Successful Gene Tagging in Lettuce Using the Tnt1 … · ‘Jessy,’ were cocultivated with an...

Breakthrough Technologies

Successful Gene Tagging in Lettuce Using the Tnt1Retrotransposon from Tobacco

Marianne Mazier*, Emmanuel Botton, Fabrice Flamain, Jean-Paul Bouchet, Beatrice Courtial1,Marie-Christine Chupeau, Yves Chupeau, Brigitte Maisonneuve, and Helene Lucas

Unite de Genetique et d’Amelioration des Fruits et Legumes, UR1502, Institut National de la RechercheAgronomique (INRA), F–84143 Montfavet cedex, France (M.M., E.B., F.F., J.-P.B., B.M.); and Laboratoirede Biologie Cellulaire, UR501, INRA, F–78026 Versailles cedex, France (B.C., M.-C.C., Y.C., H.L.)

The tobacco (Nicotiana tabacum) element Tnt1 is one of the few identified active retrotransposons in plants. These elements possessunique properties that make them ideal genetic tools for gene tagging. Here, we demonstrate the feasibility of gene tagging usingthe retrotransposon Tnt1 in lettuce (Lactuca sativa), which is the largest genome tested for retrotransposon mutagenesis so far. Of10 different transgenic bushes carrying a complete Tnt1 containing T-DNA, eight contained multiple transposed copies of Tnt1.The number of transposed copies of the element per plant was particularly high, the smallest number being 28. Tnt1 transpositionin lettuce can be induced by a very simple in vitro culture protocol. Tnt1 insertions were stable in the progeny of the primarytransformants and could be segregated genetically. Characterization of the sequences flanking some insertion sites revealed thatTnt1 often inserted into genes. The progeny of some primary transformants showed phenotypic alterations due to recessivemutations. One of these mutations was due to Tnt1 insertion in the gibberellin 3b-hydroxylase gene. Taken together, these resultsindicate that Tnt1 is a powerful tool for insertion mutagenesis especially in plants with a large genome.

Continuous advances in whole-genome sequencinghave resulted in important breakthroughs in plantscience. Current data obtained on a few model species(Arabidopsis [Arabidopsis thaliana] and rice [Oryzasativa]) demonstrate that the degree of collinearity israther low at the level of the whole genome andongoing genome projects in other plant species mightnot be able to rely on the information obtained fromalready sequenced genomes to the extent that had beenexpected (Aert et al., 2004). Consequently, the impor-tance of individual genome projects in higher organ-isms should not be underestimated. Unfortunately,traditional sequencing approaches are time consumingand expensive, and the time and cost increase propor-tionally with the size of the genome, which is a majorobstacle in the accumulation of sequences and in thediscovery of genes and nongenic functional elements inmany agriculturally and industrially important plantgenomes. Methods that allow rapid identification of thesequence involved in an interesting character (or aparticular alteration of a phenotype) in a plant arepromising, especially in plants for which genomicand whole-genome sequencing are not available. In

this context, reverse-genetics tools such as insertionalmutagenesis play an important role. Transposable ele-ments provide attractive tools for constructing mu-tant collections and for gene tagging. Among thesetransposable elements, retrotransposons have uniquefeatures that give them several advantages over tradi-tional insertion elements (Hirochika, 1997; Kumar andBennetzen, 1999; Kumar and Hirochika, 2001). Becausethey transpose via a replicative mode, the mutationsthey induce are stable. As they transpose via an RNAintermediate that is reverse transcribed into extrachro-mosomal DNA, transposition target sites are not linkedwith the site of the original copy. Their transposition isregulated and most of the retrotransposons used formutagenesis until now are activated by tissue culture,enabling large-scale generation of mutated popula-tions. Retrotransposons have already been used suc-cessfully as mutagens in plants: Tos17, an endogenousretrotransposon of rice, is used for gene tagging in rice(Yamazaki et al., 2001). Moreover, it was shown thatTto1 and Tnt1 transpose in Arabidopsis (Okamoto andHirochika, 2000; Courtial et al., 2001), and Tnt1 fromtobacco (Nicotiana tabacum) is used in Medicago trunca-tula (d’Erfurth et al., 2003; Tadege et al., 2005). Thesestudies revealed another important advantage of theuse of retrotransposons over more traditional DNA-type elements: They can be highly mutagenic becausesome of them prefer to transpose into gene-rich regions.This last feature could make them ideal genetic tools forgene tagging in plants with a large genome.

Our purpose was to explore the opportunities ofdeveloping a retrotransposon mutagenesis strategy inlettuce (Lactuca sativa) using the heterologous elementTnt1. Lettuce is an important crop species belonging to

1 Present address: Laboratoire National de la Protection desVegetaux, Unite de Detection des OGM, 93 Rue de Curembourg,45404 Fleury-les-Aubrais cedex, France.

* Corresponding author; e-mail [email protected]; fax 33–4–32–72–27–02.

The author responsible for distribution of materials integral to thefindings presented in this article in accordance with the policydescribed in the Instructions for Authors (www.plantphysiol.org) is:Marianne Mazier ([email protected]).

www.plantphysiol.org/cgi/doi/10.1104/pp.106.090365

18 Plant Physiology, May 2007, Vol. 144, pp. 18–31, www.plantphysiol.org � 2007 American Society of Plant Biologists

Dow

nloaded from https://academ

ic.oup.com/plphys/article/144/1/18/6106720 by guest on 22 August 2021

the large Compositae family. The Compositae GenomeInitiative (http://cgpdb.ucdavis.edu) has generated anenormous amount of DNA sequence data, especiallyESTs, from the two major representatives of this family:lettuce and sunflower (Helianthus annuus). However,lettuce has a genome of 2.65 pg/1C, corresponding to2,600 Mb per haploid genome, approximately 18, 6, and5 times larger than Arabidopsis, rice, and Medicago,respectively (Arumuganathan and Earle, 1991). It is thelargest genome tested so far for retrotransposon muta-genesis. Lettuce is amenable to most in vitro culturetechniques and is easy to transform via Agrobacteriumtumefaciens (Michelmore et al., 1987; Chupeau et al.,1994; Mazier et al., 2004; Lelivelt et al., 2005). Interest-ingly, Yang et al. (1993a, 1993b) already tested transposonmutagenesis using the well-known DNA transposableactivator/dissociation (Ac/Ds) elements from maize (Zeamays) in lettuce. Despite demonstrating that the elementswere functional in this plant, these authors showed thatthe frequency of germinal transposition provided wastoo low for practical use.

In this article, we show that Tnt1 transposes effi-ciently in lettuce during in vitro culture and that theinsertions are stable and genetically unlinked. We alsodemonstrate the feasibility of gene tagging in lettuceby identifying and isolating an insertion of Tnt1 intoa GA b-hydroxylase gene that gives rise to a dwarfphenotype.

RESULTS

Transposition of Tnt1 during in Vitro Transformationof Lettuce Occurred Simultaneously with

T-DNA Integration

Tnt1 is an active 5.3-kb-long copia-like long-terminalrepeat (LTR) retroelement isolated from tobacco(Grandbastien et al., 1989). To investigate whetherTnt1 transposes in lettuce, leaf fragments from twodifferent varieties of butterhead lettuce, ‘Mariska’ and‘Jessy,’ were cocultivated with an Agrobacterium straincarrying the tnk23 plasmid (Fig. 1A) containing in itsT-DNA an autonomous copy of Tnt1 (Lucas et al., 1995).After 2 d of cocultivation, leaf explants were transferredto an in vitro selective regeneration medium until anumber of transformed bushes appeared, as describedby Dinant et al. (1997). Fourteen independent bushesfrom 14 different leaf fragments (nine from ‘Mariska’and five from ‘Jessy’) were transplanted onto selectiverooting medium. Each bush is generally composed ofseveral buds that subsequently have to be plantedindividually to obtain healthy plants in the greenhouse.In previous lettuce transformation experiments, it wasverified that all the buds from one bush were clonesresulting from the same transformation event (Dinantet al., 1997; Dubois et al., 2005).

Figure 1B shows the results of the Southern hybridi-zation on T0 genotypes (primary transformants) derived

Figure 1. Southern-blot analysis ofprimary transformant genotypes (T0)obtained with the tnk23 binary vec-tor. Ten micrograms of total DNAfrom different T0 transformants weredigested with HincII and analyzedby Southern blot using a Kana orR-U5 probe. A, Structure of thetnk23 T-DNA introduced in lettuceshowing HincII restriction enzymesites and positions of probes. TheKana and R-U5 probes are repre-sented under the T-DNA scheme.RB, Right border; LB, left border. B,Hybridization with a R-U5 probe.The expected position (sizes 0.63and 1.5 kb) of the two hybridizingfragments from the T-DNA is in-dicated. C, Hybridization with aKana probe. Blot 1, 1 5 MarTnt53b;2 5 MarTnt53a; 3 5 MarTnt28c;4 5 MarTnt28b; 5 5 MarTnt28a;6 5 MarTnt27a; 7 5 MarTnt21a;8 5 MarTnt14b; 9 5 MarTnt13b;10 5 MarTnt13a; 11 5 MarTnt10a;12 5 MarTnt8a; 13 5 MarTnt4b;14 5 MarTnt4a; 15 5 ‘Mariska’wild type. Blot 2, 1 5 JesTnt8a; 2 5

JesTnt7c; 3 5 JesTnt7a; 4 5 JesTnt6b;5 5 JesTnt6a; 6 5 JesTnt2a; 7 5

JesTnt1a; 8 5 ‘Jessy’ wild type.

Gene Tagging in Lettuce with Tnt1 Retrotransposon

Plant Physiol. Vol. 144, 2007 19

Dow



from six bushes. Buds from the same bush are desig-nated by the same number and differ by the letter placedjust after. Each blot was successively hybridized withtwo probes, namely, R-U5 (fragment of Tnt1 localized inthe LTR) and Kana (fragment of the nptII gene). WhenT-DNA inserts were complete, hybridization with theR-U5 probe on HincII-digested genomic DNA revealedtwo internal fragments of 630 and 1,500 bp (Fig. 1A). Ofthe 21 buds analyzed, three (MarTnt21a and two budsfrom the same bush, JesTnt7) did not reveal the twointernal fragments and thus probably contained a trun-cated version of the T-DNA without Tnt1. Two buds,MarTnt27a and JesTnt2a, revealed only the 630-bp frag-ment, indicating a longer truncated version of theT-DNA containing only the 3#-LTR from Tnt1. Thus,although the Kana probe revealed that all the buds an-alyzed contained the right part of the T-DNA, fourbushes out of the 14 analyzed (30%) contained only atruncated version of the T-DNA in which at least the5#-part of Tnt1 was missing. Two buds (MarTnt14b andMarTnt10a) of the 21 analyzed displayed only the twoexpected fragments when the blots were hybridizedwith the R-U5 probe. The 14 remaining buds (corre-sponding to eight different bushes of the 14 analyzed)revealed the presence of multiple bands in addition tothe 630- and 1,500-bp expected fragments. The highdegree of intensity observed specifically at 1,500 bpcorresponded to the internal fragment localized insideTnt1, whereas no such intensity was visible for the630-bp fragment that corresponded to an internal frag-ment specific to the T-DNA. The other fragments prob-ably corresponded to transposed copies of Tnt1, althoughsome of them could be due to methylation of some Tnt1HincII sites (J. Perez, F. Potet, B. Courtial, and H. Lucas,unpublished data). Multiple transposed copies of Tnt1were thus inserted in the genome of 14 buds originatingfrom the two varieties of lettuce. Hybridization with theKana probe confirmed the differences between the num-ber of T-DNA inserts and the number of Tnt1 copies.Whereas the number of T-DNA inserts varied betweenone and three, the number of Tnt1 insertions was veryhigh and difficult to estimate accurately on the blots.

Hybridization with the Kana probe confirmed pre-vious observations that buds from the same bushpossess the same T-DNA inserts and derive from thesame transformation event (Fig. 1C). The multiple-band patterns obtained with the R-U5 probe appearedto be identical for buds from the same bush, suggestingthat transposition of Tnt1 occurred very early duringthe in vitro culture and transformation process, con-comitant with the T-DNA transfer process.

Tnt1 Insertions Are Stable, Genetically Independent,

and Can Be Easily Separated by Crossing toWild-Type Lettuces

When using Tnt1 for insertion mutagenesis in let-tuce, it is important to be able to isolate plants carryingonly one or a few insertions of Tnt1 to identify theinsertions responsible for the useful mutant pheno-

types. Transposed copies of the element should also begenetically unlinked to cover the whole genome. Inaddition, it is important that new rounds of transpo-sition do not happen in subsequent generations. Toinvestigate the genetic independence of the Tnt1 in-sertions, their segregation was studied in progenies ofthe MarTnt53a transformant. T1 plants obtained by self-pollination of the MarTnt53a primary transformantwere backcrossed to the wild-type genotype ‘Mariska.’Seventy-seven different BC1 genotypes, originatingfrom 51 different T1 plants, were analyzed by Southernblot. Twenty-eight BC1 plants each containing betweennine and 16 different Tnt1 insertions, but togetherincluding all the insertions present in the MarTnt53aprimary transformant genotype, were selected andagain backcrossed to ‘Mariska.’ Southern-blot analysiswas performed on 104 different BC2 plants. Twenty-one of 104 BC2 plants analyzed are presented in Figure2B. At least 28 different Tnt1 insertions were clearlyidentified. Provided that, among the different T0 geno-types showing transposed insertions, the MarTnt53agenotype was the one containing the lowest number ofTnt1 copies (Fig. 1), it can be assumed that the numberof Tnt1 insertions exceeded 28 for the other genotypesobtained in the transformation experiment. For all 28insertions, it was possible to detect segregation eventsby Southern-blot analysis. Four BC2 genotypes, eachcontaining two to five Tnt1 insertions and togetherrepresenting 11 different Tnt1 transposition events,were self-pollinated and 39 to 44 progenies per BC2line were analyzed by Southern blot. The 12 differentprogenies presented in Figure 3 show these 11 Tnt1insertions in different combinations of two or isolated(lanes 2, 3, 4, and 7). We looked for new fragmentsindicative of germinal transposition events of Tnt1 inthe different progenies analyzed, but did not find any,indicating that Tnt1 does not transpose at a highfrequency in the germ cells of lettuce. Taken together,the results showed no evidence of linkage between anyof the Tnt1 inserts studied, but rather showed Men-delian segregation of the insertions.

Tnt1 Is Transcriptionally Active in the Early

Stage of Regeneration

To analyze Tnt1 transcriptional regulation in lettuce,we introduced the pHLV5501 construct, a translationalfusion between Tnt1 LTR and the GUS reporter gene(Courtial et al., 2001), into the ‘Mariska’ genome bytransgenesis. T0 plants, corresponding to independenttransformation events, were transferred in the green-house to obtain T1 seeds by self-pollination. One hun-dred T1 seeds from each T0 plant were sown on rootingmedium containing 75 mg/L kanamycin. Three geno-types were selected that had a 3:1 ratio of kanamycinresistance consistent with one functional nptII locus.Homozygous T2 lines were selected on kanamycin.Histochemical staining of 2-week-old T2 seedlingsrevealed moderate blue staining localized only inroots, with more intense staining in the youngest parts

Mazier et al.

20 Plant Physiol. Vol. 144, 2007

Dow



(Fig. 4A). In vitro regeneration experiments wereperformed with leaf explants taken from 2-week-oldsterile T2 seedlings of each of the selected genotypes.The expression of the LTR-GUS fusion was followedby histochemical staining of explant samples at 2 d,6 d, 9 d, 13 d, 16 d, 20 d, 2 months, and 3 months afterthe beginning of the regeneration experiment. Figure 4shows the results obtained with one of the selectedgenotypes (Mariska-LTR17a-5). The expression patternof the LTR-GUS construct over time was the same forthe three genotypes studied. The strongest GUS ex-pression was observed 2 d after the beginning of theregeneration experiment (Fig. 4A, 1). All the explantsfrom the transgenic T2-selected transformants showeduniform dark-blue staining at this time point. Four and7 d later, dark-blue staining was observed, but only atthe periphery of the explants, whereas the center of theexplants became uncolored (Fig. 4A, 2 and 3). GUSexpression was never detected in untransformed con-trol explants during the experiment (Fig. 4B, 1–8). Six-teen to 20 d after the beginning of the regenerationexperiment, calli started to develop from the explantsmainly on the perimeter. Tnt1-GUS expression wasmainly localized in certain parts of the calli and de-creased over time (Fig. 4A, 4–6). Two months after thebeginning of the regeneration experiments, newly

regenerated shoots and leaves were clearly visible onthe explants. Whereas a low level of Tnt1-GUS expres-sion was still visible in some parts of the explants, nostaining was observed in regenerated plantlets. Threemonths after the beginning of regeneration, GUS ex-pression was reduced and hardly visible. These resultsshowed that Tnt1 was transcriptionally activated earlyin the in vitro regeneration process (2 d after thebeginning of the experiments).

Tnt1 Transposition Can Be Reactivated in Lettuce

Activation of Tnt1 transposition by in vitro regen-eration of lettuce genotypes containing one Tnt1 ele-ment could be an easy way to increase the size of atagged population. To test the ability of several Tnt1copies present in the lettuce genome to be activated fortransposition, we used the 12 genotypes possessingone or two Tnt1 insertions issued from the MarTnt53agenotype by several backcrosses with the wild-type‘Mariska,’ as presented previously (Fig. 3). Leaf ex-plants taken from the 12 genotypes were submitted toin vitro regeneration with or without preliminary Agro-bacterium coculture. As verified by Southern blot with akanamycin probe (data not shown), all 11 Tnt1 inserts

Figure 2. Southern-blot analysis of the progeny of one transformant (MarTnt53a) after two successive backcrosses with the wild-type parent ‘Mariska’ showing the genetic independence of the different insertions. Ten micrograms of total DNA from differentBC2 plants were digested with NcoI and analyzed by Southern blot using a RT probe. A, Structure of the tntk23 T-DNA introducedin lettuce showing NcoI restriction enzyme sites and probe position. The RT probe is represented under the T-DNA scheme. RB,Right border, LB, left border. B, Hybridization with a RT probe. L 5 1-kb ladder; 1 5 MarTnt53a-11-2-3; 2 5 MarTnt53a-22-2-2;3 5 MarTnt53a-22-2-4; 4 5 MarTnt53a-25-1-1; 5 5 MarTnt53a-38-2-5; 6 5 MarTnt53a-38-2-7; 7 5 MarTnt53a-39-2-5;8 5 MarTnt53a-40-2-4; 9 5 MarTnt53a-40-2-5; 10 5 MarTnt53a-40-2-7; 11 5 MarTnt53a-44-2-4; 12 5 MarTnt53a-45-2-1;13 5 MarTnt53a-46-2-1; 14 5 MarTnt53a-46-2-6; 15 5 MarTnt53a-51-2-3; 16 5 MarTnt53a-51-2-4; 17 5 MarTnt53a-52-2-4;18 5 MarTnt53a-55-1-1; 19 5 MarTnt53a-57-2-2; 20 5 MarTnt53a-57-2-6; 21 5 MarTnt53a-57-2-8.



Dow



covered by the 13 studied genotypes were previouslytransposed copies without T-DNA.

One to 15 independently regenerated shoots fromeach genotype were used for DNA extraction andSouthern analysis. Figure 5 shows a Southern blot forthe 14 regenerated shoots originating from Mariska-Tnt53a-51-2-4-6, obtained after cocultivation with anempty Agrobacterium followed by regeneration. Eachnew fragment indicates a transposition event of Tnt1,and seven of the 14 segregated shoots contained newtransposition events. As summarized in Table I, allgenotypes, except one (MarTnt53a-51-2-4-24), gave atleast one regenerated plantlet with new Tnt1 trans-posed copies. It was not possible to find an obviouscorrelation between the frequency of transposition inthe regenerated shoots and Tnt1 copy number or aspecific insertion locus in the starting material. WhenTnt1 transposition occurred, it generally gave rise to alarge number of new insertion events (data notshown). Cocultivation of explants with an emptyAgrobacterium strain prior to regeneration did notresult in a significant increase in the frequency ofTnt1 transposition (Table I). Overall, around 40% of theregenerated plantlets contained new transpositionevents of Tnt1.

Transgenic Lettuces Containing Tnt1 Insertions:A Small Highly Mutagenic Population

No phenotypic variation was observed in the T0plants acclimated in the greenhouse from transformed

bushes. However, phenotypic variants were found infour of eight T1 families studied.

Three different genotypes (JesTnt1, JesTnt6, andJesTnt8) showed abnormal phenotype at the seedlingstage in their T1 progenies. They were blocked at thecotyledon stage or the cotyledons showed a pigmenta-tion defect (white for JesTnt1a-7 and yellow forJesTnt6a and JesTnt8a), and the roots were shorter thanthose of the wild type (Fig. 6A). Segregation analysisof the mutant phenotypes observed in JesTnt6a andJesTnt8a T1 and T2 families indicated that they corre-sponded to a single recessive mutation (Table II). Segre-gation analysis of the mutant phenotype in the JesTnt1a-7T2 family suggested that this phenotype could be due torecessive mutations in two unlinked genes (the ratio ofwhite to green seedlings was between those expected forone and two recessive mutations).

Other phenotypic variations were observed in thegreenhouse at later stages of development in the T1progeny of three genotypes. Dwarf plants appeared inJesTnt6 progeny with a very short internodal stem(Fig. 6C). Marked modification in the structure of theflower (absence of ligules and stamens) was observedin JesTnt1 progeny (Fig. 6B) and in the color of theligule (light yellow) in MarTnt53 progeny. Segregationanalysis of these mutant phenotypes in T1, T2, and T3families suggested that all three corresponded to sin-gle recessive mutations (Table II). The high number ofTnt1 insertions in the genotypes presenting phenotypicalterations associated with the recessive character ofthose mutations did not allow identification of the insertsinvolved, except for the dwarf mutation for whichseveral rounds of backcrosses were performed. Conse-quently, we cannot exclude the involvement of someendogenous transposable elements or other mutations inaddition to Tnt1 to generate these phenotypic alterations.

Tnt1 Transposes into Coding Regions in theLettuce Genome

One significant advantage of using retrotransposonsfor gene tagging in plants with large genomes is thepreference some display to transpose into gene-richregions. To test Tnt1 mutagenic activity in lettuce, 25different lettuce Tnt1 insertion sites longer than 100nucleotides were isolated by sequence-specific ampli-fication polymorphism (S-SAP) and then sequenced(Table III). Sequences were processed to obtain lettuceDNA fragments corresponding to the 5#-Tnt1 flankingsequence. The cleaned sequenced DNA fragments var-ied in size from 114 to 557 nucleotides. Sequences werecompared with four of The Information for GenomicResearch (TIGR) databases (lettuce, sunflower, tomato[Solanum lycopersicon], and potato [Solanum tuberosum])using the BLASTN program. We also compared thetranslated sequences of the sequenced DNA fragmentswith these translated databases and with two proteindatabases (Arabidopsis and Swiss-Prot TrEmbl) usingthe BLASTX program to see whether we could detectany homology with known or hypothetical proteins. Of

Figure 3. Southern-blot analysis of the progeny of some selfed BC2 plantsselected for possession of few Tnt1 copies. Ten micrograms of total DNAfrom the plants were digested with NcoI and analyzed by Southern blotusing a RT probe. L 5 1-kb ladder; 1 5 MarTnt53a-44-2-2-5; 2 5

MarTnt53a-44-2-2-11;3 5 MarTnt53a-22-2-4-1;45 MarTnt53a-22-2-4-34; 5 5 MarTnt53a-51-2-4-6; 6 5 MarTnt53a-51-2-4-15; 7 5 MarTn-t53a-51-2-4-22; 8 5 MarTnt53a-51-2-4-24; 9 5 MarTnt53a-51-2-3-7;10 5 MarTnt53a-51-2-3-17; 11 5 MarTnt53a-51-2-3-26; 12 5 MarTn-t53a-51-2-3-31; C 5 ‘Mariska’ wild type.

Mazier et al.


Dow



the 25 sequences analyzed by BLASTN, eight se-quences (32%) showed no homology. Fourteen se-quences (56%) showed some similarities with ESTs orknown mRNAs, with E values between 7.8e-103 (per-fect match, i.e. certain) and 5.7e-05 (probable but notcertain). E values above 1e-05 were regarded as notsignificant and not included in Table III. Apart fromsequence HEL185 (which perfectly matched lettuceLs3h1 mRNA coding for the GA 3b-hydroxylase), theother 13 highest sequence similarities found showedscores between 703 (highly probable) and 222 (probablebut not certain). Among these 13 sequences, eightmatched lettuce ESTs (HEL131, HEL143, HEL207,HEL189, HEL195, HEL191, HEL193, and HEL102), twomatched sunflower ESTs (HEL150 and HEL198), and thelast three matched Solanaceae ESTs (HEL132, HEL210,and HEL128). The remaining three sequences (12%)showed no homology with BLASTN, but some homol-ogy with hypothetical proteins from Arabidopsis(HEL115, HEL204, and HEL205), with E values between1.6e-11 and 1.9e-06 and scores between 127 and 115.Taken together, the results of the analysis of 25 Tnt1flanking sequences are not in favor of random insertionof Tnt1 in the lettuce genome, but rather could indicatea preference of Tnt1 to insert into expressed regions,

even if this still has to be proved statistically byobtaining more sequence data.

Tnt1 Insertion into a GA b-Hydroxylase Gene Inducesa Dwarf Phenotype in Lettuce

To demonstrate that Tnt1 was responsible for themutant phenotype observed, we decided to study thedwarf mutant in more detail. The mutant phenotypewas detectable at an early stage of development, and,despite the absence of a flower stem, the plants werepartially fertile and generated 100% mutant progeniesafter self-pollination. Southern-blot analysis of 60 T3JesTnt6a plants showed that the Tnt1 copy numberwas too high to allow correlation with the presence ofa specific fragment with the dwarf mutation (data notshown). One T3 mutant plant was backcrossed withthe wild type. Five BC1 plants selected to posses thelowest Tnt1 copy number (between 25 and 30) amongthe 100 analyzed were grown in the greenhouse andself-pollinated. Dwarf plants were selected fromamong these BC1 progenies and the Tnt1 copy numberwas assessed by Southern blot for 100 of them. Theplant that contained the lowest copy number of Tnt1(between eight and 10) was backcrossed to the wild

Figure 4. Histochemical GUSstaining of ‘Mariska-LTR-17a-5’ T2

plants obtained with the LTR-GUSconstruct and their leaf explants atdifferent times during the in vitroculture regeneration protocol. A,‘Mariska-LTR-17a-5’ genotype. B,Control ‘Mariska’ genotype. 0,Fifteen-day-old seedlings. 1, 2, 3,4, 5, and 6, Leaf explants, respec-tively, 2, 6, 9, 13, 16, and 20 dafter placing on in vitro regenera-tion medium; 7 and 8, leaf ex-plants at, respectively, 2 and 3months after placing on in vitroregeneration medium.



Dow



type and five BC2 plants out of the 50 analyzed bySouthern blot, containing four to five Tnt1 inserts,were allowed to grow in the greenhouse for self-pollination. Fifteen to 25 progenies per BC2 plant wereallowed to grow in the greenhouse for detection ofthe mutant phenotype and mutant genotypes wereanalyzed by Southern blot. The mutant plant contain-ing the smallest copy number (two) was then self-pollinated, giving 100% mutant progenies with somecontaining only one Tnt1 copy (Fig. 6D, 1).

The S-SAP fragment corresponding to a Tnt1 inser-tion site in the lettuce GA b-hydroxylase (sequence

HEL185; Table III) was detected only in JesTnt6a prog-enies harboring the dwarf mutation (data not shown).Treatment of JesTnt6a progenies harboring the mutantphenotype with GA allowed complete reversion of thephenotype, suggesting that, indeed, disruption of theGA b-hydroxylase gene by Tnt1 results in the dwarfphenotype. PCR amplification of plants previouslycharacterized for their mutant phenotype (and its seg-regation in their progeny), using one oligonucleotidespecific for the Tnt1 5#-end and two oligonucleotidesspecific for the GA b-hydroxylase gene (upstream anddownstream of the Tnt1 insertion site), demonstratedthat, indeed, the mutant dwarf phenotype was due totransposition of Tnt1 into this gene (Fig. 7).

DISCUSSION

In this article, we have shown that the tobacco retro-transposon Tnt1 transposes in lettuce. Molecular analy-sis of several individuals isolated from a commonregenerated ancestor bush demonstrated that the trans-position of Tnt1 occurred very early during the lettucetransformation process, concomitantly to T-DNA trans-fer. Of the 10 different bushes containing a completeversion of the T-DNA analyzed, two bushes containedonly the T-DNA insert, whereas the remaining eightcontained multiple transposed copies of Tnt1. The num-ber of transposed copies per plant was particularly highand difficult to estimate accurately on the T0 Southernblots. Whereas in Arabidopsis the number of transposedTnt1 copies fluctuated between 0 and 26 (Courtial et al.,2001) and in Medicago between four and more than 30(d’Erfurth et al., 2003), the number of Tnt1 insertions

Table I. Induction of Tnt1 transposition in lettuce after regeneration

Donor PlantaNo. of Tnt1

Inserts in Donor Plantb

No. of Plants with New Tnt1 Insertions/No. of Plants Tested (Percentage)

After in Vitro Regenerationc,d After Coculture

and Regenerationc,e Total (Percentage)

MarTnt53a-44–2-2-5 2 1/4 nd 1/4 (25.0)MarTnt53a-44–2-2-11 1 2/6 nd 2/6 (33.3)MarTnt53a-22–2-4-1 1 2/7 1/1 3/8 (37.5)MarTnt53a-22–2-4-34 1 4/12 nd 4/12 (33.3)MarTnt53a-51–2-4-6 2 nd 7/15 7/15 (46.6)MarTnt53a-51–2-4-15 2 3/10 nd 3/10 (30.0)MarTnt53a-51–2-4-22 1 11/14 nd 11/14 (78.5)MarTnt53a-51–2-4-24 2 0/1 7/10 7/11 (63.6)MarTnt53a-51–2-3-7 2 3/9 2/6 5/15 (33.3)MarTnt53a-51–2-3-17 2 8/15 nd 8/15 (53.3)MarTnt53a-51–2-3-26 2 1/10 0/10 1/20 (5.0)MarTnt53a-51–2-3-31 2 4/8 5/7 9/15 (60.0)Total (all genotypes) 39/96 (40.6) 22/49 (44.8) 61/145 (42.0)

aDonor plantswereobtained after backcrossing the MarTnt53a transformant twice to the wild-type genotype ‘Mariska’ and allowing the progenies to self-pollinate once. bThe number of Tnt1 inserts in each plant was determined by Southern blot (Fig. 3). cnd, Not determined. dFor regenerationexperiments, leaves from the donor plant were sterilized and sectioned in small pieces and placed on an in vitro regeneration medium. After 1 month,regeneratedplantlets were isolated fromthe explants (never more than one plant per explant) and transferred onto rooting medium. Plantletswere then usedfor genomic DNA extraction and Southern-blot analysis (Fig. 5). eFor regeneration and coculture experiments, the lettuce standard transformationprotocolwas followedusinganAgrobacterium straincontainingabinaryplasmidwithonly thekanamycin resistancegene.After 2dof liquidcoculturewithA. tumefaciens, explants were placed on an in vitro regeneration medium containing augmentin (1 g L21).

Figure 5. Southern-blot analysis of in vitro-regenerated plantlets andthe donor plant from which the leaf explants were taken showing newtransposed Tnt1 copies. Ten micrograms of total DNA from the plantswere digested with NcoI and analyzed by Southern blot using a RTprobe. Newly regenerated plants were obtained from leaves taken fromthe donor plant MarTnt53a-51-2-4-6 (called 2-4-6). 1, Donor plant 2-4-6 (MarTnt53a-51-2-4-6); 2 to 15, independently regenerated plants.

Mazier et al.


Dow



obtained in lettuce was estimated by molecular analysisperformed on successive backcrosses to be 28 for onegenotype and much higher for the seven remaining T0genotypes, revealing a highly efficient rate of transposi-tion. This high copy number per plant will be an advan-tage for lettuce mutagenesis because it reduces thenumber of transgenic lines that have to be produced tosaturate the genome, although the time needed to seg-regate these elements afterward to show that the mutantphenotype is tagged could be a limiting factor. Based onthe equation of Krysan et al. (1999) and assuming anaverage gene length of 3.5 kb and that Tnt1 insertions arerandom, we estimate that 2,225,400 insertions would berequired to have a 95% probability of tagging any givengene in the 2,600 Mb lettuce genome. With a postulatethat a Tnt1 average copy number of 50 insertions per lineinserted into unlinked loci in lettuce, the same level ofsaturation could be obtained in 44,508 lines.

Contrary to results obtained with Medicago show-ing plants containing transposed Tnt1 copies withoutT-DNA insertions, the four different regenerated lettucebushes containing only truncated versions of the T-DNAdid not show any transposed copies of Tnt1.

The study conducted on two consecutive backcrossesand one self-pollination performed with a given T0transformant demonstrated the genetic independenceof the different Tnt1 copies and the Mendelian segre-

gation of the insertions. Furthermore, several differentsingle-copy insert plants were easily obtained afteronly three backcrosses and one self-pollination, start-ing with a plant containing 28 copies of Tnt1, demon-strating the feasibility of separating a given Tnt1insertion from the others even from a genotype con-taining large numbers of transposed copies. No newtransposition events were observed in any of theprogenies studied, as already observed in Arabidopsisand Medicago (Lucas et al., 1995; d’Erfurth et al., 2003).Thus, the insertions of Tnt1 are stable and the elementdoes not transpose at a detectable frequency in thegerm cells of lettuce under standard growth condi-tions in the greenhouse.

Using LTR-GUS fusion, we demonstrated that tran-scription of the element, driven by its LTR promoter, ishighly regulated. LTR-GUS products are poorly ex-pressed in most intact lettuce leaf tissues of transgenicplants, except roots. However, Tnt1 transcription isstrongly induced in lettuce by in vitro culture regener-ation experiments. Transcription activation occurs veryearly, 2 d after the beginning of the experiments, andthen decreases to no transcription at all in newlyregenerated plantlets. Thus, the absence of germinaltransposition events in the progeny of transgenic plantscarrying Tnt1 and the early occurrence of transposedcopies obtained in transgenic lettuce after in vitro

Figure 6. Phenotypic variations observed on theprogeny of lettuce genotypes containing transposedcopies of Tnt1 at the seedling stage (A) or at theflowering stage (B and C). A1, JesTnt6a-8 seedlings;A2, JesTnt8a-7 seedlings; A3, JesTnt1a-7 seedlings.B1, JesTnt1a mutant (left) and wild-type (right) com-pound inflorescences; B2, stereomicroscope view(103) of one JesTnt1a mutant (left) and wild-type(right) compound inflorescence. C, JesTnt6 dwarfmutants (left and right) and wild-type (center) at early(C1) and late (C2) flowering stages. D1, Southern-blotanalysis of selected JesTnt6 dwarf mutants obtainedafter two backcrosses and self-pollination.



Dow



regeneration and transformation are in keeping with theelement’s transcription pattern. Tnt1 transcription haspreviously been shown to be strongly regulated andinduced by different biotic or abiotic stress factors, suchas treatment with herbicides, microbial elicitors, wound-ing, and 2,4-dichlorophenoxyacetic acid (Pouteauet al., 1991, 1994; Grandbastien et al., 1994; Paulset al., 1994; Moreau-Mhiri et al., 1996; Mhiri et al.,1999) and during in vitro transformation of Arabidop-sis (Courtial et al., 2001) and Medicago (d’Erfurth et al.,2003). Thus, transcriptional regulation of Tnt1 (at leastin intact plants and during in vitro culture) appears tobe conserved in the heterologous host lettuce.

We also showed that transposition of integratedcopies of Tnt1 in the lettuce genome can be reactivatedby tissue culture. As already observed for T0 genotypesobtained after transformation, when transposition oc-curred, the number of Tnt1 copies was particularlyhigh. Whereas transposition was obtained for all Tnt1insertions studied, showing their functional integrity,only 38% of the regenerated plantlets studied showedtransposition events. Cocultivation of explants with anempty Agrobacterium strain prior to regeneration didnot result in a significant increase in the frequency ofTnt1 transposition compared to simple regenerationexperiments. Thus, Agrobacterium does not seem toparticularly induce Tnt1 transposition, contrary to thein vitro culture regeneration. The absence of transposedcopies observed in some of the regenerated plantlets, aswell as in some of the regenerated bushes obtained afterthe transformation experiment, might be due to the

lettuce regeneration protocol used. Indeed, lettuce is aplant that is easily regenerated in in vitro culture andthus does not need large quantities of growth regula-tors in the regeneration medium (0.15 mg/L benzyl-adenine acid and 0.3 mg/L indole acetic acid) to obtain100% regeneration of the leaf explants placed on themedium. Modification of this regeneration protocolby using higher amounts of growth regulators or 2,4-dichlorophenoxyacetic acid, previously shown to in-duce Tnt1 transposition (Pauls et al., 1994), instead ofindole acetic acid, would probably increase the per-centage of regenerated bushes containing transposedcopies of Tnt1 without seriously decreasing the per-centage of regeneration. Improvement of the in vitroprotocol for obtaining regenerated plants all containingnew transposition events have to be performed beforeto be used for large-scale mutagenesis in lettuce. Usinga lettuce genotype containing an inserted Tnt1 copy asstarting material drastically reduces the cost and com-plexity of the experiments (no antibiotic needed, noneed to transplant the explants grown on regenerationmedium before the isolation of newly regeneratedplantlets) and eliminates the risk of Agrobacterium inthe regenerated plants.

To obtain efficient mutagenesis in plants with largegenomes, it is important to use a transposable elementthat prefers to insert in genes rather than in noncodingregions. The genome of most crop plants is largecompared to the plant model Arabidopsis (145 Mb).The size of the lettuce genome is 2,639 Mb (i.e. similarto that of higher plants like maize or sunflower;

Table II. Segregation of mutations in progenies of Tnt plants

Transformed Plants

(Mutant Phenotype)Generation

Segregant Families Analyzed Phenotypesax2

Hypothesis 3:1Probability

Parental T2 Plant No. Wild Mutant

%

‘Mariska’ Tnt53a T1 37 17 1.21 27(Light yellow ligules) T2 1 8 4 – –

‘Jessy’ Tnt1a(Blocked seedlings with

white cotyledons)T2 1 222 28 25.39 0

(Abnormal flowers) T1 43 9 1.64 20T2 5 48 11 1.27 26T3 6 6 64 25 0.45 50


yellow cotyledons)T1 109 48 2.60 11

T2 1 98 35 0.12 73(No stem elongation) T1 35 17 1.64 20

T2 6 50 22 1.19 28T3 1 5 33 13 0.26 61

5 5 53 17 0.02 897 6 58 15 0.77 388 6 58 10 3.84 5


yellow cotyledons)T1 94 44 3.49 6

T2 1 105 36 0.02 88

aNumber of plants of each phenotype after the statistical homogeneity.

Mazier et al.


Dow



Table III. Characterization of Tnt1 integration sites by S-SAP

Sequence

No.

Accession

No.aSizeb BLASTNc Accession No.

or Identifierd

E-Value

Expect

(Score)

Identities

Protein Function

Predicted by

BLASTXc

Accession No./

Identifierd

E-Value

Expect

(Score)

Identities Positives

bp

HEL185 EF397544 557 Lettuce Ls3h1

mRNA for GA

3b-hydroxylasee

AB012205 7.8e-103

(2365)

473/473

(100%)

HEL131 EF397536 332 EST lettuce

contig

TIGR LsGI

TC8178

1.7e-27

(703)

149/158

(94%)

Artemisia

annua

b-caryophyllene

synthase mRNA

Q8SA63 2.5e-17

(156)

27/43

(62%)

40/43

(93%)


contig

TIGR LsGI

TC10660

1.5e-24

(641)

129/130

(99%)

*

HEL150 EF397540 309 EST sunflower

contig (similar

to Cucumis

sativus mRNA)

TIGR HaGI

TC12184

6.1e-16

(452)

194/280

(69%)

Arabinogalactan

protein

(Arabidopsis)

MAtDB

At2g14890

2.6e-09

(133)

34/80

(42%)

42/80

(52%)

HEL198 EF397551 286 EST sunflower

cDNA clone

QHE7P04

TIGR HaGI

BU022514

1.2e-15

(439)

125/164

(76%)

Unknown

protein

(Arabidopsis)

MAtDB

At3g52950

5.9e-11

(158)

26/35

(74%)

33/35

(94%)


contig

(weakly similar

to unknown

protein

Arabidopsis)

TIGR LsGI

TC10643

1.3e-12

(376)

79/83

(95%)

*


serriola

Lactuca,

cDNA clone

QGH7P11

BU008564 3.8e-11

(340)

80/92

(86%)

*


contig (similar

to amino

acid transporter

from tomato)

TIGR LsGI

TC8031

1.2e-09

(323)

89/118

(75%)

EST tomato

contig

(amino acid

transporter)

TIGR LGI

TC164746

4.2e-11

(164)

30/36

(83%)

32/36

(88%)

HEL132 EF397537 292 EST potato CK862091 5.7e-13

(292)

197/296

(66%)

myb-related

transcription

factor

(Arabidopsis)

MAtDB

At1g66230

2.7e-10

(146)

28/48

(58%)

35/48

(72%)

HEL210 EF397555 184 EST tomato contig

(similar to Hcr9-4E

protein UPjO524,

from Lycopersicum

hirsutum)

TIGR LGI

TC160018

2.7e-08

(301)

121/182

(66%)

Elicitor-inducible

LRR receptor-like

protein EILP

(tobacco)

Q9SLS3 3.5e-11

(181)

36/61

(59%)

47/61

(77%)

HEL191 EF397547 114 EST lettuce serriola

Lactuca, cDNA

clone QGG11F18

BQ995916 1.2e-07

(274)

82/114

(71%)

Cytochrome

P450-like

(Arabidopsis)

MAtDB

At3g14610

3.3e-05

(104)

19/35

(54%)

28/35

(80%)

HEL128 EF397535 179 EST tomato contig TIGR LGI

TC169575

7.4e-07

(252)

98/136

(72%)

Lateral organ

boundaries

domain

protein 12

(LBD12,

Arabidopsis)

MAtDB

At2g30130

6.8e-07

(111)

22/33

(66%)

29/33

(87%)

HEL193 EF397548 175 EST lettuce Salinas,

cDNA clone

QGB26P12

BQ855549 9.1e-06

(234)

110/170

(64%)

Unknown

protein

(Arabidopsis)

MAtDB

At1g23030

1.5e-14

(192)

37/57

(64%)

48/57

(84%)

HEL102 EF397531 132 EST lettuce contig

(Arabidopsis

unknown protein)

TIGR LsGI

TC10691

5.7e-05

(222)

56/65

(86%)

*

HEL115 EF397534 202 * Unknown

protein

MAtDB

At1g42430

8.5e-11

(127)

24/31

(77%)

27/31

(87%)


protein

(Arabidopsis)

MAtDB

At3g46420

1.6e-11

(115)

24/66

(36%)

41/66

(62%)

(Table continues on following page.)



Dow



Arumuganathan and Earle, 1991). Despite differencesin genome size, the number of genes in higher plantsis expected to be similar, although increases due toploidy changes are to be expected (Sidhu and Gill, 2004).The estimated gene number for Arabidopsis is 25,000(Arabidopsis Genome Initiative, 2000). In all eukary-otes, genes are interspersed with gene-empty regionsand the size of such regions seems to be related to thegenome size (Sidhu and Gill, 2004). Genes in the ma-jority of plants appear to be clustered; the main changefrom plants with smaller genomes (Arabidopsis andrice) to larger genomes (barley [Hordeum vulgare],maize, and wheat [Triticum aestivum]) is the overallreduction in size of the gene cluster and expansion ofthe interspersed gene-empty regions (Aert et al., 2004;Sidhu and Gill, 2004). Whereas in the smaller genome ofArabidopsis, about 45% of the chromosomal DNAcontains genes, the gene space is believed to occupyonly about 12% to 24% of the large genomes (Barakatet al., 1997). For Medicago, genes are believed to beclustered in the euchromatic regions, representing only25% of the complete genome (Kulikova et al., 2001). Asa result, transcribed regions are expected to representless than 15% of the Medicago genome (d’Erfurth et al.,2003). Accordingly, in the lettuce genome, which is 5times larger than Medicago, we would expect to find an

even lower percentage of transcribed sequences than inMedicago. Analysis of 25 different lettuce Tnt1 flankingsequences showed that the element was inserted closeto transcribed regions in one-half of the sequencesanalyzed. This Tnt1 preference to transpose into codingregions was also observed in Medicago and Arabidopsis(Courtial et al., 2001; d’Erfurth et al., 2003; Tadege et al.,2005) and also in the two other retroelements Tto1 andTos17 (Okamoto and Hirochika, 2000; Yamazaki et al.,2001; Miyao et al., 2003), even if some bias and insertionsite specificity have been shown for Tos17. Thus, retro-transposon elements such as Tnt1 have great potential astools for insertional mutagenesis in crop plants withlarge genomes like lettuce. The probability of finding aTnt1 insertion within a given gene, in the case of inser-tions restricted to coding sequences, could be calculatedusing the following formula: P 5 1 2 (1 2 [x/30,000y])n,where P 5 the probability of finding one Tnt1 insertwithin a given gene, x 5 the length of the gene, y 5 theaverage length of the genes in the plant studied, and n 5the number of Tnt1 inserts present in the population.This calculation assumes that the transcriptome is repre-sented by 30,000 genes and that the Tnt1 insertions arerandomly distributed inside expressed sequences. Basedon this equation and an average gene length of 3.5 kb, wecan estimate that 89,871 insertions (instead of 2,225,400 if

Table III. (Continued from previous page.)

Sequence

No.

Accession

No.aSizeb BLASTNc Accession No.

or Identifierd

E-Value

Expect

(Score)

Identities

Protein Function

Predicted by

BLASTXc

Accession No./

Identifierd

E-Value

Expect

(Score)

Identities Positives


protein

(Arabidopsis)

MAtDB

At3g63430

1.9e-06

(116)

23/37

(62%)

27/37

(72%)

HEL104 EF397532 120 * *

HEL105 EF397533 117 * *

HEL136 EF397538 211 * *

HEL161 EF397541 254 * *

HEL178 EF397542 263 * *

HEL184 EF397543 421 * *

HEL186 EF397545 372 * *

HEL194 EF397549 147 * *

aAccession number of the different insertion sites isolated in lettuce. bNumbers indicate the size of the lettuce sequences obtained aftersequencing of S-SAP fragments. cHighest similarities are shown. In brackets, similarities found with the EST contig. *, No similarities found or Evalues less than 1e-05. dTIGR LsGI, Lettuce TIGR database; TIGR HaGI, sunflower TIGR database; TIGR LGI, tomato TIGR database; MAtDB,Arabidopsis database ePerfect match.

Figure 7. PCR analysis of 16 T2 plants derivedfrom the JesTnt6a genotype (T1 plant JesTnt6a-5)segregating for the dwarf mutation. T3 progenyobtained from each plant were also analyzed forthe mutation segregation. PCR was performedusing the three oligonucleotides LTR1R (comple-mentary to Tnt1), Gi1F, and Gi3R (complemen-tary to GA b-hydroxylase, amplifying a 663-bpfragment). C, ‘Jessy’ wild type. L, One-kilobaseladder. According to the S-SAP sequence of theTnt1 insertion into the GA b-hydroxylase, theexpected fragment amplified with LTR1R andGI3R characteristic of this insertion is 441 bp.

Mazier et al.


Dow



insertions are randomly distributed in the whole genome)would be required to have a 95% probability of taggingany given gene. With a postulate of 10 such Tnt1 inser-tions per line in lettuce, the same level of saturation couldbe obtained in 8,987 lines (95% probability).

In addition, we demonstrated the feasibility of usingTnt1 as a gene-tagging tool in lettuce by isolating theTnt1 insertion responsible for dwarf mutation. Even ifthe involvement of Tnt1 insertions was not demon-strated for the other mutant phenotypes observed in theprogeny of the T0 plants, their surprising number withrespect to the small starting T0 population generatedprovides indirect additional proof of Tnt1 preferencefor insertion in transcribed regions and its potentiallyhighly mutagenic characteristics.

Several transposon systems have already been con-structed for the functional analysis of genes (reverse-genetics analysis) and gene tagging in plants (forreview, see Parinov and Sundaresan, 2000; Walbot, 2000;Srinivasan and Sundaresan, 2001). Although these sys-tems have been successfully used to isolate genes andidentify mutants, some problems associated with theiruse have also been reported. First, DNA transposonsgenerally jump to locations near the original insertionsite, often to linked positions (Parinov and Sundaresan,2000). Second, these types of transposable elementstranspose by a cut-and-paste mechanism, causing un-stable mutations or imprecise excisions that are difficultto detect. In lettuce, for instance, imprecise excision of aDs element produced mutants that were not taggedwith the element (Okubara et al., 1997). Because theytranspose by a copy-and-paste mechanism, no suchproblems can occur with retrotransposons, makingthem especially suitable for obtaining flanking plantsequences as well as for facilitating PCR-based reverse-genetics screening of DNA pools. At this time, only tworetrotransposons have been used as insertion mutagensin plants: Tos17 from rice and Tnt1 from tobacco(Yamazaki et al., 2001; d’Erfurth et al., 2003). Interest-ingly, Tos17 has been successfully used in rice to gen-erate more than 47,000 mutant lines and has alreadyproven to be a useful tool for functional genomics inthis plant (Miyao et al., 2003; Kaneko et al., 2004;Nonomura et al., 2004), and collaborative efforts areunder way in Europe and the United States to generatearound 20,000 Tnt1-tagged Medicago lines in the next 3years (Tadege et al., 2005).

In this article, we showed that insertional mutagen-esis and gene tagging using the retrotransposon Tnt1 isalso feasible in lettuce, despite its large genome, 5 to 6times bigger than that of rice or Medicago. The resultsobtained show that retrotransposons should not beunderestimated as alternative tools especially in plantsamenable to tissue culture but in which classical ge-nomics and whole-genome sequencing are not easy toachieve. In the future, this could help to identify thesimilarities or differences in coding sequences or codedfunctions existing between well-studied model speciesand a crop plant with agronomical useful traits orbetween plants belonging to distant families.

MATERIALS AND METHODS

Plant Material and Plant Growth Conditions

Two butterhead lettuce plants (Lactuca sativa) were used: ‘Mariska’ from

Numhems and ‘Jessy’ from Caillard. For germination, seeds were sown in petri

dishes containing filter paper moistened with water and exposed to 4�C in the

dark for 48 h before being placed in a growth room (24�C, 16-h photoperiod, 60

mE m22 s21) for 5 to 10 d. Seedlings were first grown on compost in 7-cm pots in a

greenhouse for 3 weeks before being transplanted in 3-L compost pots and

watered with nutritive solution.

Mutant plants whose stems failed to elongate in the greenhouse were

treated by applying 5 mL of a solution containing 0.02% (w/v) of GA3

(C19H2206; catalog no. G 0907; Duchefa) to the top of the plants in the

greenhouse at the heading stage.

T-DNA Vectors and Bacterial Strain

The Tntk23 vector (Fig. 1A) is a derivative of the pBin19 vector carrying the

autonomous Tnt1 retroelement (Lucas et al., 1995). The pHLV55501 vector is

also derived from pBin19 and carries a translational LTR-GUS-intron fusion

(Courtial et al., 2001).

Plasmids were introduced in Agrobacterium tumefaciens strain C58 pGV2260

(Deblaere et al., 1985).

In Vitro Culture Regeneration and

Genetic Transformation

Lettuce genetic transformation procedures are described by Dinant et al.

(1997) and the regeneration medium used is described by Mazier et al. (2004).

Leaves excised from 10-d-old seedlings (cultivated aseptically) were inoculated

with A. tumefaciens strain C58 pGV2260 carrying the constructs to be intro-

duced. Transformant buds were selected on 200 mg L21 kanamycin-containing

regeneration medium and transplanted individually onto rooting medium

with 50 mg L21 kanamycin. Vigorous plantlets were transplanted to pots with

peat soil in a growth chamber (22�C day/16�C night, 16-h photoperiod) before

transfer to a greenhouse for flowering and self-pollination.

Newly in vitro-regenerated plantlets (without transformation) were pro-

duced from true leaves taken from sterile seedlings grown in vitro or from

selected plants growing in the greenhouse, surface sterilized by immersing in

a disinfecting solution for 30 min (Mazier et al., 2004), followed by three rinses

with sterilized deionized water. In vitro regeneration procedures were

performed as described by Mazier et al. (2004).

Molecular Analysis

Genomic DNA was extracted from leaf tissues using a modified cetyltri-

methylammonium bromide method (Bernatzky and Tanksley, 1986). For

Southern blots, 12 mg of DNA were digested with the appropriate restriction

enzyme as recommended by the manufacturer. After separation on a 0.8%

(w/v) Tris-acetate EDTA 13 agarose gel and denaturation and neutralization

using standard procedures (Sambrook et al., 1989), the DNA was transferred

onto a Hybond N1 membrane (Amersham). The Kana probe (805-bp HindIII

fragment from the pABDI plasmid [Paszkowski et al., 1984]) containing the

nptII gene was labeled by random priming. The gag and reverse transcription

(RT) probes were generated by PCR using Tnt6131 (5#-TGGTATCAGAGCA-

CAGGTT-3#) and Tnt12392 (5#-TCATTGAGTAGAAGAGCCGA-3#) primers

for gag and Tnt30841 (5#-CTTCCACAGAGTATGTCCTCATCAGT-3#) and

Tnt45232 (5#-CATCAATGTGTTTGGTCCTTGC-3#) primers for RT.

Hybridization with 32P-labeled probes was performed at 65�C in hybrid-

ization buffer (750 mM NaCl, 125 mM sodium citrate, 0.6% SDS, 50 mM

Na2HPO4/NaH2PO4, pH 7.5, 53 Denhart, 2.5 mM EDTA, 5% dextran sul-

fate, 2.5 mg DNA salmon sperm) for 16 to 24 h. The filters were then washed

one time in 23 SSC; 0.1% SDS at 65�C for 20 min followed by one wash in

13 SSC; 0.05% SDS at 65�C for 20 min and a final wash in 0.53 SSC; 0.05%

SDS at 65�C for 10 to 20 min. Filters were exposed and analyzed by

autoradiography.

PCR analysis of the Tnt1 insertion responsible for the dwarf mutation was

performed using primers LTR1R, 5#-GGCTACCAATCCAACAAGGA-3#(complementary to Tnt1); Gi1F, 5#-CAAACGCCATGAAGCTTGT-3# (GA

b-hydroxylase nucleotide positions 210–229); and Gi3R, 5#-GGTGCAGCA-

CACTTGGATAC-3# (complementary to GA b-hydroxylase nucleotide posi-

tions 852–872).



Dow



Isolation of Insertion Sites

The sequences flanking Tnt1 integration sites were isolated by S-SAP and

sequenced as described by Courtial et al. (2001). The genomic DNA was

digested with Csp6I (Fermentas).

Sequence Analysis

Sequence chromatograms were analyzed using phred software (Ewing

et al., 1998; Ewing and Green, 1998) with the following parameters: -trim_

alt ‘‘’’ -trim_cutoff 0.10. This step produced the corresponding nucleotide

sequences and quality data files. The sequences were then cleaned using local

software to remove the retransposon sequence (X13777.1) and the primer

sequences (CTGGACGATGAGTCCTGAGA and TATCTCAGGACT). Homol-

ogies of the resulting cleaned sequences with different databases were then

searched for using Washington University BLAST (WU BLAST) Version 2 (W.

Gish, 1996–2003, http://blast.wustl.edu). Both BLASTN and TBLASTX anal-

ysis were performed with four TIGR Gene Indices databases (http://www.

tigr.org/tdb/tgi/index.shtml): Lettuce 2.0, Sunflower 3.0, Tomato 10.1, and

Potato 10.0. BLASTX analyses were performed with two protein sequences

databases, MAtDB, the MIPS Arabidopsis (v090704 version) database (http://

mips.gsf.de/proj/thal/db), and both Swiss-Prot and TrEMBL (10/11/2005

update) available on the ExPASy Proteomics Server (http://www.expasy.org).

GUS Assays

Histochemical staining for GUS activity was carried out according to

Jefferson (1987).

Sequence data from this article can be found in the GenBank/EMBL data

libraries under accession numbers EF397531 to EF397555.

ACKNOWLEDGMENTS

We are grateful to V. Sarnette for excellent plant care and backcrossings.

We thank Y. Bellec and E. Martin for technical assistance during cultivation of

T0 and T1 plants and their progenies. We wish to thank M. Pitrat, P. Ratet, and

S. Munos for their advice and helpful comments, and J. Chadoeuf for help in

calculations and probability formulas.

Received September 26, 2006; accepted January 30, 2007; published March 9,

2007.

LITERATURE CITED

Aert R, Sagi L, Volckaert G (2004) Gene content and density in banana

(Musa acuminata) as revealed by genomic sequencing of BAC clones.

Theor Appl Genet 109: 129–139

Arabidopsis Genome Initiative (2000) Analysis of the genome sequence of

the flowering plant Arabidopsis thaliana. Nature 408: 796–815

Arumuganathan K, Earle ED (1991) Nuclear DNA content of some impor-

tant plant species. Plant Mol Biol Rep 9: 208–218

Barakat A, Carels N, Bernardi G (1997) The distribution of genes in the

genomes of Gramineae. Proc Natl Acad Sci USA 94: 6857–6861

Bernatzky R, Tanksley SD (1986) Genetics of actin related sequences in

tomato. Theor Appl Genet 72: 314–321

Chupeau M-C, Maisonneuve B, Bellec Y, Chupeau Y (1994) A Lactuca

universal hybridizer, and its use in creation of fertile interspecific

somatic hybrids. Mol Gen Genet 245: 139–145

Courtial B, Feuerbach F, Eberhard S, Rohmer L, Chiapello H, Camilleri C,

Lucas H (2001) Tnt1 transposition events are induced by in vitro

transformation of Arabidopsis thaliana and transposed copies integrate

into genes. Mol Genet Genomics 265: 32–42

Deblaere R, Butebier B, De Greve H, Deboeck F, Schell J, Van Montagu

M, Leemans J (1985) Efficient octopine Ti plasmid-derived vectors

for Agrobacterium-mediated gene transfer. Nucleic Acids Res 13:

4777–4788

d’Erfurth I, Cosson V, Eschstruth A, Lucas H, Kondorosi A, Ratet P (2003)

Efficient transposition of the Tnt1 tobacco retrotransposon in the model

legume Medicago truncatula. Plant J 34: 95–106

Dinant S, Maisonneuve B, Albouy J, Chupeau Y, Chupeau M-C, Bellec Y,

Gaudefroy F, Kusiak C, Souche S, Robaglia C, et al (1997) Coat protein

gene-mediated protection in Lactuca sativa against lettuce mosaic poty-

virus strains. Mol Breed 3: 75–86

Dubois V, Botton E, Meyer C, Rieu A, Bedu M, Maisonneuve B, Mazier M

(2005) Systematic silencing of a tobacco nitrate reductase transgene in

lettuce (Lactuca sativa L.). J Exp Bot 56: 2379–2388

Ewing B, Green P (1998) Base-calling of automated sequencer traces using

phred. II. Error probabilities. Genome Res 8: 186–194

Ewing B, Hillier L, Wendl MC, Green P (1998) Base-calling of automated

sequencer traces using phred. I. Accuracy assessment. Genome Res 8:

175–185

Grandbastien M-A, Audeon C, Casacuberta JM, Grappin P, Lucas H,

Moreau C, Pouteau S (1994) Functional analysis of the tobacco Tnt1

retrotransposon. Genetica 93: 181–189

Grandbastien MA, Spielmann A, Caboche M (1989) Tnt1, a mobile

retroviral-like transposable element of tobacco isolated by plant cell

genetics. Nature 337: 376–380

Hirochika H (1997) Retrotransposons of rice: their regulation and use for

genome analysis. Plant Mol Biol 35: 231–240

Jefferson RA (1987) Assaying chimeric genes in plants: the GUS gene

fusion system. Plant Mol Biol Rep 5: 387–405

Kaneko M, Inukai Y, Ueguchi-Tanaka M, Itoh H, Izawa T, Kobayashi Y,

Hattori T, Miyao A, Hirochika H, Ashikari M, et al (2004) Loss-of-

function mutations of the rice GAMYB gene impair a-amylase expres-

sion in aleurone and flower development. Plant Cell 16: 33–44

Krysan PJ, Young JC, Sussman MR (1999) T-DNA as an insertional

mutagen in Arabidopsis. Plant Cell 11: 2283–2290

Kulikova O, Gualtieri G, Geurts R, Kim DJ, Cook D, Huguet T, de Jong

JH, Fransz PF, Bisseling T (2001) Integration of the FISH pachytene and

genetic maps of Medicago truncatula. Plant J 27: 49–58

Kumar A, Bennetzen JL (1999) Plant retrotransposons. Annu Rev Genet 33:

479–532

Kumar A, Hirochika H (2001) Applications of retrotransposons as genetic

tools in plant biology. Trends Plant Sci 6: 127–134

Lelivelt CLC, McCabe MS, Newell CA, de Snoo CB, van Dun KMP,

Birch-Machin I, Gray JC, Mills KHG, Nugent JM (2005) Stable

plastid transformation in lettuce (Lactuca sativa L.). Plant Mol Biol 58:

763–774

Lucas H, Feuerbach F, Kunert K, Grandbastien MA, Caboche M (1995)

RNA-mediated transposition of the tobacco retrotransposon Tnt1 in

Arabidopsis thaliana. EMBO J 14: 2364–2373

Mazier M, German-Retana S, Flamain F, Dubois V, Botton E, Sarnette V,

Le Gall O, Candresse T, Maisonneuve B (2004) A simple and efficient

method for testing Lettuce mosaic virus resistance in in vitro cultivated

lettuce. J Virol Methods 116: 123–131

Mhiri C, De Wit PJM, Grandbastien M-A (1999) Activation of the promoter

of the Tnt1 retrotransposon in tomato after inoculation with the fungal

pathogen Cladosporium fulvum. Mol Plant Microbe Interact 12: 592–603

Michelmore R, Marsh E, Seely S, Landry B (1987) Transformation of

lettuce (Lactuca sativa) mediated by Agrobacterium tumefaciens. Plant Cell

Rep 6: 430–442

Miyao A, Tanaka K, Murata K, Sawaki H, Takeda S, Abe K, Shinozuka Y,

Onosato K, Hirochika H (2003) Target site specificity of the Tos17 retro-

transposon shows a preference for insertion within genes and against

insertion in retrotransposon-rich regions of the genome. Plant Cell 15:

1771–1780

Moreau-Mhiri C, Morel J-B, Audeon C, Ferault M, Grandbastien M-A,

Lucas H (1996) Regulation of expression of the tobacco Tnt1 retrotrans-

poson in heterologous species following pathogen-related stresses.

Plant J 9: 409–419

Nonomura K-I, Nakano M, Fukuda T, Eiguchi M, Miyao A, Hirochika H,

Kurata N (2004) The novel gene HOMOLOGOUS PAIRING ABERRA-

TION IN RICE MEIOSIS1 of rice encodes a putative coiled-coil protein

required for homologous chromosome pairing in meiosis. Plant Cell 16:

1008–1020

Okamoto H, Hirochika H (2000) Efficient insertion mutagenesis of Arabi-

dopsis by tissue culture-induced activation of the tobacco retrotrans-

poson Tto1. Plant J 23: 291–304

Okubara PA, Arroyo-Garcia R, Shen KA, Mazier M, Meyers BC, Ochoa

OE, Kim S, Yang C-H, Michelmore RW (1997) A transgenic mutant of

Lactuca sativa (lettuce) with a T-DNA tightly linked to loss of downy

mildew resistance. Mol Plant Microbe Interact 10: 970–977

Mazier et al.


Dow



Parinov S, Sundaresan V (2000) Functional genomics in Arabidopsis: large-

scale international mutagenesis complements the genome-sequencing

project. Curr Opin Biotechnol 11: 157–161

Paszkowski J, Shillito RD, Saul M, Mandak V, Hohn T (1984) Direct gene

transfer to plants. EMBO J 3: 2717–2722

Pauls PK, Kunert K, Huttner E, Grandbastien MA (1994) Expression of the

tobacco Tnt1 retrotransposon promoter in heterologous species. Plant

Mol Biol 26: 393–402

Pouteau S, Grandbastien M-A, Boccara M (1994) Microbial elicitors of

plant defence responses activate transcription of a retrotransposon.

Plant J 5: 535–542

Pouteau S, Huttner E, Grandbastien MA, Caboche M (1991) Specific

expression of the tobacco Tnt1 retrotransposon in protoplasts. EMBO J

10: 1911–1918

Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning: A Laboratory

Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor,

NY

Sidhu D, Gill KS (2004) Distribution of genes and recombination

in wheat and other eukaryotes. Plant Cell Tissue Organ Cult 79:

257–270

Srinivasan R, Sundaresan V (2001) Transposons as tools for functional

genomics. Plant Physiol Biochem 39: 243–252

Tadege M, Ratet P, Mysore S (2005) Insertional mutagenesis: a Swiss army

knife for functional genomics of Medicago truncatula. Trends Plant Sci

10: 229–235

Walbot V (2000) Saturation mutagenesis using maize transposons. Curr

Opin Plant Biol 3: 103–107

Yamazaki M, Tsugawa H, Miyao A, Yano M, Wu J, Yamamoto S,

Matsumoto T, Sasaki T, Hirochika H (2001) The rice retrotransposon

Tos17 prefers low-copy number sequences as integration targets. Mol

Genet Genomics 265: 336–344

Yang C-H, Carroll B, Scofield S, Jones J, Michelmore R (1993a) Trans-

activation of Ds elements in plants of lettuce (Lactuca sativa). Mol Gen

Genet 241: 389–398

Yang C-H, Ellis JG, Michelmore RW (1993b) Infrequent transposition of Ac

in lettuce, Lactuca sativa. Plant Mol Biol 22: 793–805



Dow



Successful Gene Tagging in Lettuce Using the Tnt1 … · ‘Jessy,’ were cocultivated with an...

Documents

Transcript of Successful Gene Tagging in Lettuce Using the Tnt1 … · ‘Jessy,’ were cocultivated with an...