• Open Access

Transposon tagging in diploid strawberry

Authors


(Tel 540 231 5584; fax 540 231 3083; email potato@vt.edu)

Summary

Fragaria vesca was transformed with a transposon tagging construct harbouring amino terminally deleted maize transposase and EGFP (Ac element), NPTII, CaMV 35S promoter (P35S) driving transposase and mannopine synthase promoter (Pmas) driving EGFP (Ds element). Of 180 primary transgenics, 48 were potential launch pads, 72 were multiple insertions or chimaeras, and 60 exhibited somatic transposition. T1 progeny of 32 putative launch pads were screened by multiplex PCR for transposition. Evidence of germ-line transposition occurred in 13 putative launch pads; however, the transposition frequency was too low in three for efficient recovery of transposants. The transposition frequency in the remaining launch pads ranged from 16% to 40%. After self-pollination of the T0 launch pads, putative transposants in the T1 generation were identified by multiplex PCR. Sequencing of hiTAIL-PCR products derived from nested primers within the Ds end sequences (either P35S at the left border or the inverted repeat at the right border) of T1 plants revealed transposition of the Ds element to distant sites in the strawberry genome. From more than 2400 T1 plants screened, 103 unique transposants have been identified, among which 17 were somatic transpositions observed in the T0 generation. Ds insertion sites were dispersed among various gene elements [exons (15%), introns (23%), promoters (30%), 3′ UTRs (17%) as well as intergenically (15%)]. Three-primer (one on either side of the Ds insertion and one within the Ds T-DNA) PCR could be used to identify homozygous T2 transposon-tagged plants. The mutant collection has been catalogued in an on-line database.

Introduction

The recently sequenced diploid strawberry genome (Shulaev et al., 2011) predicted the presence of approximately 34 000 genes compared to 26 000 for Arabidopsis (Ito et al., 2005). Many strawberry genes can be expected to control the development of fleshy fruit, runnering, flavonoids and other traits unique to fruit crops. Fragaria vesca, with its rapid life cycle, compact size and abundant seed production, is a potential model organism for rosaceous and other fruit crops. Therefore, a well-characterized knockout population of diploid strawberry would be a welcome addition to the toolkit for forward genetics studies. Despite the availability of efficient transformation protocols for F. vesca Hawaii 4 (Oosumi et al., 2006), generating a knockout population of strawberry with the possibility of having at least one mutation in every gene is prohibitively labour intensive (Oosumi et al., 2010). Hence, transposon tagging whereby a single transgenic launch pad can generate an unlimited number of independent germ-line transposants, as described for Arabidopsis (Bradshaw and Schemske, 2003; Ipek et al., 2006; Ito et al., 2005; Kim et al., 2005; Kuromori et al., 2004), rice (Qu et al., 2009) and barley (Ayliffe et al., 2007), is highly desirable. Both one-component systems [Activator (Ac) and Dissociator (Ds) in the same vector construct] and two-component systems (Ac and Ds in separate constructs) are available for transposon tagging (Greco et al., 2003; Upadhyaya et al., 2006). However, only a one-component system with Ac bearing transposase and Ds carrying a selectable marker is practical for strawberry, because of the ease of self-pollination of diploid strawberry compared to the tedium of emasculation of flower buds, pollen collection and cross-pollination between independently transformed Ac and Ds plants that would be required for transposon tagging using the two-component systems.

Movement of Ds in launch pad (LP) genotypes where transposase is driven by a constitutive promoter is somewhat unpredictable. It can occur somatically or gametophytically or both. In Arabidopsis, transposition has been found to be primarily local with preferential Ds reinsertion along the chromosome where the launch pad was situated (Ito et al., 2005; Kuromori et al., 2004). In a population of 979 transposon-tagged lines of rice derived from launch pads on three different chromosomes, the frequency of intrachromosomal transpositions was 24%, with 76% interchromosomal transpositions (Kolesnik et al., 2004). One strategy to improve on the frequency of interchromosomal transposition has been to use dual selection that allows both visual tracking of the immobilized Ac and identification of Ds (Qu et al., 2008). In other plants such as cabbage (Yu et al., 2010), carrot and barley, small populations of transposants have been developed with promising results of more global interchromosomal transposition; however, the number of transposants available in these systems compared to Arabidopsis and rice makes the analyses less conclusive.

Our objective in this study was to evaluate a transposon tagging system for the diploid strawberry, F. vesca. Because of the impracticality of large-scale hand pollinations in this rosaceous crop, we used a single T-DNA construct that comprised both Ac and Ds elements of maize. We report here the transposition efficiency of some of these lines and their potential for generating an activation-tagged mutant population.

Results

Transformation

Transformation of leaf segments of F. vesca PI 551572 was conducted according to Oosumi et al. (2006) or using recent modifications to the method (Pantazis et al., 2012) in eight separate experiments. When possible, two different shoots were taken from a callus to yield a pair of somaclones for each independently transformed callus. Of 207 transformed plants obtained, 180 (87%) were unique. Multiplex screening of the 180 unique transformed plants revealed the presence of an empty donor site (EDS) in 60 plants (33%), indicating that transposition had occurred somatically, presumably either during transformation and regeneration from callus or during early seedling growth prior to leaf sampling for multiplex screening (Figure 1a). Seeds were collected from the somatic transposants and the Ds position was determined by hiTAIL (thermal asymmetric interlaced)-PCR when possible, but they were excluded from progeny analysis because additional movement of the Ds element was not expected due to the separation of P35S from the transposase gene after transposition. Among the 120 remaining putative launch pads, Southern blot revealed 48 (40% of the putative launch pads or 27% of the transformed plants) single-insertion transformants (Figure S1). In most instances where two different shoots were taken from the same callus, multiplex screening revealed a similar configuration of transgene elements (data not shown); however in six cases, one somaclone was found to have a full donor site (FDS) and the other an EDS, implying that shoots had likely regenerated from a genetic mosaic of callus differing by somatic transposition of Ds or that transposition had occurred somatically in one regenerated plant but not its somaclone.

Figure 1.

 (a) Multiplex PCR analysis of nine AcDs transgenics (Lanes 5 and 7–14). The PCR mix contained nine primers to reveal the presence of an empty donor site (EDS), full donor site (FDS), nptII on the Ds element, vector backbone VirA and a segment of Fragaria vesca gene FVAAT3 (GenBank AF193789) used to distinguish nontransgenic escapes from failed reactions. LP3, 4, 6, 7, 9 and 10 (lanes 7, 8, 10, 11, 13 and 14) are putative launch pads with intact FDS. Somatic transpositions (st) occur in lanes 5, 9 and 12 where the Ds element excised prematurely in the T0 plants prior to meiosis and the EDS was apparently lost; FV10 (lane 4) is a nontransgenic control. Lanes 1 and 15: HyperLadder™ IV; (b) Multiplex PCR screening of 28 T1 plants derived from self-pollination of a launch pad transformed with an AcDs construct. The reaction mix contained seven primers including forward and reverse for both a segment of Fragaria vesca gene FvAAT3 and nopaline synthase (NPTII). The remaining three primers amplified either the FDS spanning the AcDs junction or the EDS with primers situated in the Ac element on either side of Ds. Putative transposants occur in lanes 4 (NPTII) and lanes 8, 15, 20, 21, 22, 23 and 25 (NPTII and EDS). Wild-type plants (lane 13) and those with both an EDS and FDS (lanes 2, 3, 5, 6, 7, 9, 10, 12, 14, 16, 17, 24, 26 and 29) were discarded. Occasional plants with an FDS and no trace of an EDS (lane 19) were retained and T2 seed collected for the possibility of screening for transposition. Lanes 1 and 30 are HyperLadder™ IV.

TPase-hiTAIL revealed the insertion sites (Table 1) of the T-DNA of 27 putative launch pads (T0 plants estimated to have a single insertion per Southern blot 1 using an FDS probe and predicted to have an FDS without an EDS per multiplex PCR). The distribution of T-DNA insertions across linkage groups was apparently random with two to seven putative launch pads occurring on each of the seven linkage groups and one occurring on an unlinked scaffold. The greatest proportion of T-DNA insertions (11/27 or 41%) occurred in promoter regions, seven (26%) were within genes (six introns and one exon), seven (26%) were intergenic, one of which occurred in a highly repetitive retrotransposon region, and two (7%) occurred in 3′ UTR regions. A shortage of seed for some of the putative launch pads limited our evaluation of T1 progenies to 22 unique and five somaclonal pairs (Table 1). The insertion sites of T0 launch pads with regard to genomic regions are provided in Table 1.

Table 1.   Insertion sites (scaffold number of the version 8 strawberry genome assembly, genomic region and putative function of nearest hybrid gene model) of Fragaria vesca launch pads, and germ-line transposition frequency among T1 plants screened by multiplex PCR. The last ten entries (LP43 through LP141) are somaclonal pairs of putative launch pads regenerated from the same transformed calluses
Launch padInsertion site per TPase hiTAILGenomic regionPutative gene functionLinkage group T 1 plants screenedEDS + NPTII NPTII Transposition frequency (%)
  1. EDS, empty donor site.

LP115scf0513190:1121231Retrotransposon 7000ND
LP126scf0513081:1178005Promoter: gene 296221Beta-galactosidase 2 (Beta-gal) (probable)521000
LP139    13000
LP114    190316
LP98scf0513133:168767Intron: gene 08446Transmembrane protein 208 (probable)2236026
LP83scf0513177:2001653Intergenic 624000
LP103scf0513113:1072380Promoter: gene 26734TMV resistance protein N (probable)70000
LP136scf0512992:140412Intron: gene 23799Protein YSC84 (LAS17-binding protein 4) (probable)Not linked0000
LP89scf0513158:8881733′ UTR: gene 03729Monoglyceride lipase (MGL) (probable)50000
LP119scf0513131:8656773′ UTR: gene 27493 26000
LP8scf0513041:350609Promoter: gene 20827Putative U-box domain-containing protein 50 (probable)10000
LP7scf0513078:3005363′ UTR: gene 22118hypothetical protein542672719
LP37scf0512959:437714Intergenic 123000
LP28bscf0513153:32989Intron: gene 03904Transketolase, chloroplastic (TK), precursor (similar to)7103140
LP19scf0513168:3505179Promoter: gene 16734Putative disease resistance protein RGA1 (probable)630117
LP41scf0513177:4049746Intron: gene 13646Probable leucine-rich repeat receptor-like protein kinase At5g49770, precursor633000
LP55bscf0513142:334172Promoter: gene 18503SKP1-like protein 1A (SKP1-like 1) (probable)313000
LP69scf0513157:101323Intergenic 3142121
LP10scf0513088:878458Promoter: gene 28793Protein TRANSPARENT TESTA 12 (putative)626088235
LP38scf0513109:18740Intergenic 59000
LP25bscf0512991:1117130Promoter: gene 23873FvAP2-EREBP90, ethylene-responsive transcription factor19525431
LP58scf0513192:3394232Intergenic 138013
LP40scf0513136:689083Promoter: gene 09275Hsc70-interacting protein (Hip) (probable)733000
LP44    29207
LP57    34000
LP21 scf0513151:338818   25000
LP15scf0513147:566950Promoter: gene 11527GDSL esterase/lipase At5g45910, precursor (putative)40000
LP29bscf0513158:3283163Promoter: gene 04001Probable LRR receptor-like serine/threonine protein kinase At1g14390, precursor40000
LP3scf0513143:10780513′ UTR: gene 10639Putative tyrosine-protein phosphatase capC (probable)50000
LP5scf0512967:115669Exon: gene 02385Probable gluconokinase30000
LP36bscf0513157:305822Promoter: gene 0329721.7-kDa class VI heat-shock protein (AtHsp21.7) (putative)30000
LP43scf0513075:616123—same as LP59Promoter: gene 22043Poly [ADP-ribose] polymerase 3 (PARP-3) (putative)220000
LP59scf0513075:616135—same as LP43Promoter: gene 22043Poly [ADP-ribose] polymerase 3 (PARP-3) (putative)2182011
LP45same as LP18Exon: gene 16736 614000
LP18scf0513168:3516467 same as LP45Exon: gene 16736Small acidic protein (probable)634000
LP31scf0513177:2924018—same as LP11Intron: gene 13775 6110123
LP11scf0513177:2924182—same as LP31Intron: gene 13775Ferritin-3, chloroplastic, precursor (similar to)60000
LP68same as LP20Intron: gene 31794 551873916
LP20scf0513098:3419420—same as LP68Intron: gene 31794AP-1 complex subunit mu-1 (similar to)5339721927
LP135same as LP141Intergenic 621000
LP141scf0513176:1743612—same as LP135Intergenic 611000
Total    236235050 

Transposition

The intended strategy for the selection of putative transposants was to grow T1 seedlings on medium with kanamycin to eliminate wild types, then to screen for GFP expression in surviving seedlings, discarding GFP-positive seedlings expected to have an intact AcDs element, and retaining GFP-negative seedlings expected to have transposed the Ds element (Figure 2). Transposition would have removed Pmas from driving expression of GFP. Attempts to visualize GFP expression in putative launch pads with an intact AcDs element were successful only in a single case (Figure 3). Sequencing the T-DNA region of this putative single-insertion launch pad plant revealed that its T-DNA had rearranged, placing GFP under the control of P35S; no transposants were found among its progeny. Without obvious GFP expression in the remaining launch pads, despite PCR evidence of an intact AcDs junction between Pmas and EGFP, we used multiplex PCR on T1 plants (Figure 1b) to verify the presence of the T-DNA elements: FDS, EDS and Ds. When seed germination was adequate, we screened 24 T1 seedlings per T0 launch pad; if transposition appeared promising, we screened additional seedlings from selected families. In total, more than 2400 T1 seedlings derived from 32 putative launch pads (including five pairs of somaclones) were screened by multiplex PCR. Transposition appeared to have occurred by the presence of an EDS in T1 progeny of 12 (38%) of the T0 families. Of these, ten launch pads (including somaclonal pair LP20/LP68) exhibited sufficiently frequent transposition (>10%) to warrant more extensive T1 screens (Table 1). The insertion sites of these functional launch pads included two in promoter regions, four in introns and one in a 3′ UTR; only one was intergenic.

Figure 2.

 Model of the T-DNA region of the transposon tagging AcDs construct used for the transformation of Fragaria vesca. RB, right border; TPase, maize transposase; Ds, Dissociator part of the element; P35S, cauliflower mosaic virus promoter; NPTII, neomycin phosphotransferase gene; Pmas, mannopine synthase promoter; EGFP, enhanced green fluorescent protein gene; LB, left border; triangular borders of the Ds element represent inverted repeats. The Activator (Ac) part of the construct occurs on both sides of the Ds element. The T-DNA was inserted into plasmid psKI074. Primer positions are indicated by blue arrows: 1 and 2 (FDS), 3 and 4 (NPTII ), 1 and 5 (EDS).

Figure 3.

 EGFP fluorescence in floral tissue of Fragaria vesca transformed with pAc-Ds-EG. The vibrant fluorescence occurring in petals and stigma (a) as well as sepals (b) of a single transformed plant was demonstrated to be the result of an inverted Ds element that placed P35S originally expected to drive TPase in front of the EGPF gene, with intervening inverted repeat. Plants bearing the original construct as shown in Figure 2 exhibited little or no EGFP fluorescence.

Sequences derived from P35S-hiTAIL-PCR or IR-hiTAIL-PCR (Figure S2) of the T1 putative transposant seedlings revealed the occurrence of both unique, presumably gametophytic transpositions, and common insertions among groups of T1 transposants from the same launch pad, suggesting that transposition had occurred somatically prior to gametogenesis in some instances (Figure S2). After sequencing and aligning hiTAIL PCR products, we determined that the frequencies of unique T1 transposants varied with the launch pad (Table 2), with LP7 and LP25b generating predominantly unique germ-line transpositions, LP10 generating many T1 seedlings that all carried the same transposition and LP 20 generating a mixture of unique transpositions and sets of identical transpositions. For the three most prolific T0 launch pads, the distribution of transposants was global, that is, LP7 and LP20 were situated on linkage group 5 and their transposon-tagged T1 derivatives occurred on all seven linkage groups (Table 1, Figure 4). Likewise, LP25b was situated on linkage group 1 and its transposon-tagged T1 lines occurred on six of the seven linkage groups. The insertion sites of germ-line transpositions were apparently random with 28 in promoter regions, 14 in exons, 20 in introns, 17 in 3′ UTRs and ten in intergenic regions (Table S2). Where somaclonal pairs of putative launch pads were screened for transposition in the T1, similar transposition profiles were obtained for both members of each pair (Table 1).

Table 2.   Frequencies of unique and common transpositions among T1 seedlings classified as putative transposants with either an empty donor site (EDS)/NPTII or simply an NPTII profile by multiplex PCR
Launch padTransposant profilehiTAIL-PCR sequences obtainedhiTAIL-PCR products sequenced*Unique transposants
  1. *Identically sized hiTAIL-PCR products from different seedlings were assumed to represent somatic transposition prior to gametogenesis. Only a single band was cut and sequenced if the identity was obvious.

LP7EDS/NPTII867752
NPTII 15104
LP10EDS/NPTII8871
NPTII 200
LP20EDS/NPTII66258
NPTII 1400
LP25bEDS/NPTII27146
NPTII 411
Figure 4.

 Distribution of somatic and germ-line transposition across the strawberry pseudochromosomes. The positions of the three most prolific launch pads (LP7, LP20 and LP25b) are indicated on the circumference. The sites (determined by the closest marker to a scaffold position) of germ-line transposition of lines derived from LP7, LP20 and LP25b are indicated by red, green and purple lines, respectively. Where more than one transposant occurred linked to the same marker locus, the line shows two colours. Somatic transpositions identified by the presence of an empty donor site in T0 transgenics are indicted by blue lines. Other T1 transposants identified in families with low-frequency transposition are indicated by pink lines.

In addition to the germ-line transpositions, we sequenced hiTAIL-PCR products of 17 somatic transposants where the FDS was absent per multiplex screening of primary transgenics. These transposon-tagged lines would not be expected to transpose again unless crossed with a plant expressing transposase. Hence, they are considered to be stable transposants. The insertion site of the somatic transposants included five in introns, one in an exon, four in promoters, one in the 3′ UTR and six intergenic, one of which could not be assigned to a unique scaffold because of multiple hits with similar e-values to several scaffolds. The somatic transpositions were distributed on six of the seven linkage groups (Figure 4).

Ac/Ds footprint

The EDS band obtained in multiplex PCR of several different T1 transposants was sequenced to observe the patterns of Ds excision. All of the sequences involved at least one base substitution, mostly C→G conversions. Deletions and insertions were also observed (Figure 5).

Figure 5.

 Various footprints after Ds excision showing the aligned sequences flanking the Ds excision site from six families derived from Fragaria vesca launch pads. PCR was performed by using primers anchored in transposase (TPaseR1) and EGFP (EGFPR1) and sequenced from both ends. Deletions are indicated by dashes, insertions and substitutions are in black.

Zygosity testing

For five transposants, T2 seedlings were sampled to screen for homozygous Ds insertions by zygosity testing. Primers were designed to amplify a product that either spanned the insertion site (wild-type allele) or amplified from the Ds T-DNA to the flanking region on one side of the insertion (mutant allele). In four of five cases, one to three homozygous T2 transposants were identified (Table 3, Figure 6). No homozygous insertions were observed among seedlings screened for Fv10-AcDs-LP20-TP1, possibly due to lethality of the knockout in the homozygote. The amino acid sequence predicted for hybrid gene model 27209 in which this insertion occurred in intron 2 had no highly similar hits in the GenBank.

Table 3.   Zygosity testing of T2 transposant lines. In five different unique transposant families, 5–10 T2 plants were screened in a three-primer reaction to attempt to identify homozygous transposon-tagged lines expected to occur at a 25% frequency from single-insertion transposants
Transposant T 2 plants screenedHomozygous transposants identified
Fv10-AcDs-LP7-TP17103
Fv10-AcDs-LP7-TP1851
Fv10-AcDs-LP20-TP1100
Fv10-AcDs-LP20-TP771
Fv10-AcDs-LP25b-TP1272
Figure 6.

 Zygosity testing of ten T2 seedlings derived from a T1 transposon-tagged line of Fragaria vesca LP7. Primers developed from the insertion site flanking the T-DNA amplified a 500-bp product in wild-type Fv10, the T1 hemizygote and seven of ten T2 lines. A third primer from within the T-DNA in conjunction with one of the flanking primers amplified a 190-bp product in the hemizygous T1 and nine of ten T2 lines. Seedlings in lanes 4, 8 and 10 were selected as homozygous transposon-tagged lines. Lanes 1 and 14 are HyperLadder™ IV.

Database

Information on the transposants is presented in an on-line database at http://hortmutants.vbi.vt.edu/HortMutants/strawberry.html. Transposants (n = 103) are grouped by the three most prolific launch pads (LP20, LP7 and LP25b), with somatic transpositions and those from several low-frequency launch pads forming two additional heterogeneous groups. For each line, the TAIL-PCR sequence is presented along with the best BLAST hit to the F. vesca assembly (version 8) as evidence of the region affected. For some transformants, multiple insertion sites are supported. The closest gene, or in some cases, more than one gene, is presented along with annotation of the nearest Arabidopsis homolog. Arabidopsis gene annotation can be browsed to find nearby Fragaria insertions. The coordinates of the TAIL-PCR sequence matched to the genome assembly have been used to link to the Fragaria Gbrowse website (https://strawberry.plantandfood.co.nz/cgi-bin/gbrowse/strawberry_genome_assembly_version_8) and to create a temporary feature to visualize the location of insertion.

Discussion

Certain crucial elements are required for the practical utilization of an efficient transposon tagging system to generate unlimited numbers of transposon-tagged lines with the Ds element distributed throughout the genome by germ-line transposition in the T0 generation of a model species that has been transformed with an AcDs construct. These include prevention of premature somatic transposition of Ds, frequent gametophytic transposition, facile identification of putative transposants using selectable markers, an efficient protocol to identify genomic sites adjacent to transposed Ds, abundant seed set in T0 and T1 plants, good germination rates in both transformed and transposon-tagged lines, highly homozygous plant material to prevent confounding phenotypes because of genetic segregation with those because of transposon tagging, and global transposition of Ds from the original launch pad integration site unless a more targeted insertional mutagenesis system is desired (Creff et al., 2006). In addition, the Ac and Ds elements most commonly isolated from maize need to function properly in alien germplasm, given suitable promoters. We have demonstrated several of these functions using an AcDs construct in F. vesca, although our system falls short of efficient in several respects. Our Ds element carried both P35S to drive transposase and Pmas to drive EGFP prior to Ds excision (Figure 2). Ds also carried kanamycin resistance. Selection of intact launch pads by GFP fluorescence and transposants by the absence of GFP expression coupled with kanamycin resistance was the intended scheme. However, GFP expression was weak or nonexistent in most of our lines, even though an intact FDS and full-length EGFP gene could be demonstrated by sequencing PCR products that spanned Ac and Ds. One family with brilliant GFP expression (Figure 3) was later found to have a recombined construct, that is, with P35S driving EGFP. This suggests that Pmas with the intervening inverted repeat was inefficient in driving EGFP expression or that the distance between Pmas and EGFP was too great in most putative launch pads. After developing a highly efficient DNA isolation protocol using a GenoGrinder (BT & C, Inc., Lebanon, NJ), we relied on multiplex PCR to identify putative transposants among more than 2400 T1 plants representing 32 putative launch pads. Repairing the promoter/IR/EGFP construct for reliable expression of the reporter gene should be a reasonably simple cloning project that is currently underway in our laboratory.

P35S was often too efficient in driving transposase, with the result that many of our putative launch pads transposed somatically prior to gametogenesis, resulting in several T1 transposant seedlings from the same launch pad bearing identical Ds insertion sites. We determined the identity by similarity of hiTAIL-PCR products, either by observing equivalent band sizes of primary and secondary hiTAIL-PCR reactions in multiple samples or by aligning sequences obtained from hiTAIL-PCR products. In some cases, in families where a single Ds insertion site predominated among putative transposants, we designed primers to amplify a product spanning the common insertion and the adjacent flanking strawberry sequence, in order to eliminate duplicate T1 insertions without investing the time and expense of hiTAIL-PCR.

On examination of the T-DNA insertion site of LP7, which was the most efficient in generating unique transposants, by BLAST searching the flanking region from the TPase hiTAIL-PCR product against the F. vesca genome, we deduced that the T-DNA was situated in the 3′ UTR of hybrid gene model 22118. The Bioview link (The New Zealand Institute for Plant and Food Research Ltd., Auckland) in the Strawberry Sequence Database predicted that hybrid gene model 22118 (DB accession number 1531225) most closely resembled At1G35910.1 (putative trehalose-6-phosphate phosphatase; e-value = 1e−133). The greatest expression of this Arabidopsis ortholog occurred in developing anthers and surrounding floral tissue, with little or no expression elsewhere [The Arabidopsis eFP Browser (http://www.bar.utoronto.ca/efp/cgi-bin/efpWeb.cgi)]. The TPase hiTAIL-PCR product was oriented in the same direction as the gene model, which should have placed the entire AcDs construct 91 bp after exon 12, with the EGFP end of the T-DNA closest to the native gene and TPase with P35S some 5 kb downstream. We can speculate that the native promoter driving the expression of gene22118 specific to male gametophytic tissue influenced transcription past exon 12 into the T-DNA limiting TPase expression to gametophytic tissue. An examination of the 3′ UTR region of gene22118 using PLACE (Higo et al., 1999) revealed that both of the nearest polyA signal motifs would have been separated from the gene by T-DNA insertion, reinforcing the suggestion of continued transcription into the T-DNA. The successful generation of unique transposants by LP7, by its fortuitous occurrence in a gametophyte-specific gene, supports the utilization of gametophyte- or meiosis-specific promoters for driving TPase expression for germ-line transposon tagging systems (Van Ex et al., 2009). The fact that the amino terminally deleted TPase gene with introns used in our construct easily mediated transposition in many of our strawberry transgenics confirmed the suitability of using this maize element in strawberry. The isolation of transposon-tagged lines in the present study also confirmed that inverted repeats were recognized by transposase in strawberry and that they could excise and reinsert efficiently.

Fragaria vesca sets fruit freely through natural self-pollination. However, the abundance of tiny anthers, a requirement for pollen collection in excised flower buds isolated prior to anthesis and cross-hybridization with flower tagging, although possible, make crosses tedious and time-consuming, with often variable efficiency of success and low seed set. This predisposes the isolation of T1 transposon-tagged lines in strawberry to the products of self-pollination of an AcDs launch pad rather than from crosses between independent Ac and Ds lines or between AcDs and wild type. For practical application of this system, however, transposition would need to be exclusively gametophytic in order to eliminate the inefficiency of characterizing multiple plants with the same tag. The expected 3 : 1 segregation of the Ac element from a hemizygous T0 transgenic launch pad limits the isolation of Ds-only transposants to the 25% wild-type plants that would have received the reintegrated Ds through meiotic segregation. The presence of an EDS in transposants should pose no problem as transposase activity is expected to be eliminated by excision of P35S from TPase. Indeed, most of our identified transposants were characterized as EDS/NPTII rather than simply NPTII. The EDS and the NPTII (Ds) should segregate in the T2 generation after self-pollination of T1, provided the absence of tight linkage between them or that EDS was not homozygous owing to its potential segregation during both male and female gametogenesis of our T0 lines.

The global transposition of Ds observed among our T1 transposon-tagged lines is a distinct advantage over systems where mostly local transposition has been described as with poplar (Fladung, 2011). However, an efficient system depends heavily on reliable selection to eliminate FDS plants and isolate Ds plants. Kanamycin screening by germinating seeds in liquid medium with 50 mg/L kan was reasonably efficient with only 143 wild-type T1 observed among 2457 T1 screened. Some 509 would have been expected after subtracting the 82 NPTII only from the 614 wild type expected by 3 : 1 segregation. The system became more efficient as the liquid medium allowed operators to recognize symptoms of kanamycin damage to seedlings at an early stage.

Transposon tagging in diploid strawberry appears to be a more promising method to develop an insertional mutant population than T-DNA insertional mutagenesis, which requires an independent transformation event for each mutant (Ruiz-Rojas et al., 2010). We have demonstrated the function of many of the required elements in the present study and hope to meet the remaining challenges through the development of new constructs using gametophyte-specific promoters and more reliable selectable markers.

Experimental procedures

Plasmid construction

PskI074 was used to construct pAc-DsN-EG with right border, the amino terminally deleted maize Ac transposase derivative (103–807) (Künze et al., 1995), originally recovered from maize wx-m7 allele (Müller-Neumann et al. 1984) (nucleotides 1574–3903 of AF332955), inverted repeat (nucleotides 2–236 of X05424), P35S to drive TPase, NPTII with nopaline synthase promoter (Pnos) and nopaline synthase terminator (Tnos), Pmas to drive EGFP (GenBank: HQ416708.1) with an intervening inverted repeat, Tg7 terminator following EGFP, then the left border (Figure 2). The pAcDsNEn-EG construct was derived from pAD100 (Suzuki et al. 2001). pAD100 was digested by SacI, dividing into pAcSPTV-Kan and DsAT elements, both of which were circularized by self-ligation (Ac-2 and Ds-2, respectively). The Specr-NPTII cassette in pAcSPTV-Kan was replaced by MCS fragment of pBluescript II SK- by SacI-BanHI double digestion and ligation (Ac-3). The Pnos-NPTII element from Ac-2 and the Tnos from pBI121 were cloned in pEGFP vector, producing the Pnos-NPTII-Tnos cassette, which was then inserted in pEGFP-N1 (p NPTII-EG). The En4 enhancer in Ds-2 was excised by HindIII digestion, and Ds-5 was formed by self-ligation of the remaining Ds-2. P35S in Ds-5 was replaced by Pmas from pSKI074, resulting in Ds-8. The IR-35S cassette from Ds-5 and the Pmas-IR cassettes from Ds-8 were inserted in front of Pnos and between Tnos and EGFP gene of pNPTII -EG (Ds-11). The whole fragment was subcloned in pBluescriptII SK- by NotI+SacI digestion and ligation (pBSDsN-EG) and inserted in Ac-3, forming the pA-DsN-EG vector. Transformed plants with an intact launch pad would be expected to express GFP and kanamycin resistance. Identity of transformants was based on positive selection for both GFP and kanamycin resistance during callus growth. Transposition of Ds would be expected to shut down both transposase and GFP, allowing the recovery of transposants by negative GFP selection and positive kanamycin selection.

Transformation

Seedlings of F. vesca PI 551572 were grown in a growth chamber under an 11-h photoperiod with a light intensity of 500 μm/m2/s at 22 °C (day)/16 °C (night). Unopened trifoliate leaves were used for plant transformation according to Oosumi et al. (2006) and the modified transformation method (Pantazis et al., 2012 submitted). When possible, two shoots were taken from each GFP-positive callus. Rooted T0 plants were acclimated to the glasshouse where fruit from several inflorescences was collected and T1 seeds were extracted and batched into a common sample for each T0 line.

DNA extraction

DNA was extracted from the leaves of T0 and T1 plants using a modified CTAB protocol (Doyle and Doyle, 1990; Lodhi et al., 1994; Oosumi et al., 2006; Porebski et al., 1997). Single folded trifoliate leaves were harvested from each plant and placed in a 1.5-mL centrifuge tube containing a single #2 steel shot from Ballistic Products, Inc (Corcoran, MN). The samples were frozen with liquid nitrogen and then ground using the Geno/Grinder 2000 (BT & C, Inc., Lebanon, NJ). An extraction buffer containing polyvinylpyrrolidone was vortexed and heated to 65 °C for 30 min; 500 μL extraction buffer was added to each sample and incubated at 65 °C for 30–60 min. DNA solution was extracted with one volume of chloroform/isoamyl alcohol (24 : 1) and centrifuged at 18 000 g for 5 min. DNA was then precipitated with two volumes of chilled 100% ethanol and 0.5 volumes of chilled 5 m NaCl and centrifuged at 21 000 g for 10 min. The DNA precipitate was rinsed with 500 μL of chilled 70% ethanol and centrifuged at 21 000 g for 5 min. The remaining DNA pellet was dissolved in 100 μL of 20 μg/mL TE/RNase solution. DNA concentration was determined using the NanoDrop ND-1000 Spectrophotometer (Thermo Scientific, Wilmington, DE).

Multiplex PCR

A multiplex PCR protocol was developed to screen putative transgenic T0 launch pads and transposant T1 strawberry plants. The PCR comprised a total of four primer pairs, one for the amplification of a FDS spanning the Ac/Ds junction between transposase on Ac and P35S on Ds (600-bp expected product size), one for NPTII to demonstrate the presence of the Ds element (300-bp expected product size), one for an EDS on both components of the Ac element on either side of Ds (1-kb expected product size; no product expected if the AcDs element was intact) and one for one FvAAT (Fragaria vesca alcohol acyltransferase: GenBank accession number AF193790) to distinguish failed reactions from nontransgenic escapes (179-bp expected product size). The primers are listed in Table S1, and the positions of the primers are shown in Figure 2. The reaction mix contained 2 μL strawberry DNA from a 50 ng/μL stock, 12.5 μL Immomix™ (Bioline, Taunton, MA), 0.5 μL 50 mm MgCl, 0.5 μL of each primer (10 μm), with water added to a total volume of 25 μL. The thermocycler programme consisted of one cycle (95 °C for 10 min), 35 cycles (94 °C for 30 s, 58 °C for 30 s, 72 °C for 1 min), and one cycle (72°C for 5 min). PCR products were run on a 1% agarose/1× TAE gel with 0.1% EtBr and visualized with a Bio-Rad (Hercules, CA) Universal Hood II ChemiDoc gel documentation system.

Southern blots

Southern blots were carried out on HindIII-digested genomic DNA using either the Amersham AlkPhos Direct™ Labeling and Detection Systems or else autoradiographed after hybridization with a random primed FDS probe.

hiTAIL-PCR

The hiTAIL-PCR protocol described by Liu and Chen (2007) was modified to accommodate the structure of the Ac/Ds T-DNA. Three different hiTAIL protocols were used: (i) TPase-hiTAIL to amplify from the transposase gene in the Ac element through the right border of the T-DNA and into the strawberry genome in order to locate the genomic position of the putative launch pads; (ii) P35S-hiTAIL to amplify from the P35S region of the Ds element through the inverted repeat at the TPase end of the Ds element and into the strawberry genome to determine the insertion site of putative transposants lacking a FDS; and (iii) IR-hiTAIL to amplify from the inverted repeat (IR) on the opposite side (towards EGFP) of the Ds element outward into the strawberry genome to situate the location of transposed Ds elements if P35S-hiTAIL failed. The two IRs on opposite ends of the Ds element differed sufficiently so that primers could be developed specifically for IR-hiTAIL on the left border of Ds without annealing to the IR on the right border. All of the nested hiTAIL primers (Table S1) were used in conjunction with long arbitrary degenerate primers (Liu and Chen, 2007), either singly or in combinations of two.

Evaluation of putative launch pads

Putative launch pads with an intact FDS and a single T-DNA insertion as determined by Southern blot were evaluated for their ability to launch the Ds element. T1 seed was collected from T0 plants and bulked from different fruit and inflorescences. Seedlings were germinated in liquid B5 medium with 50 mg/L kanamycin (kan) in an illuminated (1200 W) incubator shaker (74 rpm). After 5 weeks, kan-resistant green seedlings with well-developed root systems were transplanted to cell packs in a growth chamber. After approximately 4 weeks, DNA was extracted from up to 24 of the selfed progenies from each putative launch pad. Each T1 seedling was subjected to multiplex PCR and classified according to the following: FDS/NPTII (not transposed), FDS/EDS/NPTII (possibly transposed but confounded with untransposed), EDS (evidence of transposition but without the Ds element—these would be stable insertional mutants at the genomic position of the launch pad), EDS/NPTII (transposed with a dysfunctional Ac devoid of promoters for transposase), NPTII (transposed), wild type. Putative transposant seedlings (EDS/NPTII and NPTII) were transplanted to the glasshouse for flowering and seed set. Either IR-hiTAIL or P35S-hiTAIL was conducted on putative T1 transposants to determine the genomic position of the Ds element. Bands from hiTAIL-PCR were cut from the gels and purified using the QIAquick® PCR Purification Kit from Qiagen (Qiagen Inc., Valencia, CA) following the manufacturer’s protocol, then sequenced at the Tufts University Core Facility (Tufts Medical School, Boston, MA). If there was evidence of active germ-line transposition, additional seedlings were screened from the verified launch pad.

Genomic position of launch pads and transposants

Nucleotide sequences obtained from hiTAIL-PCR products were trimmed of the T-DNA sequence and then BLASTed against the strawberry genome assembly (Shulaev et al., 2011). BLAST results that gave a single scaffold hit with a low e-value were observed in the strawberry genome browser (Shulaev et al., 2011) to determine the insertion site relative to hybrid gene models. Products obtained from hiTAIL-PCR of all the T1 seedlings derived from the same launch pad were aligned using MegAlign (DNAStar Lasergene 8, DNASTAR, Inc., Madison, WI) to identify seedlings with similar insertion sites resulting from somatic transposition prior to gametogenesis in the T0 plant.

Zygosity testing for homozygous transposants

Once unique T1 transposants had been identified, we planted up to 13 T2 seedlings for zygosity testing to isolate homozygous mutant lines. The region of the scaffold that yielded the best BLAST hit to a transposant hiTAIL-PCR sequence was downloaded from the strawberry genome website; the expected position of the Ds insertion noted and primers that would amplify a region spanning the insertion site were developed using either VectorNTI (Life Technologies, Grand Island, NY) or DNASTAR Lasergene 8 Primer Design feature within SeqBuilder. Care was taken to ensure that the two expected PCR products would be sufficiently different in size to be easily differentiated. PCR was conducted on DNA from each of the first five T2 lines derived from the same T1Ds line in a three-primer reaction, the two flanking primers and one outward-facing primer from the Ds element that would yield a different size product than the wild type. DNA from the corresponding T1 and wild-type plants were used as controls for mutant and wild-type alleles, respectively. If an individual T2 plant was identified that had only the mutant band without the expected wild-type product, it was selected as a homozygous insertion; if no homozygous insertions were found among the first five plants, then we sampled additional plants from the same family. In families where we expected somatic transposition prior to gametogenesis because of the detection of the same insertion site in multiple T1 seedlings within the family, we used zygosity testing primers from a common site on uncharacterized putative transposants to identify additional T1 seedlings bearing the same somatic transposition, thus saving the expense and effort of hiTAIL-PCR. Occasionally, if the three-primer duplex zygosity test reaction failed, the mutant and wild-type products could be resolved in simplex PCR.

Insertion site

The position of each launch pad and transposition site (intron/exon/promoter/intergenic/3′ UTR) was determined relative to hybrid gene models on the strawberry genome browser (https://strawberry.plantandfood.co.nz/cgi-bin/gbrowse/strawberry_genome/#search). Promoters were called if the insertion occurred with 2 kb of the first exon of a gene model; 3′ UTR was called if the insertion occurred within 1 kb after the last exon of a gene model.

Database

Data were entered into a MySQL database and served dynamically by Java servlets running in an Apache Tomcat web server.

Acknowledgements

This work was supported by USDA/NRI Award No. 2008-35300-04458, USDA/NIFA Award Nos. VAW-2010-01562 and VAW-2009-04069, The Virginia Tobacco Indemnification Program, Virginia Agricultural Experiment Station Hatch Project 135816 and the Virginia Tech ASPIRES Program. The contribution of Jared D. Carter for generation of Figure 4 is gratefully acknowledged. The authors of this manuscript have no conflict of interest to declare with Wiley-Blackwell.

Ancillary

Advertisement