Precise transgene stacking in planta through the combined use of TALENs and unidirectional site-specific recombination systems

Introduction This article surveys methodologies for gene stacking in plants and proposes a new strategy for precise transgene stacking by combining transcription activator-like effector nucleases and uni-directional site-specific recombination systems to precisely stack transgenes at a specific locus in the plant genome. Site-specific stacking would allow all inserted transgenes to co-segregate in progeny and also allow for their efficient removal as a unit, if necessary. Such a method could also target transgenes to a ‘safe spot’ in a plant genome that have decreased potential for transgene introgression and escape into wild relatives. Conclusion Site-specific recombination sitespecific recombination systems have been widely used in plant genetic transformation for the purpose of genome manipulation, including gene deletion, integration and stacking. The recent discovery of transcription activator-like effector nucleases (a type of engineered chimeric nucleases) provides a powerful molecular tool for gene targeting in a variety of cell types. The combined use of transcription activator-like effector nucleases and site-specific recombination can direct the site-specific recombination-mediated transgene stacking at a desirable genomic locus for more precise gene stacking. Introduction As plant biotechnology continues to develop, we can expect that transgenic crops will include increasingly greater amounts of inserted DNA compared to the singleor double-trait crops currently commercialised; e.g. those with insect-resistance and herbicidetolerance. Even within a trait, say, for insect resistance there is a trend to stack multiple insect resistance genes to combat insect resistance to single toxins. One of the examples is the ‘Yield-Guard VT’ triple transgenic maize hybrids1. Beyond that, metabolic engineering of plants might be enabled from technology allowing gene stacking into one specific locus to obviate subsequent breeding2. The development of transgenic crops with multiple traits is rapidly gaining popularity. According to ISAAA 2012 report, biotech crops with two or more stacked traits accounted for ~25% of the 170 million hectares of cultivated transgenic crops3. However, the dispersal of transgenes in the genomes of the transgenic lines by using current methods of gene transfer, such as Agrobacterium or particle bombardment (Biolistic®)mediated transformation can cause difficulty in the post-transgenesis management of these new genes. For instance not all transgenes can be maintained in the individual plants of progeny after genetic segregation during breeding. The dispersal of transgenes also makes their removal difficult, if it is needed. Random insertion also carries the risk of knocking out endogenous essential genes, which might cause harm or reduce the fitness of the host plant. To avoid these disadvantages, having the ability to direct transgenes to a pre-characterised locus as a unit is desirable. Gene stacking at the same locus allows the transgenic trait genes to be clustered and co-segregated within the progeny and removed efficiently as a unit, if necessary. Theoretically, this process will also make the expression of these stacked transgenes more predictable. In this article, the current methods used for stacking transgenes will be discussed and the strategy of combined-use of transcription activator-like effector nucleases (TALENs) and unidirectional site-specific recombination (SSR) systems to precisely stack transgenes at a specific locus in a plant genome is proposed. Discussion The authors have referenced some of their own studies in this review. The protocols of these studies have been approved by the relevant ethics committees related to the institution in which they were performed. Current strategies for stacking transgenes in plants Several methods have been used to stack (or pyramid) transgenes into plants that range from using conventional sexual crosses of plants containing one or more of transgenes to be combined into a single germplasm to biotechnologies. We will briefly survey these methods. Conventional breeding In several deregulated commercial transgenic plants, conventional breeding has been used to stack * Corresponding author Email: yau@nsuok.edu 1 Department of Natural Sciences, Northeastern State University, Broken Arrow, OK 74014, USA 2 Department of Plant Sciences, University of Tennessee, Knoxville, TN 37996, USA


Introduction
As plant biotechnology continues to develop, we can expect that transgenic crops will include increasingly greater amounts of inserted DNA compared to the single-or double-trait crops currently commercialised; e.g.those with insect-resistance and herbicidetolerance.Even within a trait, say, for insect resistance there is a trend to stack multiple insect resistance genes to combat insect resistance to single toxins.One of the examples is the 'Yield-Guard VT' triple transgenic maize hybrids 1 .Beyond that, metabolic engineering of plants might be enabled from technology allowing gene stacking into one specific locus to obviate subsequent breeding 2 .The development of transgenic crops with multiple traits is rapidly gaining popularity.According to ISAAA 2012 report, biotech crops with two or more stacked traits accounted for ~25% of the 170 million hectares of cultivated transgenic crops 3 .
However, the dispersal of transgenes in the genomes of the transgenic lines by using current methods of gene transfer, such as Agrobacterium or particle bombardment (Biolistic ® )mediated transformation can cause difficulty in the post-transgenesis management of these new genes.For instance not all transgenes can be maintained in the individual plants of progeny after genetic segregation during breeding.The dispersal of transgenes also makes their removal difficult, if it is needed.Random insertion also carries the risk of knocking out endogenous essential genes, which might cause harm or reduce the fitness of the host plant.To avoid these disadvantages, having the ability to direct transgenes to a pre-characterised locus as a unit is desirable.Gene stacking at the same locus allows the transgenic trait genes to be clustered and co-segregated within the progeny and removed efficiently as a unit, if necessary.Theoretically, this process will also make the expression of these stacked transgenes more predictable.In this article, the current methods used for stacking transgenes will be discussed and the strategy of combined-use of transcription activator-like effector nucleases (TALENs) and unidirectional site-specific recombination (SSR) systems to precisely stack transgenes at a specific locus in a plant genome is proposed.

Discussion
The authors have referenced some of their own studies in this review.The protocols of these studies have been approved by the relevant ethics committees related to the institution in which they were performed.

Current strategies for stacking transgenes in plants
Several methods have been used to stack (or pyramid) transgenes into plants that range from using conventional sexual crosses of plants containing one or more of transgenes to be combined into a single germplasm to biotechnologies.We will briefly survey these methods.Competing interests: none declared.Conflict of interests: none declared.

Conventional breeding
All authors contributed to conception and design, manuscript preparation, read and approved the final manuscript.All authors abide by the Association for Medical Ethics (AME) ethical rules of disclosure.
using the mini-chromosome platform for transgene stacking is not a simple matter for public sector researchers and it has not been widely adopted.

Consecutive transformation
Retransformation of a transgenic line can be used to stack transgenes into one plant line.For example, Singla-Pareek et al. 13 retransformed the gly-II gene into tobacco with gly-I transgenic background, which led to enhanced salinity tolerance.The disadvantages of using the consecutive transformation strategy include the need for unique selection agents and markers, which are limited and the multiple genomic locations of transgene insertion.It is difficult to obtain offspring with transgenes all localised together in the progeny following genetic segregation.After each transformation, transgenic lines have to be screened for position effects, which renders this method less than optimal and practical.

Consecutive transformation using gene targeting
Site-specific recombination (SSR) systems commonly exist in prokaryotes and lower eukaryotes with various biological functions.SSR systems have been heavily used in eukaryotic cells for genome modification to remove DNA fragments and gene targeting.In plants, SSR systems are implemented in genetic transformation technology for DNA removal, especially selectable marker gene (SMG) removal and site-specific gene integration 14,15 .Using SSR systems, transgenes can be stacked sequentially into a single locus.Examples of SSR methods for gene stacking include the circular DNA moleculecarried successive transgene insertion 16 and recombinase-mediated cassette exchange (RMCE)-mediated successive transgene integration 17 .In Ow's case, an attP-containing circular DNA molecule (bearing a gene of interest (GOI)) can be integrated into can be assembled into a single T-DNA (in the case for Agrobacterium-mediated transformation), then transformed into plants as a unit to a single locus.For example, the engineering of starch biosynthesis in maize endosperm increases the total starch by using multiple transgenes 8 .The prerequisite of this strategy is that all transgenic cassettes must be proven to be effective at the time of transformation.There are practical vector construction costs and size constraints associated with this method.Also, once this strategy is implemented, it cannot be supplemented, except by breeding.

Single transformation with all transgenes on a mini-chromosome
Mini-chromosomes or plant artificial chromosomes, can be used to deliver large amounts of DNA into plants.

Two methods have been established
for the creation of mini-chromosomes, including telomere-mediated chromosomal truncation of natural chromosomes 9 and in vitro-assembly from the minimum constituent parts of natural chromosomes 10 .Plant mini-chromosomes were first successfully produced from Arabidopsis telomeric sequence-mediated chromosomal truncation in maize 9 .It was observed that the foreign genes that were integrated into the mini-chromosomes (truncated maize B chromosome) were faithfully expressed 11 .Carlson et al. 10 created circular maize autonomous mini-chromosomes by bombarding 7-190 kb maize genomic DNA fragments containing centromere elements into embryogenic maize tissue, which were stable through mitosis and meiosis.Their reporter genes were expressed through four generations.Benefits of mini-chromosomes include their transgenerational stability without unwanted pairing or recombination with endogenous chromosomes during meiosis.Chromatin Inc. has been using mini-chromosome platforms to produce their trait-stacked biotech crop products 12 .However, transgenes.For example, combining an insect-resistant transgenic event with an herbicide-resistant event, both being deregulated, can be performed by sexual crossing and selecting progeny containing both transgenes.The main advantage of this strategy is two-fold.First, combining two deregulated events by sexual crossing necessarily results in a deregulated line, at least in the USA.Second, this strategy also removes some uncertainty about the transgene performance in the resulting progeny.This strategy can be extended to multiple transgenes.For example, Datta et al. 4 produced disease-and pest-resistant transgenic rice by conventional crossing of transgenic plants: those containing Xa21 (resistance to bacterial blight), a Bt gene (for insect resistance) and a chitinase gene (for resistance to sheath blight).However, breeding of this sort is a labour-intensive and time-consuming exercise to recover elite germplasm.During introgression, after each backcross cycle, genotyping and marker-assisted selection of a large backcrossed population can decrease the number of progeny and time required, but it is an unwieldy process 5 .Several years/ generations are required to remove the 'linkage drag' that occurs when the unwanted donor DNA is closely linked to the target gene.In a tomato backcross breeding programme, the linkage drag (51 cM) lingered around the target Tm-2 gene (which confers resistance to tobacco mosaic virus) after 11 backcrosses 6 .Linkage drag may cause a reduction in fitness in cultivars when deleterious alleles are linked to desirable genes following introduction, which negatively impacts product quality or yield 7 .Biotechnological solutions could replace breeding for stacking transgenes.

Single transformation with all transgenes on a plasmid
A group of available transgenes with appropriate promoters and terminators In the proposed strategy, as depicted in Figure 1, a TALEN pair is designed to target the sequence of a specific locus 'A' (Figure1A).The TALEN pair induces a DSB at the sequence (Figures 1B and 1C) and activates host DSB repair pathways.Donor DNA molecules with a fragment ('GOI 1 -attP 0 flanked with homologous sequence of the locus "A" sequence') are provided (Figure 1D) and integrate into the DSB location through HR repair pathway (Figures 1E and 1F).The integration plasmid with -attB 1 -GOI 2 -attB 2 -DNA fragment is used for the initial transgene integration (Figure 1G).The plasmid inserts into locus 'A' through attP 0 and attB 1 site-specific recombination catalysed by site-specific recombinase (Figure 1H) and generates a recombination product with DNA fragment -GOI 1 -attR 1 -GOI 2 -attB 2 -lox-SMG1-lox-attL 1 -, which contains two hybrid sites: attR 1 and attL 1 (Figure 1I).GOI 2 is now stacked next to GOI 1 gene.attR 1 and attL 1 are 'dead' sites and cannot be used as substrates for the corresponding recombinase any more.SMG1 is flanked by two loxP sites of Cre-lox SSR system and can be removed at this stage or later on in the presence of Cre protein (Figure 1K).attB 2 is the second attB site brought in for the next integration run.For sequential integration, the second integration plasmid with a gene cassette -attP 1 -GOI 3 -attP 2 -is used (Figure 1J).The attP 1 of this integration plasmid recombines with the attB 2 site brought in from the first integration.The recombination will integrate the whole plasmid and generate a product: -GOI 1 -attR 1 -GOI 2 -attL 2 -GOI 3 -attP 2 -lox-SMG2-lox-at-tR 2 -lox-SMG1-lox-attL 1 (Figure 1K).SMG1 and SMG2 are two different selectable marker genes.Both SMG1 and SMG2 are now located within two direct-oriented loxP sites and can be removed through Cre-lox SSR-mediated deletion (Figures 1K  and 1L).This can be accomplished by crossing the recombination line DNA sequence allow these technologies to delete or replace DNA at the target sequence.DSBs repaired through NHEJ pathway resulting in short insertions or deletions, which disrupt coding region and function of the targeted gene.Alternatively, DSBs can be repaired by HR pathway, which can be harnessed to carry out gene replacement with customdesigned donor DNA sequence, e.g. in this case, an SSR recognition site, can be achieved through HR.For simplicity, we propose using TALENs as a representative technology for use in our proposed successive transformation system since plant genomes seem to be readily mutated using TALENs 20,21 .
A specific genome locus might be chosen because it is in a 'safe spot' for which transgene introgression to wild relatives would be low 22 which is also a locus associated with high and stable transgene expression 23 .The TALEN would be designed to cut the DNA specifically in this locus and insert a recombination site to target transgenes using, for example, the unidirectional fC31-att or Bxb1-att SSR system.A unidirectional SSR system is needed since an inserted transgene will not be excised out after insertion in the presence of the corresponding recombinase.The first GOI-containing transgenic construct can be targeted and also include DNA to enable a second subsequent GOI to be targeted at a later date.Sequential DNA insertion can be obtained indefinitely as GOIs are discovered.To the best of our knowledge, there are no papers reporting the use of such a system, although the EXZACT™ technology from Dow AgroSciences is being employed to use ZFN technology for transgene stacking 24 .Gene targeting technologies coupled with SSR systems should be very useful for diverse transgene stacking applications and to improve precision of expression and also help in the biosafety/ regulatory arena.a genomic locus containing attB (preembedded in the genome) through site-specific recombination between attP and attB sites in the presence of the corresponding recombinase, resulting in the insertion of the GOI.attP and attB are the recognition sites of a unidirectional SSR system.The constructs can be designed so this process can be repeated and can bring in various GOI one by one at the same locus.In the meantime, the SMG can be deleted after serving the purpose and re-used later.Below, we propose a modification of the Ow method in which TALENs and unidirectional SSRs are used for precision gene stacking in a targeted locus (Figure 1).

TALENs and SSRs for locusspecific transgene stacking
SSR systems can only be used for locus-specific gene targeting when a recognition site sequence is preembedded into the genome.This can be accomplished using powerful and relatively new tools such as sequence-specific designer nucleases such as zinc finger nucleases (ZFNs), transcription activator-like effector nucleases (TALENs) and clustered regularly interspaced short palindromic repeats (CRISPR)/Cas systems 18 .Both ZFNs and TALENs are hybrid proteins made by fusing an engineered DNA binding domain with the catalytic domain of FokI endonuclease.CRISPRs are an adaptive immune system of bacteria to destroy foreign DNA, such as virus or plasmids.The system uses a single protein, Cas9 endonuclease, to generate double strand breaks (DSBs) at a specific DNA sequence 19 .Each of these technologies can induce DNA DSBs at a target locus that can be designed into each system and trigger the response of DNA repair systems.As a result, the broken DNA is repaired primarily through non-homologous end-joining (NHEJ) or less frequently via homologous recombination (HR) repair pathways.Induced DSBs at a target Competing interests: none declared.Conflict of interests: none declared.
All authors contributed to conception and design, manuscript preparation, read and approved the final manuscript.
All authors abide by the Association for Medical Ethics (AME) ethical rules of disclosure.
with a Cre-expressing line.The same SMG1 can then be re-used in the integration plasmid for the third integration experiment (Figure 1M).The remaining loxP site (Figure 1L) combines with the loxP sites of next incoming plasmid (Figure 1M) and can be used to remove SMG(s) and some of the hybrid sites produced in the next integration run.One of the unknown grounds is that if the presence of multiple 'dead' hybrid sites ('repeats') may have negative impact on the precise site-specific recombination and cause unwanted DNA rearrangement.Removing and reducing the number of hybrid sites may help the process of precise sitespecific recombination.fC31-att SSR system was tested for successive gene stacking in Drosophila 25 .The authors observed that fC31 often carried out random and promiscuous recombination among multiple attB and attP sites in Drosophila.By alternatively using the recognition sites of Bxb1-att and fC31-att SSR systems, they were able to reduce illegitimate recombination and carry out multiple runs of successive targeted gene integration 25 .The use of Bxb1-att as a single SSR system for successive multiple transgene integration is under testing in planta 16,26 .So far, there is no evidence showing that Bxb1 makes promiscuous recombination among multiple attB and attP sites in tobacco plants 16 .From these preliminary study results, the use of a particular unidirectional SSR system can be a factor to the success of a recombinase-mediated gene stacking programme.

Perspectives
The site-specific recombination systems are a promising gene-staking tool.By strategically including key recombination sites in the inserted DNA, transgenes can be sequentially stacked into the specific locus adjoining the previous transgenes.Recognition sites and selectable An attP site of a unidirectional SSR system is brought into the locus through host homology-directed DSB repair pathway for recombinase-mediated transgene integration and stacking (C-F).Circular DNA (G) integrates into the genomic locus (H) through genomic attP and circular DNA attB recombination.For subsequent integration, circular DNA (J) integrates into the target (I) to yield construct (K).GOI, gene of interest; SMG, selectable maker gene; attP or attB, recognition sites of site-specific recombination system; attL or attR, recombination hybrid sites; lox, loxP site of Cre-lox site-specific recombination system; Cre, Cre recombinase protein.
marker genes can be alternated and removed as needed by using a second recombination system.Theoretically, an unlimited number of novel genes can be integrated sequentially at different times at this locus through this design.The recombinase-mediated gene stacking can be accomplished at a specific genomic locus by the combined use with an engineered nuclease system (such as TALENs) in plants.This locus can be like the Rosa26 locus in mice, which facilitates high levels of constitutive, ubiquitous gene expression.Because of the highefficiency of gene-targeting into this locus, over 130 independent knock-in lines have been produced in Rosa26 23 .Alternatively, a ZFN or CRISPR can be used to target a locus for integration, as well as a variety of unidirectional SSRs 27 .This system should be amenable to any plant given sufficient genomic knowledge to facilitate choosing an appropriate locus for integration.

Conclusion
Crops with multiple transgenic traits have a better chance of overcoming problems caused by a variety of diseases, pests, environmental stresses, and thereby have increased yields.The trend for producing and growing transgenic crops with multiple transgenes continues.Several methods have been developed to deliver multiple genes into plant genomes, including SSR systems.SSR systems are user-friendly molecular tools for eukaryotic genome editing, such as DNA deletions or integrations, although their main utility involves selectable-marker gene removal.Through site-specific recombination, transgenes can be stacked at a 'landing pad' pre-embedded in the genome of interest.Engineered nucleases, such as TALENs, however, should be able to direct this preembedded landing pad to a specific locus to facilitate subsequent SSRmediated transgene stacking.
In several deregulated commercial transgenic plants, conventional breeding has been used to stack Licensee OA Publishing London 2013.Creative Commons Attribution License (CC-BY) For citation purposes: Yau YY, Easterling M, Stewart Jr N. Precise transgene stacking in planta through the combined use of TALENs and unidirectional site-specific recombination systems.OA Biotechnology 2013 Aug 01;2(3):24.
Licensee OA Publishing London 2013.Creative Commons Attribution License (CC-BY) For citation purposes: Yau YY, Easterling M, Stewart Jr N. Precise transgene stacking in planta through the combined use of TALENs and unidirectional site-specific recombination systems.OA Biotechnology 2013 Aug 01;2(3):24.

Figure 1 :
Figure 1: TALEN-guided recombinase-mediated gene stacking.A TALEN pair is designed to induce a DSB at the target sequence of a specific locus (A-B).An attP site of a unidirectional SSR system is brought into the locus through host homology-directed DSB repair pathway for recombinase-mediated transgene integration and stacking (C-F).Circular DNA (G) integrates into the genomic locus (H) through genomic attP and circular DNA attB recombination.For subsequent integration, circular DNA (J) integrates into the target (I) to yield construct (K).GOI, gene of interest; SMG, selectable maker gene; attP or attB, recognition sites of site-specific recombination system; attL or attR, recombination hybrid sites; lox, loxP site of Cre-lox site-specific recombination system; Cre, Cre recombinase protein.