of widely deployed cultivars carrying Rht-B1b.

Acknowledgments

Financial support by the Science and Engineering Research Board,New Delhi under a Start-Up Research grant for Young Scientists(SB/FT/LS-243/2012)to Ravindra Patil is gratefully acknowledged.A Junior Research Fellowship by the Department of Biotechnology,New Delhi to Parimal Vikhe is also gratefully acknowledged.We thank Dr.M.D.Bhagwat for initiating development of mapping populations and useful suggestions.

Conflict of interest

Authors declare that there are no conflicts of interest.