High efficiency Agrobacterium-mediated site-specific gene integration in maize utilizing the FLP-FRT recombination system

Abstract

An efficient Agrobacterium-mediated site-specific integration (SSI) technology using the flipase/flipase recognition target (FLP/FRT) system in elite maize inbred lines is described. The system allows precise integration of a single copy of a donor DNA flanked by heterologous FRT sites into a predefined recombinant target line (RTL) containing the corresponding heterologous FRT sites. A promoter-trap system consisting of a pre-integrated promoter followed by an FRT site enables efficient selection of events. The efficiency of this system is dependent on several factors including Agrobacterium tumefaciens strain, expression of morphogenic genes Babyboom (Bbm) and Wuschel2 (Wus2) and choice of heterologous FRT pairs. Of the Agrobacterium strains tested, strain AGL1 resulted in higher transformation frequency than strain LBA4404 THY- (0.27% vs. 0.05%; per cent of infected embryos producing events). The addition of morphogenic genes increased transformation frequency (2.65% in AGL1; 0.65% in LBA4404 THY-). Following further optimization, including the choice of FRT pairs, a method was developed that achieved 19%-22.5% transformation frequency. Importantly, >50% of T0 transformants contain the desired full-length site-specific insertion. The frequencies reported here establish a new benchmark for generating targeted quality events compatible with commercial product development.

Keywords: Agrobacterium; RMCE; FLP/FRT; co-integrate vector; maize transformation; site-specific integration.

Figure 1

Schematic presentation of the DNA constructs and the intended recombinase‐mediated cassette exchange event (RMCE). (a) Target T‐DNA containing the constitutive promoter ZmUbi driving the neomycin transferase (nptII ) gene as plant selectable marker, and the same ZmUbi promoter driving the cyan fluorescent (AmCyan1) gene as fluorescent marker for selecting transformed cells. A FRT 1 site (black triangle) is placed between the ZmUbi promoter and the nptII gene, and a FRT 87 site (white triangle) is placed at the 3′ end. (b) Donor DNA3 T‐DNA containing the same heterologous FRT sites flanking a promoterless phosphomannose isomerase (pmi) gene, which confers mannose resistance when expressed and a fluorescent reporter gene, DsRed, driven by ZmUbi promoter allowing the selection of recombined transgenic events is shown as an example of a donor construct. This donor construct also contains the ZmUbi promoter driving the flp gene delivering the FLP recombinase needed for generating intended RMCE events on the 5′ of the donor DNA, an inducible cre gene by Rab17 promoter, a maize Wuschel (Wus2) gene driven by a nos promoter and a maize Babyboom (Bbm) gene driven by ZmUbi promoter on the 3′ end of the donor DNA flanked by loxP sites (inverted black triangles). Transient expression of the flp, Wus2 and Bbm gene is sufficient for recovering RMCE events. (c) RMCE event is essentially the target DNA, wherein the nptII and AmCyan1 gene between the FRT 1 and FRT 87 site is replaced with the pmi and DsRed gene on the donor DNA. The pmi gene is activated upon being inserted downstream of the ZmUbi promoter following cassette exchange between the FRT sites. All the components outside the FRT sites on the donor DNA are not integrated following recombination in an intended RMCE event. (d) The qPCR assay devised to quantify cross‐reactivity between different heterologous FRT sites. Relative positions of the gene‐specific qPCR assays, genomic DNA border‐specific PCR assays are marked with straight lines which were used for quantifying corresponding expression units and FRT junction calls, while the line with arrow indicate the relative position of the primer‐probe used for detecting excision.

Figure 2

The different stages in transformation for selecting intended RMCE events using the target line GT6. (a) Retransformation of immature embryos from the RTL containing the nptII selectable marker. (b) Selection of the putative RMCE events in media supplemented with mannose; this selection requires 2–3 rounds of transfer before a site‐specific integration event is identified. (c) Regeneration of the putative SSI event after three rounds of selection in mannose supplemented media and, (d) Rooting of the putative RMCE events in media supplemented with mannose. The overall transformation process to generate putative RMCE events takes over 3 months.