c© Indian Academy of Sciences

RESEARCH ARTICLE

Nucleotide diversity and phylogenetic relationships among Gladioluscultivars and related taxa of family Iridaceae

NIRAJ SINGH1, BALESHWAR MEENA1, ASHISH KUMAR PAL1, ROOP KUMAR ROY1, SRI KRISHNA TEWARI1,SUSHMA TAMTA2 and TIKAM SINGH RANA1∗

1CSIR-National Botanical Research Institute, Rana Pratap Marg, Lucknow 226 001, India2Department of Botany, D. S. B. Campus, Kumaun University, Nainital 263 001, India

AbstractThe plastid genome regions of two intergenic spacers, psbA–trnH and trnL–trnF, were sequenced to study the nucleotidediversity and phylogenetic relationships among Gladiolus cultivars. Nucleotide diversity of psbA–trnH region was higher thantrnL–trnF region of chloroplast. We employed Bayesian, maximum parsimony (MP) and neighbour-joining (NJ) approachesfor phylogenetic analysis of Gladiolus and related taxa using combined datasets from chloroplast genome. The psbA–trnH andtrnL–trnF intergenic spacers of Gladiolus and related taxa-like Babiana, Chasmanthe, Crocus, Iris, Moraea, Sisyrinchium,Sparaxis and two out group species (Hymenocallis littoralis and Asphodeline lutea) were used in the present investigation.Results showed that subfamily Iridoideae have sister lineage with subfamily Ixioideae and Crocoideae. H. littoralis andA. lutea were separately attached at the base of tree as the diverging Iridaceae relative’s lineage. Present study revealed thatpsbA–trnH region are useful in addressing questions of phylogenetic relationships among the Gladiolus cultivars, as theseintergenic spacers are more variable and have more phylogenetically informative sites than the trnL–trnF spacer, and therefore,are suitable for phylogenetic comparison on a lower taxonomic level. Gladiolus cultivars are extensively used as an ornamentalcrop and showed high potential in floriculture trade. Gladiolus cultivation still needs to generate new cultivars with stablephenotypes. Moreover, one of the most popular methods for generating new cultivars is hybridization. Hence, information onphylogenetic relationships among cultivars could be useful for hybridization programmes for further improvement of the crop.

[Singh N., Meena B., Pal A. K., Roy R. K., Tewari S. K., Tamta S. and Rana T. S. 2017 Nucleotide diversity and phylogenetic relationshipsamong Gladiolus cultivars and related taxa of family Iridaceae. J. Genet. 96, xx–xx]

Introduction

The genus Gladilous L. is comprised of about 265 species,and is one of the largest genera of family Iridaceae. The Capeof Good Hope (South Africa) is considered to be the centre ofdiversity for the genus Gladiolus. It is distributed throughoutthe region of tropical Africa, Madagascar, Arabian Peninsula,the Mediterranean basin, Europe and Asia, including Iranand Afghanistan (Goldblatt and Manning 1998; Goldblattet al. 2001). The basic chromosome number is x = 15, rang-ing from diploid (2n = 30) to hypododecaploid (2n = 180),(Bamford 1935; Ohri and Khoshoo 1983). Hybridization andpolyploidy have been greatly responsible for the evolutionof Gladiolus (Ohri and Khoshoo 1983). Gladiolus is out-breeding in nature, and exhibited diverse pollination mech-anism (Ohri and Khoshoo 1983; Goldblatt and Manning

∗For correspondence. E-mail: ranatikam@gmail.com; ranats@nbri.res.in.

1998). Gladiolus cultivars possess enormous diversity incolours, sizes, textures, numbers and types of flowers andforms of inflorescence (Buch 1978; Ohri and Koshoo 1985b;Anderson and Park 1989).

In India, Gladiolus is primarily grown in the states of UttarPradesh, Uttarakhand, Himachal Pradesh, Haryana, Delhi,Karnataka, Punjab, West Bengal, Assam, Sikkim and Megha-laya, having subtropical climates. Artificial hybridization andselection for desired phenotypes have already produced largenumber of cultivars. Phenotypic characters have been widelyused to characterize and identify the cultivars; however, it isoften difficult and even misleading in case of specific culti-vars. Therefore, molecular characterization of different cul-tivars of Gladiolus is very significant in understanding thegenetic relationships among the cultivars and its close rel-atives in the family Iridaceae. Characterization of differentcultivars/genotypes using molecular markers will not onlyhelp in establishing the relationship and affinities amongvarious taxa, but will also pave the ways for selection of elite

Keywords. Gladiolus; Iridaceae; plastid genome; phylogenetic analyses.

Journal of Genetics, DOI 10.1007/s12041-017-0755-1

Niraj Singh et al.

germplasm from large sets of parental genotypes, and couldfurther be used in broadening the genetic base of Gladiolusthrough breeding (Pragya et al. 2010). Previously, to studythe variability, Gladiolus cultivars were investigated usingRAPD (Takatsu et al. 2001; Jingang et al. 2006; Pragyaet al. 2010), ISSR (Jingang et al. 2008; Raychevaet al. 2011) and AFLP (Ranjan et al. 2010). It is well knownthat in most angiosperm species, the nuclear genome isbiparentally inherited and can spread by pollens and seeds,whereas chloroplast and mitochondrial genome is maternallyinherited and spread only by seeds (Greiner et al. 2015).However, there is no comprehensive study available on thecharacterization of varied Gladiolus cultivars using chloro-plast DNA regions like psbA–trnH and trnL–trnF intergenicspacers. However, plastid DNA regions, rbcL, rps4, trnLintron, trnL-F, matK and rps16 have been used to inferphylogenetic analysis and molecular systematics of fami-ly Iridaceae (Souza-Chies et al. 1997; Reeves et al. 2001;Goldblatt 2001). The chloroplast genome has been exten-sively used for phylogenetic reconstructions in plant system-atics due to its relatively small size and conservative modeof evolution. In the present study, we aimed to estimate thenucleotide diversity and establish the phylogenetic relation-ships among Gladiolus cultivars and related taxa of familyIridaceae using two plastid genome markers.

Material and methods

Plant materials and isolation of genomic DNA

In the present study 66 Gladiolus cultivars were procuredfrom the Botanic Garden of CSIR-National BotanicalResearch Institute (CSIR-NBRI), Lucknow, India. Totalgenomic DNA was extracted from young fresh leaves ofall the collected genotypes following CTAB method (Doyleand Doyle 1990). Quantitation of purified DNAs was car-ried out by using NanoDrop ND-1000 Spectrophotometer(NanoDrop Technologies, Wilmington, USA).

PCR amplification and sequencing

The primer sequences of psbA (5′-GTTATGCATGAACGTAATGCTC-3′), trnH (5′-CGCGCATGGTGGATTCACAAATC-3′), trnL (5′-AAAATCGTGAGGGTTCAAGTC-3′), trnF(5′-GATTTGAACTGGTGACACGAG-3′) that are availablein the public domain (Shinozaki et al. 1986; Taberlet et al.1991) were custom synthesized from Sigma Aldrich Chem-icals, India. The amplification of chloroplast region psbA–trnH and trnL–trnF intergenic spacers was performed in 20μL reaction volume with the following components: 10 μL2× phusion master mix, 1 μL (10 pmol) forward primer,1 μL (10 pmol) reverse primer, 2 μL genomic DNA (40ng) and water 6 μL (protease, DNase, RNase free) usingProflex PCR System (Applied Biosystems, Life Technolo-gies, USA). After initial denaturation at 98◦C for 30 s,each cycle consisted of 10 s denaturation at 98◦C, 20 s of

annealing at 54◦C, 30 s of extension at 72◦C and a final 5 minextension at 72◦C at the end of 35 cycles.

The amplified products obtained were electrophoresed on0.8% agarose gels in 1x TBE buffer at a constant voltage 5V/cm. After electrophoresis the gel was stained in ethidiumbromide, visualized and archived using gel documentationsystem (UV Technology, UK). A known DNA ladder of 100bp differences was also loaded on the first well of the gels.Since the size of psbA–trnH and trnL–trnF regions of Glad-iolus cultivars were known (~600 and 700 bp respectively),the bands of interest were identified with reference to DNA.PCR products eluted from the agarose gel were purified usingQIAquick Gel Extraction Kit (Qiagen, GmbH-Hilden, Ger-many). Quantitation of purified PCR products was carriedout by UV spectrophotometery using NanoDrop ND-1000Spectrophotometer (NanoDrop Technologies, USA).

Sequencing was conducted using BigDye Terminator v3.1Cycle Sequencing Kit (Applied Biosystems, Foster City,USA) and the ABI 3730 (Applied Biosystems). After thecompletion of capillary gel electrophoresis, the fluorescencedata were displayed as an electropherogram with the helpof data collection software (ABI Prism�DNA Sequenc-ing Analysis Software v.5.0). Only sequence data with reli-able read lengths were considered in the present study.The sequences determined for the amplified psbA–trnH andtrnL–trnF regions were uploaded to the EMBL/GenBanknucleotide database for storage and archiving.

Phylogenetic analysis

In the present phylogenetic analyses, a total of 86accessions of psbA–trnH and trnL–trnF intergenic spacersequences, comprising 66 Gladiolus cultivars, 18 accessionsof other taxa representing family Iridaceae (Babiana,Chasmanthe, Crocus, Iris, Moraea, Sisyrinchium andSparaxis) and two out group species H. littoralis and A. luteawere considered. The sequences of interest available in theNCBI database were downloaded to use as a supplement inthe phylogenetic study of Gladiolus cultivars. The names ofthe taxa downloaded from GenBank are presented in table 1,along with their accession numbers.

The chloroplast sequences were aligned with Clustal W(Thompson et al. 1994) using Bio Edit Sequence AlignmentEditor (Hall 1999), MEGA6 software (Tamura et al. 2013)and MUSCLE (Edgar 2004). The alignment of sequenceswas further examined and as well manually edited. Theboundaries of the psbA–trnH regions (psbA gene, psbA–trnHintergenic spacer, trnH gene) and trnL–trnF region (trnLgene, trnL–trnF intergenic spacer, trnF gene) were deter-mined by aligning with the existing sequences reported inthe database as a reference (Valente et al. 2011). To inferrelationships between the Gladiolus cultivars, a phyloge-netic tree was constructed using the NJ and MP methodsas implemented in the MEGA 6 (Tamura et al. 2013), androbustness of each internal branch in NJ tree was evaluatedwith 1000-bootstrap replicates.

Journal of Genetics

Nucleotide diversity in Gladiolus cultivars

Table 1. Details of the Gladiolus cultivars and other taxa of family Iridaceae included in the present study.

Vouchar no./source Cultivar name EMBL acc. no. psbA–trnH EMBL acc. no. trnL–trnF

1 253201 Roshni KU879113 KU8791792 253202 Suvarna KU879114 KU8791803 253203 Neelima KU879115 KU8791814 253204 Urvashi KU879116 KU8791825 253234 Amethyst KU879117 KU8791836 253238 Usha KU879118 KU8791847 253242 Sydney Percy-Lancaster KU879119 KU8791858 253243 Acharya JC Bose KU879120 KU8791869 253250 Rashmi KU879121 KU87918710 253205 Friendship Pink KU879122 KU87918811 253206 Friendship White KU879123 KU87918912 253207 Snow Princess KU879124 KU87919013 253208 Vink’s Glory KU879125 KU87919114 253209 Nichole KU879126 KU87919215 253210 Green Wood Pecker KU879127 KU87919316 253211 Eurovision KU879128 KU87919417 253212 Priscilla KU879129 KU87919518 253213 Topaz KU879130 KU87919619 253214 Thamboliana KU879131 KU87919720 253215 Parade KU879132 KU87919821 253216 Classic Pink KU879133 KU87919922 253217 Shagun KU879134 KU87920023 253218 Inter Pearl KU879135 KU87920124 253219 Grock KU879136 KU87920225 253220 White Prosperity KU879137 KU87920326 253221 American Beauty Pink KU879138 KU87920427 253222 Tropic Sea KU879139 KU87920528 253223 Her-Majesty KU879140 KU87920629 253224 Regency KU879141 KU87920730 253225 Oscar KU879142 KU87920831 253226 Peter Pear’s KU879143 KU87920932 253227 Yellow stone KU879144 KU87921033 253228 Tiger Flame KU879145 KU87921134 253229 Aldebaran KU879146 KU87921235 253230 Picardy KU879147 KU87921336 253231 Sylvia KU879148 KU87921437 253232 Praha KU879149 KU87921538 253233 Rose Supreme KU879150 KU87921639 253235 My Love KU879151 KU87921740 253236 Zeus KU879152 KU87921841 253237 American Beauty Blue KU879153 KU87921942 253239 Wine & Rose KU879154 KU87922043 253240 Arka Keshar KU879155 KU87922144 253241 Interpet KU879156 KU87922245 253244 Chanson KU879157 KU87922346 253245 Lavender Puff KU879158 KU87922447 253246 Jester KU879159 KU87922548 253247 Invertation KU879160 KU87922649 253248 Garden Glory KU879161 KU87922750 253249 American Beauty Black KU879162 KU87922851 253251 Red Beauty Spot KU879163 KU87922952 253252 Hallmark KU879164 KU87923053 253253 Red Beauty KU879165 KU87923154 253254 Legeand KU879166 KU87923255 253255 Agni KU879167 KU87923356 253256 Tambri KU879168 KU87923457 253257 Orange Zinger KU879169 KU87923558 253258 Candyman KU879170 KU87923659 253259 Video KU879171 KU87923760 253260 Snow Flower KU879172 KU87923861 253261 Parcifica KU879173 KU87923962 253262 Victor KU879174 KU879240

Journal of Genetics

Niraj Singh et al.

Table 1. (contd)

Vouchar no./source Cultivar name EMBL acc. no. psbA–trnH EMBL acc. no. trnL–trnF

63 253263 Acidanthera KU879175 KU87924164 255486 Heerak KU879176 KU87924265 255487 Lalima KU879177 KU87924366 255488 Fedelio KU879178 KU87924467 NCBI Gladiolus palustris KM887244 KM88731268 NCBI G. illyricus HQ394472 KM88732069 NCBI Babiana sp. HQ394554 GQ98257670 NCBI Babiana torta GQ248252 GQ38201871 NCBI Chasmanthe aethiopica HQ394555 AJ40957272 NCBI Crocus oreocreticus KF886667 HE86426973 NCBI C. sativus KF886664 HE86425174 NCBI C. niveus EU257491 HE86421975 NCBI Iris pseudacorus KC584954 KF17087176 NCBI I. setosa KC704322 KF17088677 NCBI I. odaesanensis KC704309 KF17088078 NCBI I. koreana KF170851 KF17087879 NCBI I. rossii KC704314 KF17087780 NCBI I. minutoaurea KC704306 KF17088281 NCBI Moraea simplex GQ248344 GQ29410882 NCBI M. pallida GQ248343 GQ29409783 NCBI Sisyrinchium angustifolium KC704325 KF17089184 NCBI Sparaxis variegata HQ394559 AJ40958285 NCBI Hymenocallis littoralis KR857492 JX46440186 NCBI Asphodeline lutea JQ039289 AB933517

The MRBAYES 3.1.2 program (Huelsenbeck andRonquist 2001) was employed to sample trees using a BayesianMarkov chain Monte Carlo (B/MCMC) approach. The com-bined dataset was partitioned into the two parts (psbA–trnHand trnL–trnF), and each partition was allowed to have itsown parameters (Nylander et al. 2004). A nucleotide substi-tution model was selected using the Bayesian informationcriterion (BIC) as implemented in the program jModelTest(Posada 2008). The analyses were performed assuming thegeneral time-reversible model of nucleotide substitution(Rodriguez et al. 1990) including estimation of invariantsites and assumption of gamma distribution with six ratecategories and allowing site-specific rates (GTR+I+G) forthe combined analyses. Two parallel runs of two milliongenerations were made, starting with a random tree andemploying 12 simultaneous chains each. Every 100th treewas saved into a file. The first 200,000 generations (i.e. 2000trees) were deleted as the ‘burn in’ of the chains.

We used AWTY (are we there yet?) program (Nylanderet al. 2007) to compare split frequencies in different runs andto plot cumulative split frequencies to ensure that station-ary was reached. Of the remaining 36,000 trees (18,000 fromeach parallel run), a majority-rule consensus tree with aver-age branch lengths was calculated using the sumt option ofMrBayes. The basic sequence statistics, including conservedsites, variable sites, parsimony informative sites, singletonsites were also analysed with MEGA6 software. Averagepairwise differences between sequences for each base pair,nucleotide diversity (π ) (Nei and Li 1979) and haplotypediversity (Hd) were calculated with DnaSP ver. 5.10 (Libradoand Rozas 2009). This analysis was carried out for each DNA

fragment alignment and for a concatenated alignment of allfragments.

Results

Nucleotide diversity of psbA–trnH intergenic spacer regionof chloroplast

The chloroplast sequences of all the taxa were alignedusing Clustal W program in MEGA6 software. The alignedcpDNA sequences of psbA–trnH region of Gladiolus culti-vars varied from 547 to 557 bp with an average length of 555bp. The average frequencies of adenine (A) and thymine (T)contents were 30.3 and 33.6%, respectively, showing 63.9%mean A+T contents in 555 bases average length of psbA–trnH sequence. Similarly, the average frequencies of gua-nine (G) and cytosine (C) were 18.3 and 17.8%, respectively;showing 36.1% mean G+C contents throughout the entirepsbA–trnH sequences. The nucleotide pair frequencies likeidentical pairs (ii), transitional pairs (si), transversional pairs(sv) of psbA–trnH sequences were also calculated (table 2).The average base substitutions recorded in psbA–trnH regionwas 3 (i.e., si = 1, sv = 2). The basic statistics such asconserved sites, variable sites, parsimony informative sitesand singleton sites were calculated after the complete dele-tion of the missing/gap sites from all the sequences (table 2).The aligned psbA–trnH sequences (including psbA gene andpsbA–trnH intergenic spacer) formed in a matrix of 531nucleotide sites, of which 510 were conserved, 21 were vari-ables and eight were parsimony informative and 13 weresingleton.

Journal of Genetics

Nucleotide diversity in Gladiolus cultivars

Table 2. Results of basic sequence statistics, nucleotide diversity, haplotype diversity, nucleotide compositionand nucleotide pair frequencies of psbA–trnH and trnL–trnF sequences of 66 Gladiolus cultivars.

Parameters psbA–trnH trnL–trnF

Length range (bp) 547–557 671–678Total number of sites (excluding sites with gaps / missing data) 531 669Conserved sites 510 664Variable (polymorphic) sites 21 5Singleton variable sites 13 1Parsimony informative sites 8 4Total number of InDel sites analysed 26 11Total number of InDels events analysed 18 7Number of InDel haplotypes 21 9InDel haplotype diversity 0.853 0.516Number of haplotypes (h) 12 10Haplotype gene diversity (Hd) 0.559 0.770Nucleotide diversity {π (10−3)} 3.9 1.8Nucleotide frequencies of adenine (%) 30.3 36.2Nucleotide frequencies of thymine (%) 33.6 29.1Nucleotide frequencies of cytosine (%) 17.8 16.2Nucleotide frequencies of guanine (%) 18.3 18.5Identical pairs (ii) 551 671Transitionsal pairs (si) 1 1Transversional pairs (sv) 2 0

Nucleotide diversity of trnL–trnF intergenic spacer region ofchloroplast

The aligned sequences of trnL–trnF region of Gladiolus cul-tivars varied from 671 to 678 bp with an average length of672 bp. The average frequencies of A and T contents were36.2 and 29.1% respectively; showing 65.3% mean A+Tcontents in 672 bases average length of trnL–trnF sequences.Similarly, the average frequencies of G and C were 18.5and 16.2% respectively; showing 34.7% mean G+C contentsthroughout the trnL–trnF sequences. The average base sub-stitutions recorded in trnL–trnF region was 1 (i.e., si = 1, sv= 0). The aligned trnL–trnF sequences (including trnL geneand trnL–trnF intergenic spacer) formed in a matrix of 678nucleotide sites of which 669 sites were conserved, five siteswere variables, four sites were parsimony informative andone singleton site in Gladiolus cultivars.

Phylogenetic relationships among Gladiolus cultivars

The cumulative phylogeny of chloroplast genome regions(psbA–trnH and trnL–trnF) was inferred using the NJmethod, which resulted in an optimal tree separating all theGladiolus cultivars into eight groups with very weak boot-strap support (BS) (figure 1) group I consisted the cultivarslike Roshni, Picardy, Snow Princess, Her-Majesty, Neel-ima, Tambri, Sydney Percy-Lancaster, Classic Pink, Tham-boliana, Video, Acharya JC Bose and Victor. Group IIclustered together cultivars like Suvarna, Oscar, Hallmark,Friendship Pink, Urvashi, Nichole, Peter Pear’s, AmericanBeauty Black, American Beauty Pink and Chanson. GroupIII comprised Amethyst and Rose Supreme, whereas group IVgrouped Eurovision, Topaz, Priscilla, Zeus, Friendship

White, Invertation and Tropic Sea. Group V grouped togetherTiger Flame, Legeand, Orange Zinger, Interpet, Lalima,Parade, White Prosperity, Garden Glory, Grock, Shagun,Green Wood Pecker and Snow Flower. Group VI clusteredcultivars like Wine & Rose, Red Beauty, Candyman, Paci-fica, My Love and Red Beauty Spot, while group VII clus-tered together Yellow Stone, Aldebaran, Arka Keshar, Jester,Vink’s Glory, Inter Pearl, Regency, Praha and Lavender Puff.Cultivars like Sylvia, Heerak, Rashmi, Agni, Usha, Acidan-thera, American Beauty Blue and Fedelio clustered togetherin group VIII. One close relative and morphologically dis-tinct species Iris rossii used as out group was separatelyattached at the base of the tree as the diverging Gladiolusrelative’s lineage (figure 1).

Phylogenetic relationship within family Iridaceae

Phylogenetic trees were generated using NJ method on thebasis of the evolutionary distances computed using MEGA6program. The relationships and affinities of Gladiolus toother members of family Iridaceae were studied in NJ trees(figure 2). In phylogenetic tree of the family Iridaceae, allthe representative members were separated into two cladeswith strong BS. Clade-I consisted of all the species of Irisbelonging to the subfamily Iridoideae, (I. setosa, I. koreana,I. minutoaurea, I. odaesanensis, I. pseudacorus and I. rossii)with 96% BS, nested with Moraea species (Moraea simplexand M. pallida) with 84% BS and Sisyrinchium angusti-folium with 99% BS. Clade II comprised members of sub-families Crocoideae and Ixioideae, and was further dividedinto IIa and IIb subclade. Subclade IIa consisted membersof Ixioideae, Babiana species nested with Sparaxis var-iegata (61% BS) and Chasmanthe aethiopica (96% BS).

Journal of Genetics

Niraj Singh et al.

Figure 1. NJ clustering based on the sequence variation of the chloroplast DNAregions (psbA–trnH and trnL–trnF), among the 66 Gladiolus cultivars. Numberson nodes are the bootstrap values derived from 1000 replicates.

Journal of Genetics

Nucleotide diversity in Gladiolus cultivars

Figure 2. NJ clustering of Iridaceae based on the combined analysis of the psbA–trnH and trnL–trnFregion. Numbers on nodes are the bootstrap values derived from 1000 replicates.

Journal of Genetics

Niraj Singh et al.

Subclade IIb divided into IIb1 and IIb2. Subclade IIb1clustered together Crocus species of Crocoideae with 76%BS, while subclade IIb2 grouped all Gladiolus cultivarstogether with G. palustris and G. illyricus of subfamilyIxidoideae at 97% BS.

MP showed similar topologies with NJ analysis. However,Crocus species of subfamily Crocoideae grouped togetherwith Babiana and Sparaxis species of subfamily Ixioideaewith moderate BS (65%) (figure 3).

In B/MCMC phylogenetic tree analysis, members of fam-ily Iridaceae formed a strongly supported monophyleticgroup, which further divided into two sister groups withsignificant support (figure 4). Clade I consisted of Iris speciesnested with Moraea and Sisyrinchium species and clade IIconsisted members of Ixioideae and Crocoideae subfami-lies, and was further divided into IIa and IIb subclade. Sub-clade IIa consisted of the member of subfamilies, Crocoideae(Crocus species) and Ixioideae (Babiana species) nestedwith Sparaxis and Chasmanthe species, whereas subclade IIbcomprised member of subfamily Ixioideae.

Discussion

The present study contributes to our understanding ofnucleotide diversity and phylogenetic relationships withinthe Gladiolus cultivars and other related taxa of the fam-ily Iridaceae. Three methods, namely NJ, MP and Bayesiananalysis were used for phylogeny reconstruction. Phyloge-netic relationships were studied using combined dataset of 26variable sites (21 of psbA–trnH and five of trnL–trnF spacer,respectively), which formed eight groups of the Gladioluscultivars with weak BS. Group I consisted of 12 cultivars,of which Roshni (Friendship Pink × Red Beauty), JC Bose(Friendship Pink × Red Beauty) and SP Lancaster (Friend-ship Pink × Priscilla) clustered together revealing affinitiestowards their parentages. Neelima (Snow Princess × TropicSea) showed close affinity with female parentage SnowPrincess, while Suvarna (Fidelio × Hallmark) clustered withmale parent (Hallmark) in group II. Heerak (Yellow Stone× My Love), Usha (Friendship white × My Love) andRashmi (Friendship White × Nichole) clustered togetherin group VIII, whereas Amethyst, Lalima, Urvashi did notshow any affinity to their parentage. Parentage of other Glad-iolus cultivars used in the present study was not known(figure 1). Although Acidanthera is morphologically distinctfrom Gladiolus, it revealed close affinity with Gladiolus cul-tivars in the present investigation. Similar results were alsofound in phylogenetic relationships of Opuntia using ISSRand trnL–trnF intergenic spacer regions (Realini et al. 2015).Present study revealed that psbA–trnH region is more suit-able than the trnL–trnF region in addressing the questionsof phylogenetic relationships among Gladiolus cultivars asthese intergenic spacers are more variable and have morephylogenetically informative sites than the later, and there-fore, are informative for phylogenetic comparison at a lower

Figure 3. MP bootstrap consensus tree of 84 accessions of Iri-daceae and two outgroup, from combined plastid sequence dataanalysis. Numbers on nodes are the bootstrap values derived from1000 replicates.

taxonomic level. The present findings also corroborated withthe earlier phylogenetic investigations of subtribe Sonchinae(Asteraceae) using chloroplast sequence data of psbA–trnHregion (Kim et al. 1999), while Chandler et al. (2001) foundit more informative at higher taxonomic levels, particularly

Journal of Genetics

Nucleotide diversity in Gladiolus cultivars

Figure 4. The evolutionary relationships of Iridaceae based on the combined analysis of the psbA–trnH and trnL–trnFregion. Majority rule consensus tree obtained in the MrBayes phylogenetic analysis.

at the intergeneric level. Evolutionary changes in chloroplastDNA indicated limited variability within species resulting ina lack of intraspecific sampling in studies examining rela-tionships at higher taxonomic levels (Terry et al. 2000). Theevolutionary rate of the nuclear genome is the fastest among

chloroplast and mitochondrial genomes, and the plant mito-chondrial genome is typically characterized by low pointmutation rates (Sloan et al. 2012). It has been establishedthat overall relative rate of substitution in mitochondrial,chloroplast and nuclear genes in 27 species of angiosperm

Journal of Genetics

Niraj Singh et al.

is 1:3:16, respectively (Drouin et al. 2008). Analysis of thetotal cpDNA sequences indicated that psbA–trnH sequenceshad higher nucleotide diversity (4.2 × 10−3) than trnL–trnFsequences (1.78 × 10−3). Similarly, psbA–trnH region hasshown higher nucleotide diversity than atpB–rbcL intergenicregion in Japanese Aucuba (Ohi et al. 2003b).

All the three approaches, namely NJ, MP and Bayesiananalyses were used to analyse the combined dataset producedvirtually identical topologies with high BS (BS > 80 andPP > 0.95) at majority of nodes. Gladiolus was retrieved asmonophyletic in NJ, MP and Bayesian reconstruction meth-ods (BS = 97, BS = 94, PP = 0.99), and these results are incongruence with MP and Bayesian analysis of 150 species ofGladiolus based on five regions of plastid genome (Valenteet al. 2011). Phylogenetic trees generated using NJ analy-sis revealed that Iridoideae have sister lineage with Ixioideaeand Crocoideae. Outgroup species like Hymenocallis lit-toralis and Asphodeline lutea were separately attached atbase of the tree as diverging relative of the family Iridaceae.Bayesian and MP analysis also generated similar topologieswith strong BS at the majority of nodes in different genera ofIridaceae. All three reconstruction methods indicated mono-phyletic origin of Iridaceae, which are in agreement withother phylogenetic analysis of Iridaceae based on chloroplastgenes. The subfamily Crocoideae showed close affinity withsubfamily Ixioideae. The results of the present investigationare supporting the earlier studies on Iridaceae (Souza-Chieset al. 1997; Goldblatt 2001). Gladiolus cultivars are exten-sively cultivated as an ornamental crop having high poten-tial in floriculture trade. There is still lot of demand and needto generate new Gladiolus cultivars with stable phenotypes,and one of the most popular methods is hybridization. There-fore, the information generated in the present investigationabout phylogenetic relationships among different Gladioluscultivars and its relationships with other taxa of the familyIridaceae will provide new insights for hybridization studies.

Acknowledgements

This study was financially supported by Council of Scientific andIndustrial Research, New Delhi under AGTEC (BSC0110) project.

References

Anderson W. and Park R. 1989 Growing Gladioli, pp. 166. TimberPress, Portland, USA.

Bamford R. 1935 The chromosome number in Gladiolus. J. Agri.Res. 51, 945–950.

Buch P. O. 1978 The species. In The world of the Gladiolusedgewood, pp. 2–7. Edgewood Press, Edgewood, USA.

Chandler G. T., Bayer R. J. and Crisp M. D. 2001 A molec-ular phylogeny of the endemic Australiangenus Gastrolobium(Fabaceae: Mirbelieae) and allied genera using chloroplast andnuclear markers. Am. J. Bot. 88, 1675–1687.

Doyle J. J. and Doyle J. L. 1990 Isolation of plant DNA from freshtissue. Focus 12, 13–15.

Drouin G., Daoud H. and Xia J. 2008 Relative rates of syn-onymous substitutions in the mito-chondrial chloroplast and

nuclear genomes of seed plants. Mol. Phylogenet. Evol. 49,827–831.

Edgar R. C. 2004 MUSCLE: Multiple sequence alignment withhigh accuracy and high throughput. Nucleic Acids Res. 32,1792–1797.

Goldblatt P. 2001 Phylogeny and classification of the Iridaceae andthe relationships of Iris. Irises and Iridaceae: biodiversity andsystematics. Ann. Bot. 1, 13–28.

Goldblatt P. and Manning J. C. 1998 Gladiolus in Southern Africa.Fernwood Press, Vlaeberg.

Goldblatt P., Manning J. C. and Bernhardt P. 2001 Radiation of pol-lination systems in Gladiolus (Iridaceae: Crocoideae) in southernAfrica. Ann. Missouri Bot. Gard. 88, 713–734.

Greiner S., Sobanski J. and Bock R. 2015 Why are most organellegenomes transmitted maternally? BioEssays 37, 80–94.

Hall T. A. 1999 BioEdit: a user-friendly biological sequencealignment editor and analysis program for Windows 95/98/NT.Nucleic Acids Sym. Ser. 41, 95–98.

Huelsenbeck J. P. and Ronquist F. 2001 MrBAYES: Bayesianinference of phylogenetic trees. Bioinformatics 17, 754–755.

Jingang W., Hongkun J., Shufang G. and Daidi C. 2006 RAPDAnalysis of 12 general species of Gladiolus hybridus Hort. J.Northeast Agric. Univ. 13, 112–115.

Jingang W., Ying G., Daidi C., Shenkui L. and Chuanpin Y. 2008ISSR analysis of 26 general species of Gladiolus hybridus Hort.J. Northeast Agric. Univ. 15, 6–10.

Kim S. C., Crawford D. J., Jansen R. K. and Santos-Guerra A. 1999The use of a non-coding region of chloroplast DNA in phyloge-netic studies of the subtribe Sonchinae (Asteraceae: Lactuaceae).Plant Syst. Evol. 215, 85–99.

Librado P. and Rozas J. 2009 DnaSP v5: a software for compre-hensive analysis of DNA polymorphism data. Bioinformatics 25,1451–1452.

Nei M. and Li W. H. 1979 Mathematical model for studying geneticvariation in terms of restriction endonucleases. Proc. Natl. Acad.Sci. USA 76, 5269–5273.

Nylander J. A. A., Ronquist F., Huelsenbeck J. P. and Nieves-AldreyJ. L. 2004 Bayesian phylogenetic analysis of combined data. Syst.Biol. 53, 47–67.

Nylander J. A. A., Wilgenbusch J. C., Warren D. L. and SwoffordD. L. 2007 AWTY (Are we there yet?): A system for graphicalexploration of MCMC convergence in Bayesian phylogenetics.Bioinformatics 24, 581–583.

Ohi T., Kajita T. and Murata J. 2003b Distinct geographic structureas evidenced by chloroplast DNA haplotypes and ploidy level inJapanese AuCuba (Aucubaceae). Am. J. Bot. 90, 1645–1652.

Ohri D. and Khoshoo T. N. 1983 Cytogenetics of garden Gladiolus.IV. Origin and evolution of ornamental taxa. Proc. Nat. Acad.Sci. USA 49, 279–294.

Ohri D. and Koshoo T. N. 1985b Cytogenetical evolution of gardenGladiolus. Nucleus 28, 216–221.

Posada D. 2008 jModelTest: phylogenetic model averaging. Mol.Biol. Evol. 25, 1253–1256.

Pragya P., Bhat K. V., Misra R. L. and Ranjan J. K. 2010 Analysisof diversity and relationships among Gladiolus cultivars usingmorphological and RAPD markers. Ind. J. Agric. Sci. 80, 766–772.

Ranjan P., Bhat K. V., Misra R. L., Singh S. K. and Ranjan J. K.2010 Genetic relationships of gladiolus cultivars inferred fromfluorescence based AFLP markers. Sci. Hort. 123, 562–567.

Raycheva T., Stoyanov K. and Denev I. 2011 Genetic diversity andmolecular taxonomy study of three genera from Iridaceae fam-ily in the Bulgarian flora based on ISSR markers. Biotechnol.Biotechnol. Equip. 25, 2484–2488.

Realini M. F., Gonzalez G. E., Font F., Picca P. I., Poggio L.and Gottlieb A. M. 2015 Phylogenetic relationships in Opun-tia(Cactaceae, Opuntioideae) from southern South America.Plant Syst. Evol. 301, 1123–1134.

Journal of Genetics

Nucleotide diversity in Gladiolus cultivars

Reeves G., Chase M. W., Goldblatt P., Rudall P. J., Fay M. F., CoxA. V. et al. 2001 Molecular systematics of Iridaceae: evidencefrom four plastid DNA regions. Am. J. Bot. 88, 2074–2087.

Rodriguez F., Oliver J. L., Marin A. and Medina J. R. 1990 The gen-eral stochastic model of nucleotide substitution. J. Theor. Biol.142, 485–501.

Shinozaki K., Ohme M., Tanaka M., Wakasugi T., Hayashida N.,Matsubayashi T. et al. 1986 The complete nucleotide sequenceof the tobacco chloroplast genome. Plant. Mol. Biol. Report. 4,110–147.

Sloan D. B., Alverson A. J., Chuckalovcak J. P., Wu M., McCauleyD. E. and Palmer J. D. 2012 Rapid evolution of enormous mul-tichromosomal genomes in flowering plant mitochondria withexceptionally high mutation rates. PLoS Biol. 10, 1–17.

Souza-Chies T. T., Britar G., Nadot S., Carter L., Besin E. and Leje-une B. 1997 Phylogenetic analysis of Iridaceae with parsimonyand distance methods using the plastid gene rps4. Plant Syst.Evol. 204, 109–123.

Taberlet P., Gielly L., Pautou G. and Bouvet J. 1991 Univer-sal primers for amplification of three non-coding regions ofchloroplast DNA. Plant Mol. Biol. 17, 1105–1109.

Takatsu Y., Miyamoto M., Inoue E., Yamada Y., Manabe T.,Kasumi M. et al. 2001 Interspecific hybridization amongwild Gladiolus Species of South Africa based on randomlyamplified polymorphic DNA markers. Sci. Hort. 91, 339–348.

Tamura K., Stecher G., Peterson D., Filipski A. and Kumar S. 2013Mega 6: Molecular Evolutionary genetics analysis version 6.0.Mol. Biol. Evol. 30, 2725–2729.

Terry R. G., Nowak R. S. and Tausch R. J. 2000 Genetic variation inchloroplast and nuclearribosomal DNA in Utah Juniper (JuniPe-rus osteosperma, Cupressaceae): evidence for interspecific geneflow. Am. J. Bot. 87, 250–258.

Thompson J. D., Higgins D. G. and Gibson T. J. 1994 CLUSTALW: improving the sensitivity of progressive multiple sequencealignment through sequence weighting, position-specific gappenalties and weight matrix choice. Nucleic Acids Res. 22, 4673–4680.

Valente L., Savolainen V., Manning J., Goldblatt P. and Vargas P.2011 Explaining disparities in species richness between mediter-ranean floristic regions: a case study in Gladiolus (Iridaceae).Glob. Ecol. Biogeogr. 20, 881–892.

Received 11 July 2016, in revised form 1 August 2016; accepted 4 August 2016Unedited version published online: 8 August 2016

Final version published online: 3 March 2017

Corresponding editor: UMESH C. LAVANIA

Journal of Genetics