Science & Technology

A first linkage map of globe artichoke (Cynara cardunculus var. scolymus L.) based on AFLP, S-SAP, M-AFLP and microsatellites markers

Description
A first linkage map of globe artichoke (Cynara cardunculus var. scolymus L.) based on AFLP, S-SAP, M-AFLP and microsatellites markers
Published
of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  Theor Appl Genet (2006) 112: 1532–1542 DOI 10.1007/s00122-006-0256-8 ORIGINAL PAPER S. Lanteri · A. Acquadro · C. Comino · R. MauroG. Mauromicale · E. Portis  A first linkage map of globe artichoke ( Cynara cardunculus   var. scolymus   L.) based on AFLP, S-SAP, M-AFLP and microsatellite markers Received: 26 January 2006 / Accepted: 1 March 2006 / Published online: 25 March 2006 ©  Springer-Verlag 2006 Abstract We present the W rst genetic maps of globe arti-choke ( Cynara cardunculus  var. scolymus  L.  2n =  2x =34),constructed with a two-way pseudo-testcross strategy. AF 1  mapping population of 94 individuals was generatedbetween a late-maturing, non-spiny type and an early-maturing spiny type. The 30 AFLP, 13 M-AFLP and 9 S-SAP primer combinations chosen identi W ed, respectively,352, 38 and 41 polymorphic markers. Of 32 microsatelliteprimer pairs tested, 12 identi W ed heterozygous loci in oneor other parent, and 7 were fully informative as they seg-regated in both parents. The female parent map com-prised 204 loci, spread over 18 linkage groups andspanned 1330.5cM with a mean marker density of 6.5cM. The equivalent W gures for the male parent mapwere 180 loci, 17 linkage groups, 1239.4 and 6.9cM.About 3% of the AFLP and AFLP-derived markers dis-played segregation distortion with a P  value below 0.01,and were not used for map construction. All the SSR lociwere included in the linkage analysis, although one locusdid show some segregation distortion. The presence of 78markers in common to both maps allowed the alignmentof 16 linkage groups. The maps generated provide a W rmbasis for the mapping of agriculturally relevant traits,which will then open the way for the application of amarker-assisted selection breeding strategy in this species. Introduction Globe artichoke ( Cynara cardunculus  var. scolymus  L.)makes an important contribution to the Mediterraneanagricultural economy, producing over 800kt of cropfrom more than 80kha of cultivated land (FAO data2004: http://www.faostat.fao.org/). Almost 85% of theworld artichoke production srcinates from Europe. Thespecies is also grown in North Africa, the Middle East,South America, the USA and China (FAO data 2004).The edible part of the plant is the head (formally thecapitulum), which is the immature composite in X ores-cence, used as both a fresh and a canned delicacy world-wide. Each plant produces small, medium and largeheads, with the largest formed at the apex of the terminalbuds along the central stem. The smaller heads developon the lateral branches. The srcin of the artichoke datesback to the era of Theophrastus, the Greek (371–287 BCE )who described their cultivation in Southern Italy andSicily. In 77 CE , the Roman naturalist Pliny the Eldermentioned their use for medicinal purposes, but it wasmost probably between 800 and 1,500 CE  that the arti-choke was domesticated and transformed, presumably inmonastery gardens, into the plant which we know today.Artichoke is a non-fat, zero cholesterol food, rich infolate (vitamin B), vitamin C and minerals, and is a prom-ising source of biopharmaceuticals, such as inulin from itsroots (Brown and Rice-Evans 1998), and antioxidantcompounds, such as luteolin and di-ca V  eoylquinic acidsfrom its leaves (Gebhardt 1997). Furthermore, good eat-ing quality oil can be extracted from its seeds (Maccaroneetal. 1999) and the whole plant can be used for the pro-duction of ligno-cellulosic biomass for energy or paperpulp manufacture (Gominho etal. 2001). Italy is the lead-ing producer of globe artichoke (480kt per year, FAOdata 2004), and also houses the most abundant in situdiversity (Bianco 1990). Distinct varietal groups, welladapted to local environments and local tastes, aregenerally identi W ed on the basis of harvest time (early- tolate-maturing types), size and shape of the head, and Communicated by I. ParanS. Lanteri · A. Acquadro · C. Comino · E. Portis ( & )Di.Va.P.R.A. Plant Genetics and Breeding, University of Turin, via L. da Vinci 44, 10095 Grugliasco (Turin), ItalyE-mail: ezio.portis@unito.itFax: +39-011-2368807R. Mauro · G. MauromicaleDipartimento di Scienze Agronomiche, Agrochimiche e delle Produzioni Animali – sez. Scienze Agronomiche, University of Catania, via Valdisavoia 5, 95123 Catania, Italy  1533 presence/absence of spines on the head bracts. Recently, aliving worldwide collection of 89 varietal types was char-acterized by AFLP pro W ling (Lanteri etal. 2004), and twomajor, genetically di V  erentiated groups were identi W ed:group A includes the non-spiny types with elongated orspherical or sub-spherical capitula, and group B the spinyand non-spiny types with medium-small capitula. The W ngerprint data provided a demonstration that the traitsselected by man have played an important role in shapingthe variation and di V  erentiation within cultivated arti-choke, and supported the hypothesis that globe artichokewas domesticated from wild cardoon ( Cynara cardunculus var . sylvestris ).Globe artichoke is predominantly cross-pollinating.Cross-fertilization is largely enforced in nature by prot-andry, so that by applying simple strategies of pollenpreservation and application, it is possible to obtain sel-fed progenies (Mauromicale and Ierna 2000); however,repeated sel W ng does induce a considerable level of inbreeding depression (Pécaut 1983). At present, com-mercial production is based mainly on the perennial cul-tivation of vegetatively propagated clones via crownshoots; although vegetative propagation is costly and isresponsible for pathogen di V  usion (mainly viruses), itguarantees higher yields of marketable artichokes. Seed-propagated cultivars are becoming popular in someparts of the world, particularly in Israel, Spain and USA(Basnizki and Zohary 1987, 1994; Mauromicale etal. 2004), but they often lack uniformity and their perfor-mance is rather unpredictable. At present artichokebreeding is limited to a small number of studies aimed atunderstanding the inheritance of some major traits(Pécaut 1993; Lopez Anido etal. 1998; Mauromicale etal. 2000). Common breeding aims are to promote ear-liness, yield and quality, and selection is largely based onintra-clonal variation (Deidda 1967; Abbate and Noto 1981; Pécaut 1993; Mauromicale etal. 2000; Gil and Villa 2003). Few attempts have been made to use hybrid-ization between varietal groups to generate novel geneticcombinations (Basnizki and Zohary 1987, 1994; Miller 1975; Scarascia Mugnozza and Pacucci 1976; Tesi 1976). In order to move to a crossing strategy for breeding,some knowledge of artichoke genetics would be advanta-geous, in particular a framework of linkage relationships,which will facilitate the identi W cation and localization of genes controlling important traits, subsequently openingthe way for marker-assisted selection.The aim of the present work was to develop the W rstmarker-based genetic maps of globe artichoke by apply-ing a combination of marker technologies. The strategyadopted was the double pseudo-testcross, pioneered in Eucalyptus  by Grattapaglia and Sedero V   (1994), and subsequently applied to a number of out-breeding spe-cies such as Poa pratensis  (Porceddu etal. 2002), Alstroemeria aurea  (Han etal. 2002), Salix  spp .  (Hanleyetal. 2002; Barcaccia etal. 2003), Olea europea  (Wu etal.2004), Larix decidua  (Arcade etal. 2000), Vitis  spp.(Doucle V   etal. 2004), Carya illinoinensis  (Beedanagarietal. 2005) and Malus  spp. (Kenis and Keulemans 2005).This approach produces two independent maps, one foreach parent (Weeden 1994; Atienza etal. 2002; Yin etal. 2002; La Rosa etal. 2003), and is particularly suited to cross-pollinating species, where individuals typically dis-play a high level of heterozygosity. Materials and methods Plant material and DNA isolationA controlled intraspeci W c cross was performed in theexperimental W elds at the University of Catania in Cassi-bile (Siracusa, Sicily), using as female a single clone of ‘Romanesco C3’, and as male a single clone of ‘Spinosodi Palermo’. The former is a non-spiny varietal type,while the latter carries long sharp spines on its bracts andleaves. Seeds obtained from the cross were germinated inlightly moistened potting mix at room temperature.Emergence was observed within about 10days, andhealthy seedlings were transferred to the W eld after30days, at which stage there were typically three trueleaves. The presence and absence of spines was scored onwell-developed leaves of each F 1  plant. Two weeks aftertransplanting in the W eld DNA was extracted from eachplant following the procedures described by Lanteri etal.(2001), and DNA concentration was estimated by ethi-dium bromide- X uorometry against DNA standards.After checking each presumptive F 1  plant for hybridityusing informative SSR markers (data not shown), 94progenies were selected for segregation analysis andgenetic map construction. However, because a largenumber of AFLP W ngerprints for one individual werenot readable, the genetic maps were W nally based on apopulation of 93 individuals.Marker analysisFor AFLP W ngerprinting, we adapted the protocol of Vos etal. (1995), as detailed by Lanteri etal. (2003). Brie X y, 5   l of extracted DNA (400–500ng) were co-digested with Eco RI (or Pst I) and Mse I, and ligated tostandard adapters. The ligation reaction was used as atemplate for pre-ampli W cation using primers comple-mentary to the adapter sequences plus one selectivenucleotide, namely Eco RI+A (or Pst I+A) and Mse I+C.Selective ampli W cation was subsequently carried outusing primers carrying two or three selective nucleotides.Ampli W ed fragments were electrophoretically resolvedon 5% denaturing polyacrylamide gels and silver stainedas described by Bassam etal. (1991). Example of AFLPpro W les are shown in Fig.1. S-SAP W ngerprinting(Waugh etal. 1997) used a procedure based on theAFLP protocol above. For the selective ampli W cation,one AFLP primer was replaced with the X uorescence-labelled (IRD-700) Cyre5  primer designed to anneal to aretroelement LTR (A. Acquadro, E. Portis, A. Moglia,F.Magurno, S. Lanteri, submitted), and the other was an  1534 unlabelled AFLP primer ( Eco RI, Mse I or Pst I) withthree selective nucleotides. PCR products were separatedon a DNA analyser Gene ReadIR 4200 (LI-COR) in6.5% polyacrylamide gels (Sigma), as described byJackson and Matthews (2000). The M-AFLP W ngerprint-ing method followed the procedure described byAlbertini etal. (2003), using the AFLP pre-ampli W cationproduct as a template. Brie X y, selective ampli W cationswere carried out using a standard two or three selectivebase AFLP primer ( Eco RI, Mse I or Pst I) in combinationwith an 5  -anchored microsatellite primer [PolyGA:GTC(GA) 8  or PolyGT: GAC(GT) 8 ]. PCR products wereseparated as described for S-SAP W ngerprinting. SSRpro W ling used primer pairs developed in our laboratory(Acquadro etal. 2003, 2005a, b). PCR ampli W cationregimes were as detailed by Acquadro etal. (2003) andamplicons were separated and stained as for AFLP,except that the polyacrylamide content of the gels wasincreased from 5 to 6%.Mapping and linkage analysisElectrophoretic patterns were documented using theGel Documentation System (Quantity One Programme,BioRad), analysed twice and only reliable markers con-sidered. Markers were separated into three types: (a)maternal testcross markers, segregating only within‘Romanesco C3’: i.e., female parent A 1 A 2 , male parent g. A patterns o 1  plants and parents, ‘Spinoso di Palermo’ ( A ) and ‘Romanesco C3’ ( B  ), ampli W ed with the E33/M48 ( a ) and E33/M50( b ) primer combinations. Segregating AFLP markers are indicated by arrows A BF 1 A BF 1 ab  1535 A 1 A 1  (expected monogenic segregation ratio of 1:1), (b)paternal testcross markers, segregating only within‘Spinoso di Palermo’: i.e., male A 1 A 2 , female A 1 A 1 , (c)intercross markers, segregating within both parents: i.e.,either both parents A 1 A 2  (expected segregation ratio =3:1 for dominant markers, 1:2:1 for co-dominant), orone parent A 1 A 2 , and the other A 1 A 3  (or A 3 A 4 ),(1:1:1:1). The goodness-of- W t between observed andexpected segregation data was assessed using the chi-square (  2 ) test. Markers segregating in a Mendelianfashion (  2 ·  2  =0.1 ) or deviating only slightly from it(  2  =0.1 <  2 ·  2  =0.01 ) were used for map construction,while those showing highly signi W cant segregation dis-tortion (  2 >  2  =0.01 ) were excluded. Markers withmissing data for more than 30 of the 93 F 1  individualswere excluded. Two separate data sets were thereforeassembled: one was used to construct a linkage map for‘Romanesco C3’ (markers a and c) and the other for‘Spinoso di Palermo’ (markers b and c). The data wereanalysed using JoinMap 2.0 software (Stam and VanOoijen 1995). For both maps, linkage groups wereaccepted at a LOD threshold of 4.0. To determinemarker order within a linkage group, the followingJoinMap parameter settings were used: Rec=0.40,LOD=1.0, Jump=5. Map distances were convertedto centiMorgans using the Kosambi mapping function(Kosambi 1944). Where a discrepancy arose in theorder of markers common to both a maternal andpaternal linkage group, the marker order of the ‘1:1’segregating markers was used as a ‘ W xed order’ toreconstruct the separate parental linkage groups. Theorder of common markers was then inferred by mini-mizing the number of singletons between the ‘3:1’ and‘1:1’ segregating markers in the maternal and paternaldata sets. A singleton is assumed to be suspect as a datapoint because it implies a double recombination event(Han etal. 2002; Isidore etal. 2003). Linkage maps were drawn using MapChart 2.1 soft-ware (Voorrips 2002). AFLP, S-SAP and M-AFLP lociwere named according to primer combination (PC) code(Table1) with multiple markers generated by a given PCordered by decreasing molecular weight. SSR loci were Table1 AFLP, S-SAP andM-AFLP primer combinationsused for linkage analysis Eco RI /Mse I template  Pst I/ Mse I templatePCCodePCCodeAFLPE+ACA/M+CAAe35/m47P+AC/M+CAAp12/m47E+ACA/M+CACe35/m48P+AC/M+CATp12/m50E+ACA/M+CAGe35/m49P+AC/M+CTTp12/m62E+ACA/M+CATe35/m50P+AG/M+CAAp13/m47E+ACA/M+CTTe35/m62P+AG/M+CATp13/m50E+ACC/M+CAAe36/m47P+AG/M+CTAp13/m59E+ACC/M+CACe36/m48P+AG/M+CTCp13/m60E+ACC/M+CTAe36/m59P+AG/M+CTGp13/m61E+ACG/M+CAAe37/m47P+AG/M+CTTp13/m62E+ACG/M+CACe37/m48P+ATG/M+CAAp45/m47E+ACG/M+CAGe37/m49P+ATG/M+CATp45/m50E+ACG/M+CATe37/m50P+ATG/M+CTAp45/m59E+ACG/M+CTGe37/m61P+ATG/M+CTCp45/m60E+ACT/M+CAAe38/m47P+ATG/M+CTGp45/m61E+ACT/M+CATe38/m50P+ATG/M+CTTp45/m62S-SAPCyre5/M+CAAcyre5/m47Cyre5/P+AGCcyre5/p40Cyre5/M+CACcyre5/m48Cyre5/P+AGTcyre5/p42Cyre5/M+CAGcyre5/m49Cyre5/P+AGGcyre5/p41Cyre5/M+CATcyre5/m50Cyre5/E+AAGcyre5/e33Cyre5/E+ACAcyre5/e35M-AFLPPolyGA/E+AAGpGA/e33PolyGA/M+CATpGA/m50(P)PolyGA/E+ACApGA/e35PolyGA/P+ATGpGA/p45PolyGA/M+CATpGA/m50(E)PolyGT/P+ATGpGT/p45PolyGA/M+CCpGA/m16PolyGA/M+CGpGA/m17PolyGA/M+CTCpGA/m60PolyGT/E+ACTpGT/e38PolyGT/M+CAApGT/m47PolyGT/M+CCpGT/m16PolyGT/M+CGpGT/m17  1536 named using the srcinal primer nomenclature. Markersthat segregated with only a minor deviation from theexpected ratio are identi W ed with one (  2  =0.1 <  2 ·  2  =0.05 )or two (  2  =0.05 <  2 ·  2  =0.01 ) asterisk (Fig.2). Indepen-dent linkage maps were constructed for each parentusing the double pseudo-testcross mapping strategy(Weeden 1994).Estimation of genome lengthA method-of-moments type estimator (Hulbert etal.1988), as proposed in ‘method 3’ by Chakravarti etal.(1991), was used to estimate the genome length (G) of each parent. In this method, G is given by the expression N  ( N  -1) X  / K  , where N   is total number of mapped markersin the major groups, X   is the observed maximum dis-tance between two adjacent framework markers in centi-Morgans at a certain minimum LOD score, and K   thenumber of markers pairs with a LOD value at the sameminimum LOD score. Results AFLP, S-SAP and M-AFLP markersSixty-four AFLP PCs (four Eco RI primers £  eight Mse I primers and four Pst I primers £  eight Mse Iprimers) were used to generate W ngerprints of both par-ents and six F 1  progenies. On the basis of the number of polymorphic markers detected, the 30 most informativePCs (15 Eco RI/ Mse I and 15 Pst I/ Mse I, Table1) weretaken forward for mapping. In all, 352 polymorphicAFLP markers were identi W ed, of which 66% wereheterozygous in one parent and absent in the other(testcross markers), with the remainder being heterozy-gous in both parents (intercross markers). The numberof polymorphic AFLP markers per PC ranged from 8to 25, with a mean of 11.7 markers per PC (Table2).For S-SAP, the same group of 30 PCs were applied tothe parents and sample progenies, and nine of thesewere used for mapping (Table1). The number of polymorphic S-SAP markers per PC ranged from 3 to10 (mean 4.6 per PC), and 41 S-SAP markers were iden-ti W ed, of which 25 were testcross and 16 were intercrossmarkers (Table2). Following a screen of 36 M-AFLPPCs, 13 were used for mapping (Table1). The numberof polymorphic M-AFLP markers ranged from 1 to 6per PC, and generated a total of 38 (32 testcross, 6intercross) informative markers (mean 2.9 per PC). Of the 295 testcross markers, 160 (54%) were heterozygousin ‘Romanesco C3’ and the remaining 135 (46%) in‘Spinoso di Palermo’. An analysis of genotype frequen-cies showed that about 14% of these AFLP and AFLP-derived markers su V  ered from segregation distortion(  2 >  2  =0.1 ). Alleles in the male parent showed moresegregation distortion than did those in the femaleparent (16 vs. 10%). Twelve highly distorted markers(  2 >  2  =0.01 ) were discarded prior to the constructionof linkage maps.Microsatellite markersTwelve of the 32 SSR primer pairs assayed were infor-mative: these were CDAT-01, CLIB-04, CLIB-12(Acquadro etal. 2003), CMAL-06, CMAL-8, CMAL-21,CMAL-24, CMAL-108, CMAL-117 (Acquadro etal.2005a), CMAFLP-04, CMAFLP-07 and CMAFLP 08(Acquadro etal. 2005b). CLIB-04, CMAL-06 andCMAL-117 segregated only in the female parent, whileCMAL-08 and CMAL108 segregated only in the maleparent. The remaining loci segregated in both parentseither in the ratio 1:1:1:1 (CDAT-01 and CLIB-12) or1:2:1, and were thus located on both the male and femalelinkage maps. All the SSR loci were included in thelinkage analyses, as minor segregation distortion(  2  =0.05 <  2 ·  2  =0.01 ) was observed only for one locus(CMAL-08).Map constructionAfter the exclusion of 12 markers showing highly signi W -cant levels of distortion, 432 markers were available formap construction (including presence/absence of spines).In all, the maps were built from 300 loci for ‘RomanescoC3’ (maternal) and from 275 for ‘Spinoso di Palermo’(paternal). Of these, 143 loci were in common betweenboth maps.For the maternal map, 204 markers were assignableto 18 major linkage groups (LGs), each containing aminimum of four markers (Fig.1), and a further 19markers were distributed as W ve triplets and twodoublets. Fifteen testcross (9%) and 62 intercross (42%)markers remained unlinked. The length of the individualLGs varied from 27.0 to 132.4cM (mean 73.9cM), com-prising 4–26 loci per LG (mean 11.3). The mean inter-marker distance was 6.5cM, and the longest gap of 26.6cM was found on LG4. Although the majority(76%) of map intervals were less than 10cM, some largegaps remain in the map. The markers generated by Eco RI/ Mse I and Pst I/ Mse I were evenly distributedacross all 18 LGs, without any obvious clustering of markers generated by any one PC. Only four LGs (LG6,13, 14 and 15) were composed solely of AFLP loci.Microsatellite loci (SSR and M-AFLP) and S-SAP frag-ments were distributed over, respectively, nine and eightLGs. Seventeen loci with minor segregation distortionwere mapped to 12 LGs (14 at  =0.05 and three at  =0.01), and groups of distorted loci in close linkage toone another were detected on LG1, LG10 (three loci perLG) and LG5 (two loci).For the paternal map, 180 loci were arranged into 17LGs, with three triplets and three doublets (Fig.1).Seventeen testcross (13%) and 62 intercross (42%) mark-ers remained unlinked. The length of the LGs variedfrom 26.5 to 126.7cM (mean 72.9cM). The number of loci per LG varied from 4 to 21 (mean 10.6), giving a
Search
Similar documents
View more...
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks