Trees

, Volume 25, Issue 3, pp 551–557

Analysis of re-integrated Ac element positions in the genome of Populus provides a basis for Ac/Ds-transposon activation tagging in trees

Authors

    • Johann Heinrich von Thuenen-Institute, Federal Research Institute for Rural Areas, Forestry and FisheriesInstitute of Forest Genetics
Short Communication

DOI: 10.1007/s00468-010-0511-0

Cite this article as:
Fladung, M. Trees (2011) 25: 551. doi:10.1007/s00468-010-0511-0

Abstract

With a view to establish an efficient gene tagging system for forest tree species, we assessed the transposition behaviour of the maize transposable element Ac in poplar. In earlier work, we showed that new integration sites were often located within predicted or known coding sequences. However, somatic transposition behaviour of Ac with regard to conserved chromosome specificity or, more specifically, whether Ac transposition is restricted to the chromosome on which the primary insertion locus (donor) is located or whether it is able to pass chromosomal boundaries, remained unclear. To answer these questions, we took advantage of the publicly available Populus trichocarpa genome sequence (Phytozome v5.0; http://www.phytozome.net) and three 35S::Ac-rolC transgenic hybrid aspen lines to determine the flanking sequences of Ac re-integration sites for tissue sectors from which Ac had been excised. Only about one-third of the analysed re-integrations were positioned within the scaffold containing the primary Ac donor locus, and the majority of re-integrations were found scattered over many unlinked sites on other scaffolds confirming that Ac transposition in poplar does in fact cross chromosome boundaries. The majority of re-integration sites (57.1%) were found within or near coding regions demonstrating that Ac/Ds transposon tagging in poplar holds much promise for the efficient induction of mutants and functional genomics studies in forest tree species.

Keywords

Functional genomicsPoplarTree genomicsTransgenic aspenTransposition

Introduction

Modern DNA sequencing technology promises to produce a large number of whole genome sequences within a short period of time. For forest tree species, the public release of whole genome sequences for Populus trichocarpa (Tuskan et al. 2006) and Eucalyptus grandis (http://eucalyptusdb.bi.up.ac.za/) represents a fundamental milestone in tree genomics research. These enormous resources provide vast new opportunities to broaden our understanding of growth and development of woody tree species. The challenge remains to combine the “letters” and “words” within the genomic sequence with a specific physiological, structural or developmental function. Unlike in Arabidopsis and other annual plant species, only very few mutant forms are known for forest tree species that could be used to analyse specific gene function behind the mutation (Flachowsky et al. 2009). Thus, tree mutagenesis combined with phenotypic analyses remains a key challenge for the identification of gene function in trees.

Strategies that have been successfully used in many annual plant species to induce genetic changes and to isolate mutants include genomic engineering approaches such as T-DNA and transposon tagging (both knock-out or gain-of-function [activation tagging]), promoter trapping as well as down-regulation of gene expression via antisense and RNAi technology (Koncz et al. 1989; Feldmann 1991; Martienssen 1998; Springer 2000; McGinnis 2010). Tagging approaches based on T-DNA insertion are effective in plant species that are easily transformed in tissue culture and yield high frequencies of tagged lines (Qu et al. 2008). Transposon-based tagging approaches mainly using the maize Ac/Ds element has also been successfully applied to generate insertional mutants in many plant species (Parinov et al. 1999; Meissner et al. 2000; Greco et al. 2004; McKenzie and Dale 2004).

In trees, only few examples exist of successful application of insertional mutagenesis, and these have been restricted to poplar (Busov et al. 2005, 2010). A dominant gibberellin catabolism gene (GA2-OXIDASE) was the first gene to be isolated in a pilot study that produced a T-DNA-based activation tagging population (Busov et al. 2003). Later, a large T-DNA activation tagging population was screened for developmental abnormalities including alterations in leaf and stem structure as well as overall stature (Harrison et al. 2007). While the application of a transposon-based tagging system for trees is still in progress, transfer, excision, and re-integration of the maize transposable element Ac have been reported for Populus (Fladung and Ahuja 1997; Kumar and Fladung 2003; Fladung et al. 2004). It was suggested that this transposon could become an important tool for the study of gene function because of its preferential re-integration in or near coding regions (Kumar and Fladung 2003).

In annual plants, the mobility of the Ac element (or its inactive derivate Ds) was mostly analysed via germinal excision, and new insertion sites were found scattered throughout the genome at many unlinked sites (McKenzie and Dale 2004; Kolesnik et al. 2004; Qu et al. 2008). However, insertion clusters were also identified around donor sites as well as near nucleolus organiser regions (Parinov et al. 1999; Raina et al. 2002), or new insertion sites showed a bias towards single chromosomes (Kolesnik et al. 2004). Whether the same holds true also for long-lived trees, where somatic excision of the transposable elements occurs, remains to be answered. Here, mobile elements might jump only within a preferential region or “hot spot” or within the chromosome where the primary insertion (donor) locus is located. Alternatively, a transposon might be able to jump to other chromosomes, thus passing chromosomal boundaries. If the latter was true, only a few primary transposon carrying transgenic lines would be sufficient for a large transposon tagging experiment to saturate a genome with transposon re-integration sites.

To answer these questions, we took advantage of the publicly available Populus trichocarpa genome sequence (Phytozome v5.0; http://www.phytozome.net) and three 35S::Ac-rolC transgenic Populus tremula × P. tremuloides hybrid aspen lines (Kumar and Fladung 2003) to determine the flanking sequences of Ac re-integration sites for tissue sectors from which Ac had been excised as a proof-of-concept for the suitability of this approach for large-scale transposon tagging in a forest tree species.

Experimental procedures

Generation and molecular characterisation of 35S-Ac-rolC transgenic poplar

Genetic transformation of poplar with the 35S-Ac-rolC construct, and phenotypic and molecular characterisation of the transgenic lines carrying the 35S-Ac-rolC construct have been described in detail in Kumar and Fladung (2003). Moreover, molecular evidence for Ac excision and re-integration of the element into the genome has also been given in Kumar and Fladung (2003).

In silico analysis of Ac flanking genomic sequences

The procedure to obtain Ac flanking genomic sequences has been described in Kumar and Fladung (2003). All sequences were submitted to the GabiPD database (http://www.gabipd.org/) and blasted against the publicly available genome sequence of Populus trichocarpa (Phytozome v5.0; http://www.phytozome.net).

Results and discussion

Our earlier research intended to use the maize transposable element Ac to establish a gene tagging system in poplar (Fladung and Ahuja 1997; Kumar and Fladung 2003). Towards this end, we have shown in several studies that Ac is functional in somatic tissues in Populus, and we were able to demonstrate Ac excision from the original donor locus (Fladung and Ahuja 1997) as well as re-integration into another locus (Kumar and Fladung 2003). In all cases, we successfully used the 35S::rolC reporter system (Fladung 1990; Fladung et al. 1997). Transgenic plants carrying 35S::rolC are characterised by dwarfed growth and pale leaf colour. Placing the Ac element between the 35S promoter and the rolC gene leads to inactivation of rolC, and excision of the element from this position re-activates the gene (Fladung and Ahuja 1997). Thus, somatic transposition can be easily followed visually by using the pale-green colour of leaf sectors as a guide (Fladung and Ahuja 1997).

From these sectors, DNA could be extracted, and flanking regions of Ac new insertion (re-integration) sites could be sequenced using inverse-PCR or TAIL-PCR (Kumar and Fladung 2003). Prior to the public release of the completed poplar genome sequence, our earlier work could “only” identify and analyse Ac-flanking sequences where those were located within DNA fragments that were represented in public databases. At that time, about one-third of the identified re-insertion sites showed similarity to already published plant gene sequences. Our conclusion was that the Ac element frequently inserts in or near coding regions, which is essential for the development of a transposon-based activation tagging system in poplar.

Unfortunately, important questions regarding somatic transposition behaviour of the Ac element remained unsolved. These questions comprise the chromosome specificity of the element in poplar: is Ac restricted to the chromosome on which the primary insertion locus (donor) is located? If this is not the case, is the element able to pass chromosomal boundaries?

Determination of Ac donor loci

Taking advantage of the publicly available genome sequence of Populus trichocarpa (Phytozome v5.0; http://www.phytozome.net), we first determined the original chromosomal position of the T-DNA-Ac element (donor locus) for the three independent 35S::Ac-rolC transgenic aspen-poplar lines used in our experiments (Fig. 1). High co-linearity between the genomes of P. trichocarpa and aspen-Populus (P. tremula and P. tremuloides) was recently demonstrated using a genetic mapping approach (Pakull et al. 2009) and allowed us to position the T-DNA-flanking sequences obtained from aspen by blast search against the corresponding sequence in the P. trichocarpa genome.
https://static-content.springer.com/image/art%3A10.1007%2Fs00468-010-0511-0/MediaObjects/468_2010_511_Fig1_HTML.gif
Fig. 1

Distribution of the Ac original donor site (black triangle on the right side in the upper scaffold) and new Ac insertion sites (arrows) in the genome of the three 35S-Ac-rolC transgenic Populus lines #2 (a), #3 (b), and #10 (c). Due to the high collinearity between the P. tremula × P. tremuloides and the P. trichocarpa genomes (Pakull et al. 2009), the position could be determined by blast search of the aspen sequence in the P. trichocarpa genome. The length of the scaffolds is not corresponding to their physical size

Ac donor positions for three transgenic aspen-poplar lines were determined on three different scaffolds, namely 17 (line 35S::Ac-rolC#2), 6 (line 35S::Ac-rolC#3), and 14 (line 35S::Ac-rolC#10), indicating insertion of the gene construct in line #3 in a far-end terminal position, and in the other two lines (#2 and #10) towards one end, but not in a completely terminal position. Significant alignments to known genes represented in the P. trichocarpa genome were found for genomic T-DNA flanking sequences for all three lines (line #2: E value 0, POPTR_0017s04950, conotoxin; line #3: E value 0, POPTR_0006s01440, protein of unknown function; line #10: E value 0, POPTR_0014s11980, no functional annotation).

Determination of new Ac insertion loci

In total, 79 Populus tremula × P. tremuloides genomic flanking sequences were obtained and analysed for new Ac insertion sites (Kumar and Fladung 2003). A search for homologous sequences in the genome of P. trichocarpa (Phytozome v5.0) found 70 of those to show high similarity to P. trichocarpa genomic sequences with an E value of at least −2 (only one case), with the majority of E values ranging from −30 to > −100, and nine sequences producing no hit or matching T-DNA sequences. Following somatic transposition of Ac, varying numbers of flanking genomic sequences were obtained for the three 35S::Ac-rolC transgenic aspen-poplar lines analysed showing significant alignment with P. trichocarpa genomic sequences ranging from 11 in line #2, 22 in line #10, and 37 in line #3. New genomic positions of transposed Ac are shown in Fig. 1 and summarised in Table 1. For most sequences, only one highly significant alignment to P. trichocarpa sequence (i.e. a very low E value) was obtained. In few cases, two ore more significant alignments with different E values were observed, possibly due to the duplication of the Populus genome described by Tuskan et al. (2006). In those cases, only the match with the lowest E value was considered.
Table 1

Number of Ac insertions in Populus scaffolds on basis of BLAST search results against the genome sequence of Populus trichocarpa (Phytozome v5.0; http://www.phytozome.net)

 

Number of scaffolds

Scaffolds

Transgenic line

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

>19

Sum

35S-Ac-rolC#2

1

   

1

2

2

   

1

 

1

 

1

 

2

  

0

11

35S-Ac-rolC#3

1

2

 

2

 

18

3

2

1

2

  

1

   

1

 

1

3

37

35S-Ac-rolC#10

5

1

1

  

2

 

1

1

1

  

1

5

1

 

2

  

1

22

Sum

7

3

1

2

1

22

5

3

2

3

1

0

3

5

2

0

5

0

1

4

70

Percentage

10.0

4.3

1.4

2.9

1.4

31.4

7.1

4.3

2.9

4.3

1.4

0

4.3

7.1

2.9

0

7.1

0

1.4

5.7

100.0

Size of scaffold (MB)

48.40

23.6

20.2

23.2

25.8

26.6

15.1

18.8

12.9

21.5

18.9

14.9

15.7

17.7

15.1

14.1

14.7

15.0

16.0

  

No. of insertions/100 Kbp

1.45

1.27

0.50

0.86

0.39

8.27

3.31

1.60

1.55

1.40

0.53

0

1.91

2.82

1.32

0

3.40

0

0.63

  

The scaffold containing the donor locus of Ac is grey-shaded . (http://genome.jgi-psf.org/help/scaffolds.html: a scaffold is a portion of the genome sequence reconstructed from end-sequenced whole-genome shotgun clones. Scaffolds are composed of contigs and gaps.)

Re-integration of Ac was confirmed not only in the original but also in other scaffolds for all three lines. Re-integration within the same scaffold was most frequent, ranging from 18% in line #2 to 48% in line #3. For 35S::Ac-rolC#2, a total of 11 re-insertion sites were found, with 2 detected in the original scaffold 17 and 9 in 7 other scaffolds. Line 35S::Ac-rolC#3 produced the highest number (37) of analysed re-integrations: 18 re-integrations were found in the original scaffold 6, and 19 insertions were detected in 13 other scaffolds. Finally, for 35S::Ac-rolC#10, for 22 analysed Ac-flanking sequences, five re-integrations were found on the original scaffold 14 and 17 in 11 other scaffolds (Fig. 1).

These results clearly demonstrate that Ac is able to cross linkage barriers (scaffolds or chromosomes). Overall, no preference to any scaffold was detectable, and the frequency of Ac re-integration in the different scaffolds ranged from 1.4% (1 re-integration) to 10% (7 re-integrations) of all analysed insertions, not considering the scaffolds carrying the original construct (Table 1). Taking into account the size of the different scaffolds, with scaf_1 being the largest, the frequency of Ac re-integrations ranged from 0.50 (scaf_3) to 3.31 (scaf_7) insertions per 100 kb, again not considering the scaffolds carrying the original construct (Table 1). Only for three scaffolds (12, 16, and 18), no Ac re-integration was observed, and four flanking sequences (5.6%) were found in scaffolds not related to the 19 scaffolds representing the haploid Populus genome (Table 1).

For all three lines, a high number of re-integration sites was detected in other scaffolds than the one carrying the original Ac-containing T-DNA (donor) locus. Thus, our results show that in poplar new Ac insertion sites can be scattered over many unlinked sites which is in line with earlier results for Brassica (McKenzie and Dale 2004). Unlike in rice (Kolesnik et al. 2004) and Arabidopsis (Parinov et al. 1999; Raina et al. 2002), no bias towards any chromosome was observed for Populus (with the exception of scaf_6 carrying the Ac donor locus in line #3).

Annotation of coding regions

BLAST search results for re-integration sites against the genome sequence of Populus trichocarpa (Phytozome v5.0) also indicated frequent Ac integration in or near coding regions (Table 2). A total of 40 (57.1%) new integration positions were found near genes, with 8 (72.7%) found for line #2, 22 (59.5%) for line #3, and 10 (45.5%) for line #10. These results are consistent with earlier findings in Arabidopsis (Parinov et al. 1999), tomato (Meissner et al. 2000), rice (Greco et al. 2001, 2004), and barley (Cooper et al.2004); however, the overall frequency of re-integration into coding sequence in Populus appears to be somewhat lower. Our results clearly demonstrate the suitability of Ac for the establishment of a highly efficient transposon tagging system in Populus both for either knock-out or gain-of-function (activation tagging) approaches. Using a DsAT transposable element as developed by Suzuki et al. (2001), the establishment of such an Ac/Ds activation tagging population for poplar is currently underway in our laboratory.
Table 2

Summary of Ac insertion sites in poplar based on BLAST search results against the genome sequence of Populus trichocarpa (Phytozome v5.0; http://www.phytozome.net)

Scaffold/position

POPTR-affiliation

Gene name

Score

E value

35S-Ac-rolC#2

 1/7082342

_0001s09250

Plant phoshoribosyltransferase C-terminal

576.6

3.7E−163

 6/2001236

_0006s03170

Cysteine protease inhibitor activity

188.8

1.4E−46

 6/5031628

_0006s06980

Mevalonate kinase/galactokinase

252.0

1.5E−65

 7/6623312

_0007s08160

Serine-threonine protein kinase

679.4

0

 11/18442817

_0011s16640

O-linked acetylglucosamine transferase

129.3

1.8E−28

 13/3924399

_0013s05490

Ger1 protein (RNA binding region)

567.5

1.9E−160

 17/4528678

_0017s05810

Transferase activity

701.0

0

 17/6858287

_0017s08340

Histone deacetylase (HDAC) interacting

410.7

2.3E−113

35S-Ac-rolC#3

 2/3955418

_0002s05970

1,3 beta-glucan synthase

224.9

9.5E−58

 2/11896604

_0002s15960

Domain of unknown function (DUF296)

241.1

4.6E−62

 4/17085661

_0004s16690

Beta-galactosidase

203.3

4.9E−51

 6/937464

_0006s01550

Apoptotic ATPase

807.4

0

 6/2307963

_0006s03540

Membrane transport protein

399.8

4.5E−110

 6/2702230

_0006s03990

Condensin

29.2

2.7E−2a

 6/3424046

_0006s04900

Sterol C methyltransferase

367.4

2.9E−100

 6/3917465

_0006s05540

No functional annotation

132.9

2.5E−30

 6/4368332

_0006s06150

Akryrin repeat-containing

208.7

6.9E−53

 6/4385385

_0006s06170

ANK repeat-containing

452.1

1.9E−125

 6/4507644

_0006s06310

Exostosin-related

419.7

5.9E−116

 6/5121061

_0006s07130

Predicted membrane protein

262.8

5.0E−69

 6/10856778

_0006s13640

Saccharopine dehydrogenase

208.7

6.9E−53

 7/224267

_0007s00570

Zinc-binding dehydrogenase

363.8

3.3E−99

 8/9033516

_0008s13270

Sensory transduction histidine kinase

993.1

0

 9/10455248

_0009s12940

No functional annotation

749.7

0

 10/17088571

_0010s19070

Serine-threonine protein kinase

906.6

0

 10/21066427

_0010s25350

SWIM zinc finger

672.1

0

 17/6857961

_0017s08340

Histone deacetylase complex

805.6

0

 19/14134438

_0019s13170

Domain of unknown function (DUF588)

560.3

3.8E−158

 113/25287

_0113s00240

Serine-threonine protein kinase

66.2

1.1E−9

 344/2156

_0344s00200

Serine-threonine protein kinase

403.4

3.3E−111

35S-Ac-rolC#10

 1/9045632

_0001s11630

Cation-transporting ATPase

201.5

8.6E−57

 1/9686994

_0001s12470

Glutamine amidotransferase class I

776.7

0

 1/26034816

_0001s27090

Sterol methyl-transferase C-terminal

946.3

0

 1/30716097

_0001s32360

Histone deacetylase complex

288.0

1.3E−76

 6/7520075

_0006s10110

Isoamyl acetate hydrolysing esterase

199.7

3.2E−50

 6/12420785

_0006s14940

Linker histone H1 and H5 family

428.7

8.8E−119

 9/11528978

_0009s14810

Root cap

242.9

3.8E−63

 10/15597957

_0010s16740

Glycosyl hydrolase family 1

73.4

2.8E−12

 14/6354296

_0014s08550

Copper transport protein ATOX1-related

237.5

2.0E−61

 17/7847598

_0017s09210

Serine-threonine protein kinase

224.9

3.4E−57

Information about scaffold number and position of the first nucleotide of the BLAST match, locus name in P. trichocarpa (POPTR-affiliation), gene name, score, and E value are given

aThe best match from all searches is shown, as expressed with the lowest E value. Values larger than e−5 are not considered to be significant

Suitability of the proof-of-concept approach for large-scale transposon tagging in poplar

Application of this proof-of-concept approach to large-scale transposon tagging in poplar is hindered to the fact that no germinal excision but only somatic excision of the element can be induced. Thus, following induction, not all cells may exhibit transposon excision. One solution would be to use the similar system as it has been successfully applied with recombination systems to induce site-specific integration of transgenes (Fladung et al. 2010). Here, regenerative calli or plant tissue were thoroughly crushed into micro-calli as small as possible prior the heat-shock experiments. From these heat-shocked cells/micro-calli, plants could be regenerated and screened for putative variations. However, also by using this procedure, production of chimeras cannot per se be excluded.

Acknowledgments

I thank O. Polak, D. Ebbinghaus and A. Schellhorn for skilful technical assistance. This work was supported by Ministry of Education and Research (BMBF) funded under ‘Genome Analysis in the Biological System Plant’ (GABI; project GABI-POP) and the German Research Foundation (projects FL263/12-2 and FL263/12-3). In particular, I thank Prof. Dr. Gerd Bossinger (University of Melbourne, Australia) for critically reading the manuscript and for his very helpful corrections.

Copyright information

© Springer-Verlag 2010