New chloroplast microsatellite markers suitable for

Annals of Botany Page 1 of 13 doi:10.1093/aob/mcs044, available online at www.aob.oxfordjournals.org

PART OF A HIGHLIGHT ON BREEDING STRATEGIES FOR FORAGE AND GRASS IMPROVEMENT

New chloroplast microsatellite markers suitable for assessing genetic diversity of Lolium perenne and other related grass species Kerstin Diekmann1,2,*, Trevor R. Hodkinson2,3 and Susanne Barth1 1

Teagasc Crops Environment and Land Use Programme, Oak Park Research Centre, Carlow, Ireland, 2School of Natural Sciences, Trinity College Dublin, Dublin 2, Ireland and 3Trinity Centre for Biodiversity Research, Trinity College Dublin, Dublin 2, Ireland * For correspondence. E-mail [email protected] Received: 6 December 2011 Returned for revision: 4 January 2012 Accepted: 30 January 2012

Key words: Lolium perenne, perennial ryegrass, Poaceae, chloroplast microsatellite markers, chloroplast genome, genetic diversity.

IN T RO DU C T IO N Lolium perenne ( perennial ryegrass) is the most important forage grass species of temperate regions. Consequently, genomic sequence data and markers to study its variation are required by several end-users such as plant breeders, biotechnologists and population geneticists who are developing new varieties for the agricultural sector or studying its evolution. The complete nuclear and mitochondrial genomes of Lolium have not been determined but the complete chloroplast ( plastid) genome of one of its cultivars, Lolium perenne ‘Cashel’, has been sequenced by our research group (Diekmann et al., 2008, 2009). Despite this advance, only a few studies have analysed the variability of the chloroplast genome in Lolium (e.g. Balfourier et al., 2000). However, the L. perenne chloroplast genome was sequenced using plant material from a population and, therefore, several single nucleotide polymorphisms (SNPs) were detected

during chloroplast genome assembly. Such polymorphisms clearly offer great potential to develop markers for the study of genetic variation within and among populations of L. perenne and its close relatives. We based our study on Irish L. perenne ecotypes that were mostly sampled from old pasture ecosystems. Old permanent grasslands often contain a large reservoir of genetic diversity in comparison with highly managed, fertilized and reseeded grasslands. To preserve this diversity, an extensive ex situ programme of collection and seed storage was undertaken between 1979 and 1983 by Teagasc (Irish Agriculture and Food Development Authority). A total of 534 sites were sampled and these ecotypes were then propagated under isolation. In 1994 the European Lolium core collection programme was started and 163 different accessions from different gene banks were included in this programme to assess genetic diversity within this species (Connolly, 2000). These two collections formed the plant material source for the present study.

# The Author 2012. Published by Oxford University Press on behalf of the Annals of Botany Company. All rights reserved. For Permissions, please email: [email protected]

Downloaded from http://aob.oxfordjournals.org/ at Trinity College Dublin on July 18, 2012

† Background and Aims Lolium perenne ( perennial ryegrass) is the most important forage grass species of temperate regions. We have previously released the chloroplast genome sequence of L. perenne ‘Cashel’. Here nine chloroplast microsatellite markers are published, which were designed based on knowledge about genetically variable regions within the L. perenne chloroplast genome. These markers were successfully used for characterizing the genetic diversity in Lolium and different grass species. † Methods Chloroplast genomes of 14 Poaceae taxa were screened for mononucleotide microsatellite repeat regions and primers designed for their amplification from nine loci. The potential of these markers to assess genetic diversity was evaluated on a set of 16 Irish and 15 European L. perenne ecotypes, nine L. perenne cultivars, other Lolium taxa and other grass species. † Key Results All analysed Poaceae chloroplast genomes contained more than 200 mononucleotide repeats (chloroplast simple sequence repeats, cpSSRs) of at least 7 bp in length, concentrated mainly in the large single copy region of the genome. Nucleotide composition varied considerably among subfamilies (with Pooideae biased towards poly A repeats). The nine new markers distinguish L. perenne from all non-Lolium taxa. TeaCpSSR28 was able to distinguish between all Lolium species and Lolium multiflorum due to an elongation of an A8 mononucleotide repeat in L. multiflorum. TeaCpSSR31 detected a considerable degree of microsatellite length variation and single nucleotide polymorphism. TeaCpSSR27 revealed variation within some L. perenne accessions due to a 44-bp indel and was hence readily detected by simple agarose gel electrophoresis. Smaller insertion/deletion events or single nucleotide polymorphisms detected by these new markers could be visualized by polyacrylamide gel electrophoresis or DNA sequencing, respectively. † Conclusions The new markers are a valuable tool for plant breeding companies, seed testing agencies and the wider scientific community due to their ability to monitor genetic diversity within breeding pools, to trace maternal inheritance and to distinguish closely related species.

Page 2 of 13

Diekmann et al. — Marker development and genetic diversity assessment of Lolium transferability to other closely related species, and (3) use the markers in studies of interspecific and intraspecific characterization of Lolium and its close relatives. Our search for variable sites within the chloroplast genome of L. perenne resulted in the design of nine new chloroplast markers applicable to a broad range of grass species. Some of them have potential for seed certification institutes because they enable the differentiation of L. perenne and Lolium mutliflorum. Others can be applied for genotyping wild and breeding populations via simple agarose gel electrophoresis without the need for sequencing or automated polyacrylamide gel genotyping systems. M AT E R I A L S A N D M E T H O D S Fourteen Poaceae chloroplast genomes (Table 1) were searched for chloroplast microsatellites using the microsatellite finder tool find_microsat_Win32 (N Salamin, Universite´ de Lausanne, Lausanne, Switzerland, unpubl. res.). The search focused on mononucleotide repeats of seven and more nucleotides that had to be interrupted from each other by at least one nucleotide. The mutation rate of microsatellite regions increases with an increasing number of repeat units (Ellegren, 2004). Therefore, from these data, only mononucleotide repeats of more than 10 bp length were considered for primer design. We preferentially chose microsatellite regions with high variation among species in the alignment. Furthermore, we designed primers so that they would also bind to non-L. perenne chloroplast DNA to enable their application in studies across different grass species. Thus, 14 regions consisting of 17 microsatellites were found (data not shown). Primer sets were then designed for nine regions, amplifying 12 microsatellites (Table 2), using the Primer3 software (http://frodo.wi.mit.edu/). Six primer sets amplified repeats in non-coding regions, three amplified repeats within genes of which only one amplified exclusively a gene region. The amplicon lengths varied from 195 to 658 bp depending on the primer set used. The amplicons were thus considerably longer than the microsatellite repeat of interest. This provided potential for recording other polymorphisms outside the SSR repeat and would allow them to be more broadly applied (e.g. for phylogenetic studies). TA B L E 1. List of Poaceae species that were used in the present studies and their corresponding GenBank accession numbers Poaceae species Agrostis stolonifera Bambusa oldhamii Brachypodium distachyon Dendrocalamus latiflorus Festuca arundinacea Hordeum vulgare Lolium perenne Oryza nivara Oryza sativa ‘japonica group’ Oryza sativa ‘indica group’ Saccharum officinarum Sorghum bicolor Triticum aestivum Zea mays

Accession no. EF115543 FJ970915 EU325680 FJ970916 FJ466687 EF115541 AM777385 AP006728 X15901 AY522329 AP006714 EF115542 AB042240 X86563


We mainly focused our search for highly variable chloroplast genome markers on chloroplast microsatellite regions (chloroplast simple sequence repeats, cpSSRs) because they are known to be highly variable in comparison with other classes of DNA in the chloroplast genome due to slipped strand mispairing that occurs during replication of these regions (reviewed in Kelchner, 2000). These mutations usually result in small insertion/deletion events and are highly suitable as molecular markers (Taberlet et al., 1991; Powell et al., 1995a; Dumolin-Lapegue et al., 1997) that are commonly detected either by direct sequencing or sizing PCR amplicons (genotyping). Due to the overall high conservation of chloroplast DNA sequences among species (Olmstead and Palmer, 1994), primers have been designed to amplify cpSSRs based on sequences of related species if sequence information from the target species is missing. Some universal cpSSR markers have also been developed for angiosperms but these are not always sufficiently variable for chloroplast DNA characterization in Lolium (McGrath et al., 2007). The most productive approach is often to develop markers from genome information sequenced from the target species or at least from closely related taxa. McGrath et al. (2007) showed that plastid SSRs designed specifically from genome data of grasses can be highly variable and can detect genetic variability within populations. Furthermore, publication of the Lolium chloroplast genome (Diekmann et al., 2008, 2009) has provided new sequence information to allow the design of markers specifically for Lolium. As chloroplast genomes show uniparental inheritance and do not recombine during sexual reproduction, different microsatellite loci are linked together and individual haplotypes can be easily detected by applying a set of different chloroplast microsatellite markers (Bryan et al., 1999a). There are many significant applications for chloroplast microsatellite markers in breeding schemes. cpSSRs enable the monitoring of seed-mediated gene flow in angiosperms where the chloroplast genome is generally maternally inherited (Corriveau and Coleman, 1988) and pollen flow in gymnosperms where the chloroplast genome is paternally inherited (Powell et al., 1995a). Thus, cpSSRs in combination with nuclear DNA markers can be used to assess the relative contribution of seed- vs. pollen-mediated gene flow. This is valuable for several applications including risk assessment of transferring transgenes into wild populations (Ryan et al., 2006) or for the detection of parentage in hybrids (Akkak et al., 2007; Atienza et al., 2007), allopolyploids (Hodkinson et al., 2002) and somatic hybrids (Bastia et al., 2001; Bryan et al., 1999b). Powell et al. (1995b) reported that intraspecific chloroplast variation is not random with regard to geographical localization, and thus chloroplast microsatellites are able to determine phylogeographical population structure (Balfourier et al., 2000; McGrath et al., 2007). Furthermore, cpSSRs are well suited to detect population genetic bottlenecks in natural populations and to assess the cytoplasmic diversity and genetic variation that exists in plant breeding material (Provan et al., 2001; Grau Nersting et al., 2006; Fjellheim et al., 2006). Here we aimed to: (1) develop markers for chloroplast genome characterization of L. perenne and investigate their nature and distribution in the genome, (2) test their

Diekmann et al. — Marker development and genetic diversity assessment of Lolium

Page 3 of 13

TA B L E 2. Poaceae universal chloroplast microsatellite primers

Primer*

Sequence

TeaCpSSR27:

Amplicon length (bp)

cpSSR

cpSSR position

Amplified genome region

NCBI accession nos.

390

(A)7g(A)7

33357

atpF intron

HM 172869–HM 172889

419

(A)7c(A)8

36489

rps14-psaB

HM 173009– HM 173027

415

(T)12

49337

ndhK-ndhC

HM 172936–HM 172954

570

(T)10

58739

cemA

HM 172890–HM 172912

658

(T)11

62836

psbE-petL

386

(T)10 (A)10 (T)10

63032 63108 76076

infA-rps8

HM 172913–HM 172935

482

(T)10 (T)11

76094 76655

rps8-rpl14

HM 172984–HM 173008

382

(T)11

80296

rps19-trnH

HM 173028–HM 173057

195

(C)10

93162

rrn16- trnI

HM 173058–HM 173079

FP: AATGCCGAATCGACGACCTA RP: CAATGGTCCCTCTACGCAAT TeaCpSSR28: FP: TGCAATTTTTCTCGCATTTTC RP: TTTCCATTGTGCAAGCAAGA TeaCpSSR29: FP: GGTACCAATCCATAACGATC RP: GCGCTAGTTTTTGTTGTTTT TeaCpSSR30: FP: GGATTAAGAATTGGTGGAATACC RP: AAATGACTACAAGCGAAGGAGA FP: GGTCGTGGAATGCTTTTCTT RP: TCCACGAATCTCAATGACCA TeaCpSSR32: FP: ACGTCCCTTGCTTGAATCAT RP: TCGAGGGATAATGACAGATCG TeaCpSSR33:

HM 172955–HM 172983

FP: TGTCGCAAAGTTGAAACCAA RP: AATCCACTGCCTTGATCCAC TeaCpSSR34: FP: CCCAATTTGCGACCTACCAT RP: AATCCACTGCCTTGATCCAC TeaCpSSR35: FP: GGAGGCCTGCTACGCC RP: TGGAAGTCTTCTTTCGTTTAGGGT *FP, forward primer; RP, reverse primer.

DNA from 16 Irish L. perenne ecotypes, 15 European L. perenne ecotypes, nine L. perenne cultivars, six different Lolium species and 14 other grass species were analysed (Table 3) with up to 15 individuals per population. Total DNA from these individuals was extracted during an earlier project by McGrath et al. (2006, 2007). Thirty-microlitre PCR reactions were set up using 3 mL DNA template, 6 mL 5× PhusionTM HF Buffer (New England Biolabs, Inc., Ipswich, MA, USA), 0.6 mL forward primer (10 mM), 0.6 mL reverse primer (10 mM), 0.6 mL dNTPs (Metabion International AG, Martinsried, Germany) (10 mM), 18.96 mL ddH2O and 0.24 mL PhusionTM Hot Start High-Fidelity DNA Polymerase (New England Biolabs, Inc.). The microsatellite lengths required the application of a polymerase with proofreading ability to avoid potential genotyping errors caused by replication slippage. The PCR programme settings were 98 8C for 5 min, 35 cycles of 98 8C for 1 min, 60 8C for 1 min and 72 8C for 1 min, and finally 72 8C for 10 min. Three microlitres of each PCR product was checked for amplification using 2.5 % MetaPhorw Agarose (Lonza, Rockland, ME, USA) gels. Amplified PCR products were sequenced once using forward primers. Sequencing was outsourced to LGC Genomics (Berlin, Germany) or GATC Biotech AG (Konstanz, Germany). Sequences from each of the different loci were first aligned in MEGA 3.1 (Kumar et al., 2004). To avoid an overestimation of evolutionary events, indels and gaps were replaced by nucleotides that did not appear at that specific position in the

other individuals. Missing data were included by coding each absent nucleotide with a question mark. Sequences from the different primer sets were combined in MEGA 3.1 (Kumar et al., 2004). Haplotypes of the data set were distinguished using the software Arlequin 3.11 (Excoffier et al., 2005) and manually corrected to prevent an overestimation of haplotype number due to missing information. The resulting haplotype sequences were used as an input file for phylogenetic analysis using Bayesian inference with the software MrBayes {settings: nst ¼ 6 and rates ¼ invgamma [¼ GTR + G + I-model (General Time Reversible model + Gammadistributed rate)]; ngen (generations) ¼ 1000 000; samplefreq (samplefrequency) ¼ 100; burnin ¼ 1000} (Huelsenbeck and Ronquist, 2001; Ronquist and Huelsenbeck, 2003). Maximumparsimony bootstrapping, as implemented in PAUP* 4.0 (Swofford, 2002), was also used to test the support of clades; it included 1000 replicates of random addition sequence, TBR branch swapping, and rearrangements limited to 1000 000. The Bayesian tree was modified in FigTree v1.2.1 (Rambaut, 2007; http://tree.bio.ed.ac.uk/) for figure production and bootstrap percentages added to the tree (following Hodkinson et al., 2010). For a locus-by-locus AMOVA, 38 L. perenne populations were included and divided into four groups: (1) L. perenne – Irish ecotypes, (2) L. perenne – European ecotypes, (3) L. perenne – cultivars and (4) Lolium species. The analysis was performed in Arlequin 3.11 (Excoffier et al., 2005) for all groups and then reduced in a stepwise manner (Table 4).


TeaCpSSR31:

Page 4 of 13


TA B L E 3. Names of accessions, their origin, numbers of individuals used in the Lolium perenne diversity study and haplotypes found among Lolium and Festuca populations Haplotypes* Species

Accession number

8 2 8 12 15 15 7 2 7 8 2 9 3 8 6 8

Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland

County

Cork Wicklow Clare Galway Kerry Waterford Cork Galway Tipperary Tipperary Limerick Cork Carlow Roscommon Mayo Wexford

Seed source

Total

Teagasc Oak Park Teagasc Oak Park Teagasc Oak Park Teagasc Oak Park Teagasc Oak Park Teagasc Oak Park Teagasc Oak Park Teagasc Oak Park Teagasc Oak Park Teagasc Oak Park Teagasc Oak Park Teagasc Oak Park Teagasc Oak Park Teagasc Oak Park Teagasc Oak Park Teagasc Oak Park

2 – 3 5 4 3 1 2 3 3 1 3 3 3 4 3

Unique

– 2

1 1

6 8 2 5 7 6 9

Ukraine Norway Romania Romania Greece Denmark Hungary

N/A Teagasc Oak Park Teagasc Oak Park Teagasc Oak Park IBERS IBERS Teagasc Oak Park

3 3 1 2 2 2 2

4 8 7 1 8 8 8 8

Sweden Poland Czech Republic Iran Netherlands Turkey France Wales

Nordic Gene Bank GRIN GRIN GRIN GRIN GRIN GRIN GRIN

1 3 2 – 5 4 1 4

8 5 5 8 2 8 2 2 6

N/A N/A N/A N/A N/A N/A N/A N/A N/A

IBERS Barenbrug Holland BV Teagasc Oak Park Teagasc Oak Park Teagasc Oak Park Alan Stewart NZ DARDNI IPK Gatersleben DARDNI

2 3 3 2 1 3 1 1 2

5 8 7 6 3 7 3

N/A N/A N/A Iran Germany Argentina Iran

IPK Gatersleben N/A N/A GRIN IPK Gatersleben GRIN IBERS

3 3 3 3 1 4 1

1

7

Morocco

IBERS

2

8

Uzbekhistan

GRIN

1 4 2

N/A Hungary Romania

Lochow-Petkus GRIN GRIN

1 –

1

Name

1 – 1 1 1 1 1 1 1 1 2 1 1 1 1 1

2 – 2 2 2 2

– 3 3 7 9

2 10 2

11 12

2 2 2 2 2

13 14 3 3 14

1 1 2 1 1 1 1

4 2

10 13

1 3 1 – 1 1 2 1 1 1 1 1 1 1 2 1 2

–

–

4 8

50

4

2 4 2 2 9 2 – 2 2

15 – 4 4

– 13 14

2

4

10

2 2 2 2

10 3

3

14

– 60

17

3 20 20 50 23 50 9

25 21 21 27

60 22 22 50

9

60

50

2

18

19

–

–

–

–

–

–

–

– – –

– – –

– – –

– – –

– – –

– – –

– – –

1 1

Continued


Lolium perenne – Irish ecotypes Lolium perenne IRL-OP-02007 Lolium perenne IRL-OP-02018 Lolium perenne IRL-OP-02059 Lolium perenne IRL-OP-02078 Lolium perenne IRL-OP-02128 Lolium perenne IRL-OP-02173 Lolium perenne IRL-OP-02192 Lolium perenne IRL-OP-02230 Lolium perenne IRL-OP-02267 Lolium perenne IRL-OP-02269 Lolium perenne IRL-OP-02274 Lolium perenne IRL-OP-02312 Lolium perenne IRL-OP-02337 Lolium perenne IRL-OP-02419 Lolium perenne IRL-OP-02442 Lolium perenne IRL-OP-02491 Lolium perenne – European ecotypes Lolium perenne 21806 Lolium perenne 16-7-62-2Nordic Lolium perenne 3013 Romania Lolium perenne 3199 Romania Lolium perenne ABY-Ba 11478 Lolium perenne ABY-Ba 12896 Lolium perenne IV-51-161 Hungary Lolium perenne NGB 14250 Lolium perenne PI 267059 Lolium perenne PI 321397 Lolium perenne PI 547390 Lolium perenne PI 598445 Lolium perenne W6 11325 Lolium perenne W6 9286 Lolium perenne W6 9339 Lolium perenne – cultivars Lolium perenne ‘Aurora’ Lolium perenne ‘Barlenna’ Lolium perenne ‘Cashel’ Lolium perenne ‘Greengold’ Lolium perenne ‘Magician’ Lolium perenne ‘Manhattan’ Lolium perenne ‘Navan’ Lolium perenne ‘Odenwaelder’ Lolium perenne ‘Portstewart’ Lolium spp. Lolium hybridum GR 11849/49 Lolium multiflorum ‘Multimo’ Lolium multiflorum ‘Nivack’ Lolium persicum PI 229764 Lolium remotum GR 11839/99a Lolium subulatum PI 197310 Lolium ABY-Ba 8917 temulentum Lolium ABY-Ba 13643 temulentum Other species Agrostis PI 439027 stolonifera Avena sativa ‘Evita’ Bromus erectus PI 619490 Cynosurus PI 509441 cristatus

Individuals

Country of origin


Page 5 of 13

TA B L E 3. Continued Haplotypes* Species Festuca arundinacea Festuca ovina Festuca pratensis Festuca pratensis Festuca rubra Festuca vivipara Hordeum vulgare Poa pratensis Secale cereale Triticum aestivum ×Triticosecale

Accession number

Individuals

Country of origin

County

Seed source

Total

Unique

Name

1

N/A

Barenbrug Holland BV

–

–

–

–

–

–

–

PI 634304 ‘Wendelmold’ ‘Northland’ IRL-OP-02174 PI 251118 ‘Regina’ PI 539060 ‘Protector’ ‘Robicum’ ‘Lupus’

6 1 8 5 8 1 4 1 1 1

China N/A N/A Ireland Yugoslavia N/A Siberia N/A N/A N/A

GRIN N/A PGG-Wrightson Teagasc Oak Park GRIN Cebeco Zaden BV GRIN Cebeco Zaden BV CPB Twyford UK Nordsaat Saatzucht GmbH

5 – 1 4 4 – – – – –

4 – 1 4 3 – – – – –

28 – 32 37 28 – – – – –

33 –

34 –

35 –

36 –

38 29 – – – – –

39 30 – – – – –

40 31 – – – – –

– – – – –

* – indicates amplification/sequencing failed due to low quality of DNA.

The input-files had the following header: NbSamples ¼ x [x¼ the amount of samples (¼ populations) used], DataType ¼ DNA, GenotypicData ¼ 0 (¼ haploid data), GameticPhase ¼ 1 (¼ known), LocusSeparator ¼ NONE (¼ each nucleotide will be counted as locus), RecessiveData ¼ 0 (¼ co-dominant data), MissingData ¼ ‘?’. For the analysis, the value for the allowed missing data was set to 0.26 to include all data.

R E S U LT S Characterization of Poaceae chloroplast microsatellites

The chloroplast genomes of 14 different grass species were searched for microsatellite regions. Although this search was limited to mononucleotide repeats with at least seven nucleotides, the number found per species was high, ranging from 241 (Dendrocalamus latiflorus) to 282 (Brachypodium distachyon) microsatellites per genome. The number of microsatellites found in the subfamilies Pooideae and Panicoideae was clearly higher than that found in Ehrhartoideae and Bambusoideae (Fig. 1). The number of microsatellites decreases with an increase of the repeat size in all genomes analysed. Thus, the highest number of microsatellites was found for repeats of seven and the lowest for repeats of more than 11 nucleotides (Fig. 1). Chloroplast microsatellites in Pooideae are mainly based on the nucleotide A, while for the other subfamilies they are relatively equally based on nucleotides A and T. Although the number of published chloroplast genome sequences for Bambusoideae is limited (to only two), it seems that in this subfamily the number of microsatellites based on the nucleotide C is higher than in the other species (Fig. 2). In the chloroplast genome of L. perenne, 78 (33.05 %) of the microsatellites were located within genes and 158 (66.94 %) within intergenic spacers and introns (data not shown). The microsatellites were, in general, distributed over the entire genome; however, repeats with nine and more nucleotides were clearly less abundant in the inverted repeat region

(Fig. 3). Furthermore, it can be seen that repeats based on ten and more nucleotides mainly cluster in four regions of the large single copy region: matK – rpoB, psaA – trnV UAC, infA– trnH GUG, and in the intergenic spacer region between psbE and petL.

Chloroplast microsatellite analysis

The nine newly designed primer sets detected cpSSR polymorphisms to different extents. However, all markers distinguished between L. perenne and non-Lolium taxa. Table 5 gives an overview of marker performance within and among Lolium species. The three most informative markers were TeaCpSSR27, TeaCpSSR28 and TeaCpSSR31. TeaCpSSR28 was able to distinguish between all tested Lolium species and L. multiflorum. The difference between L. multiflorum and the other Lolium species was based on an elongation of the A8 mononucleotide repeat by one nucleotide in L. multiflorum. This was detected via sequencing. TeaCpSSR31 combines a high degree of cpSSR length variation with a considerably high number of SNPs. SNPs detected by this marker happened to be in the same individuals where SNPs had been observed before by TeaCpSSR30 and TeaCpSSR33. Basic agarose gel electrophoresis of amplification products of marker TeaCpSSR27 revealed variation within some L. perenne accessions (Fig. 4) and further sequencing showed that this was due to an indel of 44 nucleotides. All the haplotypes for which SNPs had been observed in the products of TeaCpSSR30, TeaCpSSR31 and TeaCpSSR33 also showed this deletion in the sequencing results.

Phylogenetic and haplotype analysis

Thirty-three haplotypes were detected with up to five haplotypes within a single accession. Few private haplotypes (unique to an accession) were detected. Haplotypes 1 and 2 were the most frequent, of which at least one was found in each L. perenne accession. Most of the non-L. perenne


‘Dovey’

6 15 16

16 n.s.

*** *** ***

*** *** –0.06 1 1 1 9 14 15 3 2 1

1 ¼ Lolium perenne (Irish ecotypes), 2 ¼ Lolium perenne (European ecotypes), 3 ¼ Lolium perenne cultivars. Fixation indices: FST ¼ among groups, FSC ¼ among populations/within groups, FCT ¼ within populations. *** Highly significant (P , 0.001), n.s. ¼ not significant (P . 0.05). ‡

2 23 2+3

1+3

†

50.14 50.41 89.89 49.86 49.59 10.11

0.50 0.50 0.10

0.51 51.41

2 24

– 5.93

54.52

0.49

18 n.s. *** *** –0.01 0.17 83.82

3 2 1– 3 1+2

Lolium perenne Lolium perenne (Irish + European ecotypes) Lolium perenne (Irish ecotypes + cultivars) Lolium perenne (European ecotypes + cultivars) Lolium perenne (cultivars) Lolium perenne (European ecotypes) Lolium perenne (Irish ecotypes)

38 29

– 1.16

17.34

0.16

n.s. n.s. *** *** *** *** –0.02 –0.01 0.28 0.25 73.48 76.32 28.65 25.14 – 2.13 – 1.46

0.27 0.24

Va and FCT Vb and FSC Within populations (Vc) Among populations/ within groups (Vb) Group composition†

Populations Groups

Among groups (Va)

FST‡ FSC‡ FCT‡

Vc and FST

P % Variation

species showed high variation with up to four haplotypes per accession, although the sample sizes were small. The haplotype DNA sequences as well as one sequence of Avena sativa, Triticum aestivum, Secale cereale and Hordeum vulgare were used in an input file for phylogenetic Bayesian inference and parsimony analyses. In both analyses (Fig. 5), Agrostis stolonifera and Avena sativa (tribe Aveneae) were sister to the remaining species (but were not themselves monophyletic) and hence were used as an outgroup to all other species analysed, following Grass Phylogeny Working Group (2001) and Grass Phylogeny Working Group II (2012). High posterior probability and bootstrap values were obtained for Bromus erectus as sister to the Triticeae species (Hordeum vulgare, Secale cereale, Triticum aestivum), S. cereale as sister to T. aestivum, and Poa pratensis as sister to all Festuca and Lolium species as well as to Cynosurus cristatus. High posterior probability values were only obtained for Festuca ovina as sister to a group of Festuca rubra and most Festuca vivipara individuals (but note F. vivipara was not monophyletic) and for all Festucas (except Festuca pratensis) being sister to the Lolium species (Fig. 5). F. pratensis ‘Northland’ was grouped within the L. multiflorum accessions. The haplotypes of ecotype accessions PI267059, IRL-OP-02173, IRL-OP-02078 and IRL-OP-02269, for which variation had already been observed by gel electrophoresis, grouped mainly with the other Lolium species. Locus-by-locus AMOVA

Each accession had missing data for some of the analysed regions due to a lack of amplification or sequencing problems. A locus-by-locus AMOVA, as recommended by Excoffier et al. (2005), was used to test the assigned genetic structure across all nine regions. The number of polymorphic loci decreased from 21 (all four groups) to 16 (only the European L. perenne ecotypes and L. perenne cultivars) (data not shown). No differentiation was found among the different groups, but considerable variation was found within the populations (more than 50 %). The highest variation (83.32 %) was found within populations in the AMOVA of Irish L. perenne ecotypes and L. perenne cultivars. The fixation index, FST, was relatively small at 0.27 for L. perenne populations compared with European populations (FST ¼ 0.49). All the comparisons were highly significant (P ¼ 99.9 %). The AMOVA between the different L. perenne accessions revealed high partitioning of variation within populations of groups 1 and 2 at 76.32 % compared with an among-population variance of 25.14 %. For populations of groups 1 and 3, the within- and among-population variance values were 83.32 and 17.34 % respectively. D IS C US S IO N Characterization of Poaceae chloroplast microsatellites

Fourteen Poaceae chloroplast genomes were searched for mononucleotide repeat regions of more than 7 bp in length. All genomes contained more than 200 cpSSRs. However, the


TA B L E 4. Results of the locus-by-locus AMOVA

21 19

Diekmann et al. — Marker development and genetic diversity assessment of Lolium Polymorphic loci

Page 6 of 13

Page 7 of 13 Total x7 x8 x9 x10 x11

250

150

100

50

Panicoideae

Ehrhartoideae Bambusoideae

Triticum aestivum

Hordeum vulgare

Lolium perenne

Festuca arundinacea

Agrostis stolonifera

Brachypodium distachyion

Dendrocalamus latiflorus

Bambusa oldhamii

Oryza sativa j. group

Oryza nivara

Saccharum officinarum

Sorghum bicolor

Zee mays

0

Pooideae

F I G . 1. Number of microsatellites (more than seven nucleotides) in complete Poaceae chloroplast genomes. x ¼ repeat length (7 to 11). GenBank accession numbers: Agrostis stolonifera (EF115543), Bambusa oldhamii (FJ970915), Brachypodium distachyon (EU325680), Dendrocalamus latiflorus (FJ970916), Festuca arundinacea (FJ466687), Hordeum vulgare (EF115541), Lolium perenne (AM777385), Oryza nivara (AP006728), Oryza sativa ‘japonica group’ (X15901), Saccharum officinarum (AP006714), Sorghum bicolor (EF115542), Triticum aestivum (AB042240), Zea mays (X86563).

total number of cpSSRs varies between species, as found in comparisons made by Powell et al. (1995b). Furthermore, the results show that the number of microsatellites in the different grass subfamilies is not random. Pooideae and Panicoideae have a significantly (P ¼ 99.99 %) higher number of cpSSRs than Ehrhartoideae and Bambusoideae. Ehrhartoideae was represented by rice (Oryza) only and Bambusoideae included two closely related bamboo taxa (Sungkaew et al., 2009). Therefore, it will be interesting to determine if this pattern holds when a more diverse range of ehrhartoids and bambusoids are included, such as taxa from the temperate bamboo tribe Arundinarieae (Hodkinson et al., 2010). Many research groups have found a strong bias of nucleotides A and T for mononucleotide repeats (Powell et al., 1995b; Flannery et al., 2006; Rajendrakumar et al., 2006). Our analysis showed that nucleotides A and T are also favoured across all species but the nucleotide usage seems to be subfamily-specific. Pooideae favours A and Ehrhartoideae favours T (Fig. 2). This observation is, to our knowledge, the first of its kind. However, more Poaceae chloroplast genomes, especially for the subfamilies Ehrhartoideae and Bambusoideae, need to be sequenced to confirm this result. As previously observed in other species (Powell et al., 1995b; Rajendrakumar et al., 2006), the majority (67 %) of

microsatellites were located in the intergenic spacer regions. Most of the cpSSRs were located in the large single copy (LSC) region, as found by Powell et al. (1995b). It is likely that long microsatellites evolve due to slipped-strand mispairing (Kelchner, 2000). The accumulation of microsatellites in the LSC is thus not surprising because a mutation correction mechanism exists in the inverted repeat (IR) region that keeps both IRs potentially identical, and thus the nucleotide substitution is much lower (Wolfe et al., 1987). Chloroplast microsatellite analysis

Our nine newly designed cpSSR primer sets were able to detect polymorphisms across all species tested but only six sets (TeaCpSSR27, TeaCpSSR30, TeaCpSSR31, TeaCpSSR33, TeaCpSSR34, TeaCpSSR35) were able to detect variation at an intraspecific level. This is not surprising given that one objective of this analysis was to design primers that preferably amplify across a wide range of Poaceae species and thus some amount of conservation was required (Taberlet et al., 1991; Dumolin-Lapegue et al., 1997; Provan et al., 2001). However, most primers produced amplicon lengths of 380 bp or more and thus the lack of variation in some loci might be due to the location of the microsatellite rather than to the location of the


Number of mononucleotide repeats


Page 8 of 13

Diekmann et al. — Marker development and genetic diversity assessment of Lolium A T C G

Number of mononucleotide repeats

140

120

100

10

Ehrhartoideae Bambusoideae

Triticum aestivum

Hordeum vulgare

Lolium perenne

Festuca arundinacea

Agrostis stolonifera

Brachypodium distachyion

Dendrocalamus latiflorus

Bambusa oldhamii

Oryza sativa j. group

Oryza nivara

Saccharum officinarum

Sorghum bicolor

Zee mays

Panicoideae

Pooideae

F I G . 2. Nucleotide usage of chloroplast microsatellites (more than seven nucleotides) in Poaceae species. GenBank accession numbers: Agrostis stolonifera (EF115543), Bambusa oldhamii (FJ970915), Brachypodium distachyon (EU325680), Dendrocalamus latiflorus (FJ970916), Festuca arundinacea (FJ466687), Hordeum vulgare (EF115541), Lolium perenne (AM777385), Oryza nivara (AP006728), Oryza sativa ‘japonica group’ (X15901), Saccharum officinarum (AP006714), Sorghum bicolor (EF115542), Triticum aestivum (AB042240), Zea mays (X86563).

primers. Very little intraspecific variation was found for marker TeaCpSSR29, which is located within the ndhK gene, TeaCpSSR32 from the infA gene and TeaCpSSR28 from the intergenic spacer rps14-psaB. This low variation in microsatellites in coding regions is expected because variation in the form of one- to two-nucleotide indels would lead to frameshifts and thus to non-functionality of the respective gene (Metzgar et al., 2000). Marker TeaCpSSR28 is based on an interrupted microsatellite and such interruption seems to have a stabilizing effect (Rolfsmeier et al., 2000). Although primers TeaCpSSR28, TeaCpSSR29 and TeaCpSSR32 were unable to detect intraspecific variation, they were able to detect interspecific variation and will be useful in future studies at that level. TeaCpSSR28 is particularly promising due to its ability to differentiate L. multiflorum from other Lolium species. This ability could be important for seed testing agencies for testing the purity of seed lots. The polymorphism is based on a mononucleotide repeat and can be easily adopted for high-throughput genotyping applications. L. perenne is widely used as turf grass in Europe and the United States (Floyd and Barker, 2002). Contamination of seed lots with L. multiflorum, which has a brighter foliage colour, broader leaves and less tillering, is unwanted. Other markers have been developed for this purpose. For example, Warnke et al. (2002) showed that the locus of the enzyme superoxide

dismutase (Sod-1) can be used for distinguishing L. perenne from L. multiflorum. Warnke et al. (2004) showed that the seedling root fluorescence locus (Pgi-2) and the 8-h flowering characteristics on chromosome 1 (morphological marker that indicates photoperiod insensitivity after vernalization) can also distinguish L. multiflorum from L. perenne and may be of wider value for species identification. Our TeaCpSSR28 marker is a valuable addition to these markers, especially because of its ease of application. Overall, the most informative marker was TeaCpSSR31 because it revealed nearly as much haplotype information as that detectable by using a combination of the other primers. This marker is located in the psbE-petL intergenic region. The psbE-petL region has not previously been characterized in Lolium. Marker TeaCpSSR27 is also a new marker of considerable value. This marker is not variable at the microsatellite region but revealed variation at a 44-bp-long repeat that showed one and two copies, respectively, in four different accessions [IRL-OP-02078 (2× one copy/9× two copies), IRL-OP-02173 (3/12), IRL-OP-02269 (1/4) and PI267059 (1/7)]. This variation was already detectable via agarose gel electrophoresis and thus this primer set is of particular value because the sequencing step for the detection of variation can be omitted. Sequence information obtained for the chloroplast genome of different species and their individuals (when


0

T9

2

rp

A9; T10

rp

T10

A13 tr

A9 A10

psbC psbD

psbN

T9 (3); T10

rpoA rps11 T11 rpI36 infA rps8 T 4 rpI1 T109 * rpI16 T9 rps32 rpI219 T11 r ps * rpI2 A10 3 rpI2 CAU

135 282 bp

T11

trnS GCU

A9

trnQ UUG

A10

U matK

psbA GUG

rps1 9 rpI2

* rpI2 trnI 3 CAU

AA

LC

*

hB

*

trn

nd

LC

’e

SSC

U GU tmN 23 rm

rp rp

*

s7

s1

2

3’

en

d*

U

GU

r m4 tmR rm .5 ACG5

*tr nI nA GA UG U C

hB

trnN

*tr

trn

AA

nd

23 rrn 4.5 rrnrrn5 G AC

trnR

*ndhA

5 rps1 ndhH

A9; A10 F I G . 3. Distribution of microsatellites (more than nine nucleotides) in the chloroplast genome of L. perenne. The two inverted repeat regions are identical to each other. Thus, microsatellites are only highlighted in one of the two copies.

Page 9 of 13

ccsA trnL UAG rpI32

T9

A9 ndhI ndhG ndhE psaC ndhD

ndhF

C10

A9 (2)

AC G 6 V n1 trn rr U GA C nI *rr A UG n *tr

3 s7 rp s12 rp

V G A rrn C 16

nd

rps1 5 ndhH

trn

A9; A10

*rps16

-A

-B

trnI

psbI psbK

*trnK UU

trnH

IR

GUG

A9; A15 T9; C11

GCC trnG psbZ

IR

A16

M C psb D GUUA trn Y G C trn E UU U trn I GG trn

A GC tN pe

UC GU U *trn CA fM trn UGA trnS

Lolium perenne L.

T10 (2)

nC

A9; T9

*petD

trnH

A9; A13

LSC p p sbJ p sbL pssbF bE trn W trn -CC P-U A GG rpI2 0 rps1 clpP 2 5’en d T 9

*petB

T9

oB


oC 1

rp

oC 2

r ps

trnR rps14UCU

A10

A9 ; A3; T9

psaB

A10

psaA

yct3*

J ndh K ndhhC * nd AC VU trn E atp B atp

pe tA

pe pe tL tG psa rpI3 J r ps 3 18

psbB psbT psbH

A9 (2)

p yc saI f4 A

m

T10; A10; T11 A9; T9

cL

ce

A9

rps4 UGU trnT

U CA

rb

T9

A10

F* atp

GA

atpA

trnS G

T9

M

trn

UAA trnL GAA trnF

T12

T9

atp H atp I

C15 A9; T10 T9


Page 10 of 13


TA B L E 5. Ability of newly designed Poaceae universal primers to detect variation within the chloroplast genome of Lolium perenne and other Lolium species (L. hybridum, L. multiflorum, L. persicum, L. remotum, L. subulatum, L. temulentum) Form of variation Primer

Lolium perenne

cpSSR

SNP

Indel

3 3 – 3 3 3 3 3 3

3 – – 3 3 – 3 3 3

– 3 – – 3 – 3 3 3

– – – 3 3 3 3 – –

3 – – – – – – – –

IRL-OP-02078

IRL-OP-02173

20 32 11 34 12 35 1 21 44 42 30 43 3144 20 33 23 25 24 27 10 28

F I G . 4. TeaCpSSR27 PCR products (5 mL per sample) from different individuals of two Lolium perenne accessions. Shorter amplicon products are due to the lack of one copy of a 44-bp repeat. A 2.0 % MetaPhorw Agarose gel (Lonza, Rockland, ME, USA) stained with 1 % ethidium bromide (10 mg mL21). Size standard: mi-100 bp+ DNA Marker Go (Metabion International AG, Martinsried, Germany), run for approx. 3 h at 80 V.

polymorphisms were observed) using the new markers was submitted to GenBank (Table 5) and is publicly available. Phylogenetic and haplotype analysis

Thirty-three different haplotypes were detected in this study and analysed using Bayesian inference and parsimony approaches. Bromus erectus, the Triticeae cereals, Poa pratensis and Cynosurus cristatus were resolved as outlying the majority of Festuca and Lolium accessions. This tree showed high resolution and consistency with current taxonomy despite the

Locus-by-locus AMOVA

The different L. perenne accessions were analysed as four different groups for the locus-by-locus AMOVA (Table 4).


TeaCpSSR27 TeaCpSSR28 TeaCpSSR29 TeaCpSSR30 TeaCpSSR31 TeaCpSSR32 TeaCpSSR33 TeaCpSSR34 TeaCpSSR35

Lolium spp.

fact that most of the variation between the different haplotypes is based on microsatellites and thus potentially influenced by size homoplasy (Hale et al., 2004; Flannery et al., 2006). There was also evidence that the inbreeding groups including L. persicum, L. remotum, L. subulatum and L. temulentum (Kubik et al., 1999; Balfourier et al., 2000) and outbreeding groups of Lolium species could be distinguished from each other. The grouping of inbreeding L. subulatum was an exception to this pattern because the majority of its individuals were positioned close to the outbreeding Lolium species. Unfortunately, six out of the seven L. subulatum individuals showed haplotypes that were not clearly distinguishable from each other, probably due to a large amount of missing data in that species. The positions obtained for L. hybridum accessions are of note. L. hybridum is an interspecific highly fertile hybrid between L. perenne and L. multiflorum that occurs naturally very frequently. In this study, L. hybridum individuals are grouped either within or as sister to L. perenne, while the L. multiflorum individuals are grouped as sister to L. perenne. This reveals L. perenne as maternal genome donor of L. hybridum and L. multiflorum as paternal genome donor in all these analysed accessions. Lolium perenne individuals that lacked one copy of the repeat of the TeaCpSSR27 marker had a haplotype very similar to the inbreeding Lolium species (Fig. 5). Although interspecific hybrids between the different outbreeding Lolium species are possible as mentioned above, interspecific hybrids between in- and outbreeding Lolium species are normally very rare due to post-zygotic hybridization barriers (Matzk et al., 1980). Embryo rescue techniques have to be carried out to obtain viable plants after reciprocal crosses between L. temulentum and L. perenne (Yamada, 2001). Thus, the results obtained with TeaCpSSR27 are surprising because they clearly indicate that some natural hybridization occurred between the two groups. We are investigating this possibility further. The grouping of F. pratensis within the Lolium accessions is also of note. F. pratensis did not group with the other Festuca species. Festuca species can be grouped into two major categories – fine-leaved and broad-leaved (Torrecilla and Catalań, 2002). In phylogenetic analyses, the genus Lolium is found within the broad-leaved Festuca group (Torrecilla and Catalań, 2002; Catalań et al., 2004). F. pratensis and F. arundinacea belong to the broad-leaved category while F. ovina, F. rubra and F. vivipara belong to the fine-leaved category. This is in accordance with the extensive Lolium diversity study of McGrath et al. (2007) that grouped F. pratensis closer to the Lolium species than Festuca species and where the fine-leaved Festuca species were resolved as an outlying group. Darbyshire (1993) suggested the re-circumscription of Festuca. He considered the fine-leaved fescues to be true fescues and transferred F. pratensis and F. arundinacea to the genus Lolium (Lolium pratense and L. arundinacea).

IRL-OP-02007 (3) IRL-OP-02059 (4) IRL-OP-02078 (5) IRL-OP-02128 (2) IRL-OP-02173 (8) IRL-OP-02230 (1) 0·96 IRL-OP-02059 (1); IRL-OP-02078 (1); IRL-OP-02419 (5); IRL-OP-02442 (1); PI267059 (6); Cashel (1); Manhattan (5); Lolium hybridum (1) IRL-OP-02269 (1) 62 IRL-OP-02274 (2) PI267059 (1) IRL-OP-02312 (5) 0·75 IRL-OP-02337 (1) IRL-OP-02267 (1) 0·82 IRL-OP-02419 (2) IRL-OP-02337 (1); IRL-OP-02491 (2); W611325 (2); Manhattan (2) IRL-OP-02442 (1) IRL-OP-02491 (3) IRL-OP-02312 (1); PI598445 (1), 16-7-62-2Nordic (1) ABYBa-12896 (4) PI321397 (6) 1·00 IRL-OP-02078 (1); IRL-OP-02442 (1); ABYBa-11478 (2); PI598445 (1); W69339 (1); W611325 (4); 21806 (2) PI598445 (2) 75 IRL-OP-02128 (1) W69286 (8) W69339 (1) 0·98 IRL-OP-02267 (2); W69339 (1); 21806 (2); Barlenna (1) W611325 (1) Portstewart (1) 3199Romania (2) PI598445 (1); Lolium hybridum (3); Lolium subulatum (5) IV-51-161-Hungary 0·95 0·55 Lolium hybridum (1) (2) 70 Aurora (3) Lolium multiflorum cv. Nivack (5); Lolium multiflorum cv. Multimo (3) 0·63 Barlenna (1) 58 Lolium multiflorum cv. Nivack (1); Lolium multiflorum cv. Multimo (4) Cashel (3) 0·58 Lolium multiflorum cv. Nivack (1); Lolium multiflorum cv. Multimo (1) Greengold (2) Festuca pratensis cv. Northland (8) Navan (2) Portstewart (5) IRL-OP-02078 (2); Lolium subulatum (1); Lolium persicum (5) 0·95 IRL-OP-02173 (3); PI267059 (1); Lolium temulentum ABYBa8917 (3); Lolium subulatum (1) 0·97 IRL-OP-02269 (1) 72 1·00 Lolium temulentum ABYBa13643 (6) Lolium remotum (3) Lolium temulentum ABYBa13643 (1) Lolium persicum (1)

Posterior probability Bootstrap value

IRL-OP-02007 (5) IRL-OP-02059 (3) IRL-OP-02078 (2) IRL-OP-02128 (12) IRL-OP-02173 (4) IRL-OP-02192 (7) IRL-OP-02230 (1) IRL-OP-02267 (4) IRL-OP-02269 (3) IRL-OP-02312 (3) IRL-OP-02337 (1) IRL-OP-02419 (1) IRL-OP-02442 (3) IRL-OP-02491 (3) ABYBa-12896 (1) ABYBa-11478 (5) PI321397 (1) PI598445 (3) W69339 (5) W611325 (1) NGB14250 (4) 21806 (4) 3013Romania (1) 3199Romania (3) IV-51-161-Hungary (7) 16-7-62-2Nordic (2) Aurora (5) Barlenna (3) Cashel (1) Greengold (6) Magician (2) Manhattan (1) Odenwaelder (2) .

Festuca vivipara (2) 0·85 0·51 57 Festuca vivipara (1) 57 Festuca vivipara (2)

1·00

0·60 92 1·00

0·82

0·90 0·64 59 85

Festuca rubra (2) Festuca rubra (1) Festuca rubra (1)

51

0·53 1·00 95

0·92 69

0·79 84

Festuca ovina (1) Festuca ovina (1)

Festuca ovina (3); Festuca ovina (1) Festuca ovina (1) Festuca ovina (2)

1·00

Cynosurus cristatus

89 0·55 63 1·00

Festuca rubra (1) Poa pratensis 1·00 0·75 96 72

100

Triticum aestivum Secale cereale

Hordeum vulgare Bromus erectus

Avena sativa Agrostis stolonifera


Lolium perenne – Irish ecotypes Lolium perenne – European ecotypes Lolium perenne – cultivars Lolium subspecies Festuca subspecies Others

F I G . 5. Bayesian inference tree using DNA sequences from haplotypes. Haplotype names are replaced by the names of the accessions in which they were found. Posterior probability values were obtained in MrBayes and bootstrap values in PAUP* 4.0.

Page 11 of 13


Page 12 of 13


Conclusions

The main objectives of this study were to find new Poaceae chloroplast DNA markers using information about the location of variable regions previously detected in the completely sequenced L. perenne chloroplast genome and to characterize a diverse collection using these markers. Thus, nine new primer sets were designed, all of which successfully amplified

polymorphic regions either in or outside (as indels or SNPs) the target SSRs. The most informative marker was TeaCpSSR31 ( psbE-petL), because it alone detected nearly all the variation otherwise found only with a combination of primers. TeaCpSSR27 was also of high value as it enabled the distinction of haplotypes solely via agarose gel electrophoresis without the need for sequencing or more accurate sizing methods. TeaCpSSR28 was also found to be particularly informative because it could detect L. multiflorum within other Lolium species. This marker, although based on a microsatellite region, might prove useful in barcoding approaches especially when commonly used universal barcoding markers (CBOL Plant Working Group, 2009), such as matK and rbcL, are unable to detect variation among species. The new markers were tested on a small set of different Lolium and Festuca accessions and were able to detect diversity within these accessions. Although the breeding history of L. perenne is rather young (approx. 90 years, Wilkins, 1991), this study shows that existing cultivars derive from a narrow genetic pool. In total, 15– 16 polymorphic loci were found within the ecotype accessions, but only six were found within the cultivars. It is recommended that diversity is increased or at least monitored within breeding material. Markers TeaCpSSR27 and TeaCpSSR31 will prove particularly useful for managing this process. ACK N OW L E DG E M E N T S We thank Choun-Sea Lin who released the two bamboo sequences to us prior to their publication. We thank Sarah McGrath for sharing DNA from L. perenne populations. This work was part of the PhD thesis of the first author (K.D.) and was supported by the Teagasc ‘Vision’ programme. K.D. was financed under the Teagasc Walsh Fellowship Scheme. Partial funding of this work was obtained from the Irish Department of Agriculture, Food and the Marine (DAFM) under the ‘Genetic Resources Fund for Food and Agriculture’. L I T E R AT U R E C I T E D Akkak A, Boccacci P, Botta R. 2007. ‘Cardinal’ grape parentage: a case of a breeding mistake. Genome 50: 325– 328. Atienza SG, Martıń AC, Ramı´rez MC, Martıń A, Ballesteros J. 2007. Effects of Hordeum chilense cytoplasm on agronomic traits in common wheat. Plant Breeding 126: 5– 8. Balfourier F, Imbert C, Charmet G. 2000. Evidence for phylogeographic structure in Lolium species related to the spread of agriculture in Europe. A cpDNA study. Theoretical and Applied Genetics 101: 131– 138. Bastia T, Scotti N, Cardi T. 2001. Organelle DNA analysis of Solanum and Brassica somatic hybrids by PCR with ‘universal primers’. Theoretical and Applied Genetics 102: 1265–1272. Bryan GJ, De Jong W, Provan J, et al. 1999a. Potato genomics: a general strategy for the molecular genetic characterization of Solanum germplasm. Scottish Crops Research Institute Annual Report, 1998/99. http://www.scri.ac.uk/publications/1998/99annualreport (accessed 18 December 2009). Bryan GJ, McNicoll J, Ramsay G, Meyer RC, De Jong WS. 1999b. Polymorphic simple sequence repeat markers in chloroplast genomes of Solanaceous plants. Theoretical and Applied Genetics 99: 859–867. Catalań P, Torrecilla P, Lo´pez Rodrı´guez JA, Olmstead RG. 2004. Phylogeny of the festucoid grasses of subtribe Loliinae and allies (Poeae, Pooideae) inferred from ITS and trnL-F sequences. Molecular Phylogenetics and Evolution 31: 517–541.


The combined Lolium collection showed a low fixation index of 0.27 and a high proportion of within-population variation (73.48 %). When Lolium was split in groups of two (Table 4), the highest among-group variation (76 – 84 %) was detected for the two groups that contained the Irish ecotype accessions. In contrast, the proportion of within-population variation for the group of European ecotypes and cultivars was only approx. 50 %. A fixation index of 0.10 and a within-population variation of approx. 90 % was found for the Irish ecotypes when the three different L. perenne groups were analysed separately from each other in additional AMOVAs. This observation is in agreement with other studies using nuclear markers on L. perenne and other outbreeding grass species. Fjellheim and Rognli (2005) analysed the genetic diversity of 13 F. pratensis cultivars from Denmark, Iceland, Norway and Sweden via amplified fragment length polymorphism markers. Cultivars were selected to represent old and new varieties from each country. The within-cultivar variation was around 80 %. Furthermore, their study showed that the level of genetic diversity was as large in new cultivars as in old ones. No decrease in genetic diversity was recorded in breeding populations of F. pratensis. Guthridge et al. (2001) analysed genetic diversity in L. perenne using two different sets of plant material. The first set, consisting of four cultivars with different breeding background in relation to their genetic base and number of parents, showed within-population variation of 89 %. The second set, based on three cultivars derived from a low number of parents, showed within-population variation of 79 %. Surprisingly, the amount of variation found within the cultivars and European ecotypes used in this study was small compared with the results achieved by other research groups. This observation was also shown by McGrath et al. (2006, 2007) who used chloroplast microsatellite markers to assess the genetic diversity of L. perenne ecotypes and cultivars. They recorded levels of 82 and 18 % for the partitioning of variation within and among Irish populations, respectively. This compared with 61 and 39 % variation within and among European populations, respectively (McGrath et al., 2006, 2007). Although the percentages differ from the results obtained in this study, which might be due to different sample size (and different marker systems), the trend is the same. This result is not surprising as Ireland is an island and thus the possibility of pollen- and seed-mediated gene flow is limited in comparison with the larger European continent. The distance between populations is, in general, shorter and hence differentiation of populations is lower. Only six polymorphic loci were found within the L. perenne cultivars, while there were 15 within the European ecotypes and more than 16 within the Irish ecotypes. Despite the low genetic variation among Irish ecotypes, this result highlights the value of the Irish ecotype collection for future breeding programmes.


Kubik C, Meyer WA, Gaut BS. 1999. Assessing the abundance and polymorphism of simple sequence repeats in perennial ryegrass. Crop Science 39: 1136–1141. Kumar S, Tamura K, Nei M. 2004. MEGA3: integrated software for molecular evolutionary genetics analysis and sequence alignment. Briefings in Bioinformatics 5: 150– 163. Matzk F, Gro¨ber K, Zacharias M. 1980. Ergebnisse von Art- und Gattungskreuzungen mit Gramineen im Zusammenhang mit den natu¨rlichen Isolationsmechanismen zwischen den Arten. Kulturpflanze XXVIII: 257 –284. McGrath S, Hodkinson TR, Salamin N, Barth S. 2006. Development and testing of novel chloroplast microsatellite markers for Lolium perenne and other grasses (Poaceae) from de novo sequencing and in silico sequences. Molecular Ecology Notes 6: 449– 452. McGrath S, Hodkinson TR, Barth S. 2007. Extremely high cytoplasmic diversity in natural and breeding populations of Lolium (Poaceae). Heredity 99: 531– 544. Metzgar D, Bytof J, Wills C. 2000. Selection against frameshift mutations limits microsatellite expansion in coding DNA. Genome Research 10: 72–80. Olmstead RG, Palmer JD. 1994. Chloroplast DNA systematics: a review of methods and data analysis. American Journal of Botany 81: 1205–1224. Powell W, Morgante M, McDevitt R, Vendramin GG, Rafalski JA. 1995a. Polymorphic simple sequence repeat regions in chloroplast genomes: applications to the population genetics of pines. Proceedings of the National Academy of Science USA 92: 7759–7763. Powell W, Morgante M, Andre C, et al. 1995b. Hypervariable microsatellites provide a general source of polymorphic DNA markers for the chloroplast genome. Current Biology 5: 1023–1029. Provan J, Powell W, Hollingsworth PM. 2001. Chloroplast microsatellites: new tools for studies in plant ecology and evolution. Trends in Ecology and Evolution 16: 142–147. Rajendrakumar P, Biswal AK, Balachandran SM, Srinivasarao K, Sundaram RM. 2006. Simple sequence repeats in organellar genomes of rice: frequency and distribution in genic and intergenic regions. Bioinformatics 23: 1 –4. Rambaut A. 2007. FigTree, a graphical viewer of phylogenetic trees. http ://tree.bio.ed.ac.uk/software/figtree/ (accessed 2 October 2011). Rolfsmeier ML, Dixon MJ, Lahue RS. 2000. Mismatch repair blocks expansions of interrupted trinucleotide repeats in yeast. Molecular Cell 6: 1501– 1507. Ronquist F, Huelsenbeck JP. 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: 1572–1574. Ryan E, Mullins E, Burke J, Downes M, Meade C. 2006. Tracing field hybridization in ryegrass species using microsatellite and morphological markers. Environmental Biosafety Research 5: 111 –117. Sungkaew S, Stapleton CMA, Salamin N, Hodkinson TR. 2009. Non-monophyly of the woody bamboos (Bambuseae; Poaceae): a multigene region phylogenetic analysis of Bambusoideae s.s. Journal of Plant Research 122: 95–108. Swofford DL. 2002. PAUP*. Phylogenetic analysis using parsimony (*and other methods), Version 4. Sunderland, MA: Sinauer Associates. Taberlet P, Gielly L, Pautou G, Bouvet J. 1991. Universal primers for amplification of three non-coding regions of chloroplast DNA. Plant Molecular Biology 17: 1105–1109. Torrecilla P, Catalań P. 2002. Phylogeny of broad-leaved and fine-leaved Festuca lineages (Poaceae) based on nuclear ITS sequences. Systematic Botany 27: 241– 251. Warnke SE, Barker RE, Brilman LA, Young WCIII, Cook RL. 2002. Inheritance of superoxide dismutase (sod-1) in a perennial × annual ryegrass cross and its allelic distribution among cultivars. Theoretical and Applied Genetics 105: 1146– 1150. Warnke SE, Barker RE, Jung G, et al. 2004. Genetic linkage mapping of an annual × perennial ryegrass population. Theoretical and Applied Genetics 109: 294– 304. Wilkins PW. 1991. Breeding perennial ryegrass for agriculture. Euphytica 52: 201–214. Wolfe KH, Li W-H, Sharp PM. 1987. Rates of nucleotide substitution vary greatly among plant mitochondrial, chloroplast, and nuclear DNAs. Proceedings of the National Academy of Science USA 84: 9054–9058. Yamada T. 2001. Introduction of a self-compatible gene of Lolium temulentum L. to perennial ryegrass (Lolium perenne L.) for the purpose of the production of inbred lines of perennial ryegrass. Euphytica 122: 213–217.


CBOL Plant Working Group. 2009. A DNA barcode for land plants. Proceedings of the National Academy of Science USA 106: 12794–12797. Connolly V. 2000. Collection and conservation of forage species from old pasture ecosystems. Irish Journal of Agricultural and Food Research 39: 476. Corriveau JL, Coleman AW. 1988. Rapid screening method to detect potential biparental inheritance of plastid DNA and results for over 200 angiosperm species. American Journal of Botany 75: 1443–1458. Darbyshire SJ. 1993. Realignment of Festuca subgenus schedonorus with the genus Lolium (Poaceae). Novon 3: 239– 243. Diekmann K, Hodkinson TR, Fricke E, Barth S. 2008. An optimized chloroplast DNA extraction protocol for grasses (Poaceae) proves suitable for whole plastid genome sequencing and SNP detection. PLoS ONE 3: e2813. http://dx.doi.org/10.1371/journal.pone.0002813. Diekmann K, Hodkinson TR, Wolfe KH, van den Bekerom R, Dix PJ, Barth S. 2009. Complete chloroplast genome sequence of a major allogamous forage species, perennial ryegrass (Lolium perenne L.). DNA Research 16: 165– 176. Dumolin-Lapegue S, Pemonge MH, Petit RJ. 1997. An enlarged set of consensus primers for the study of organelle DNA in plants. Molecular Ecology 6: 393– 397. Ellegren H. 2004. Microsatellites: simple sequences with complex evolution. Nature Reviews Genetics 5: 435–445. Excoffier L, Laval G, Schneider S. 2005. Arlequin ver. 3.0: an integrated software package for population genetics data analysis. Evolutionary Bioinformatics Online 1: 47– 50. Fjellheim S, Rognli OA. 2005. Molecular diversity of local Norwegian meadow fescue (Festuca pratensis Huds.) populations and Nordic cultivars – consequences for management and utilisation. Theoretical and Applied Genetics 111: 640– 650. Fjellheim S, Rognli OA, Fosnes K, Brochmann C. 2006. Phylogeographical history of the widespread meadow fescue (Festuca pratensis Huds.) inferred from chloroplast DNA sequences. Journal of Biogeography 33: 1470–1478. Flannery ML, Mitchell FJ, Coyne S, et al. 2006. Plastid genome characterisation in Brassica and Brassicaceae using a new set of nine SSRs. Theoretical and Applied Genetics 113: 1221– 1231. Floyd DJ, Barker RE. 2002. Change of ryegrass seedling root fluorescence expression during three generations of seed increase. Crop Science 42: 905– 911. Grass Phylogeny Working Group. 2001. Phylogeny and subfamilial classification of the grasses (Poaceae). Annals of the Missouri Botanical Garden 88: 373 –457. Grass Phylogeny Working Group II. 2012. New grass phylogeny resolves deep evolutionary relationships and discovers C4 origins. New Phytologist 193: 304– 312. Grau Nersting L, Bode Andersen S, von Bothmer R, Gullord M, Bagger Jørgensen R. 2006. Morphological and molecular diversity of Nordic oat through one hundred years of breeding. Euphytica 150: 327– 337. Guthridge KM, Dupal MP, Ko¨lliker R, Jones ES, Smith KF, Forster JW. 2001. AFLP analysis of genetic diversity within and between populations of perennial ryegrass (Lolium perenne L.). Euphytica 122: 191–201. Hale ML, Borland AM, Gustafsson MH, Wolff K. 2004. Causes of size homoplasy among chloroplast microsatellites in closely related Clusia species. Journal of Molecular Evolution 58: 182– 190. Hodkinson TR, Chase MW, Lledo´ MD, Salamin N, Renvoize SA. 2002. Phylogenetics of Miscanthus, Saccharum and related genera (Saccharinae, Andropogoneae, Poaceae) based on DNA sequences from ITS nuclear ribosomal DNA and plastid trnL intron and trnL-F intergenic spacers. Journal of Plant Research 115: 381–392. Hodkinson TR, Nı´ Chonghaile G, Sungkaew S, Chase MW, Salamin N, Stapleton CMA. 2010. Phylogenetic analyses of plastid and nuclear DNA sequences indicate a rapid late Miocene radiation of the temperate bamboo tribe Arundinarieae (Poaceae, Bambusoideae). Plant Ecology and Diversity 3: 109–120. Huelsenbeck JP, Ronquist F. 2001. MRBAYES: Bayesian inference of phylogenetic trees. Bioinformatics 17: 754 –755. Kelchner SA. 2000. The evolution of non-coding chloroplast DNA and its application in plant systematics. Annals of the Missouri Botanical Garden 87: 482 –498.

Page 13 of 13