Development of late blight resistant potatoes by cisgene stacking

Abstract

Background: Phytophthora infestans, causing late blight in potato, remains one of the most devastating pathogens in potato production and late blight resistance is a top priority in potato breeding. The introduction of multiple resistance (R) genes with different spectra from crossable species into potato varieties is required. Cisgenesis is a promising approach that introduces native genes from the crops own gene pool using GM technology, thereby retaining favourable characteristics of established varieties.

Results: We pursued a cisgenesis approach to introduce two broad spectrum potato late blight R genes, Rpi-sto1 and Rpi-vnt1.1 from the crossable species Solanum stoloniferum and Solanum venturii, respectively, into three different potato varieties. First, single R gene-containing transgenic plants were produced for all varieties to be used as references for the resistance levels and spectra to be expected in the respective genetic backgrounds. Next, a construct containing both cisgenic late blight R genes (Rpi-vnt1.1 and Rpi-sto1), but lacking the bacterial kanamycin resistance selection marker (NPTII) was transformed to the three selected potato varieties using Agrobacterium-mediated transformation. Gene transfer events were selected by PCR among regenerated shoots. Through further analyses involving morphological evaluations in the greenhouse, responsiveness to Avr genes and late blight resistance in detached leaf assays, the selection was narrowed down to eight independent events. These cisgenic events were selected because they showed broad spectrum late blight resistance due to the activity of both introduced R genes. The marker-free transformation was compared to kanamycin resistance assisted transformation in terms of T-DNA and vector backbone integration frequency. Also, differences in regeneration time and genotype dependency were evaluated.

Conclusions: We developed a marker-free transformation pipeline to select potato plants functionally expressing a stack of late blight R genes. Marker-free transformation is less genotype dependent and less prone to vector backbone integration as compared to marker-assisted transformation. Thereby, this study provides an important tool for the successful deployment of R genes in agriculture and contributes to the production of potentially durable late blight resistant potatoes.

Figure 1

Detached leaf assays of transgenic potatoes obtained by marker-assisted transformation with single R gene constructs. Non transformed Atlantic or Bintje were susceptible to four P. infestans isolates. Rpi-vnt1.1-containing transgenic plants were susceptible to EC1 and Rpi-sto1-containing transgenic plants were susceptible to pic99189.

Figure 2

Schematic diagram of the marker-free double gene construct pBINAW2:Rpi-vnt1.1:Rpi-sto1. In light green and light blue arrows the Rpi-vnt1.1 and Rpi-sto1 genes are shown, respectively. The red arrows indicate the coding regions of Rpi-vnt1.1 or Rpi-sto1. Unique restriction enzyme recognition sites XmaI, SbfI and AscI are shown. RB: right border of T-DNA, LB: left border of T-DNA, TetA, trfA, NPTIII, ColE1, oriV and traJ are vector backbone sequences for plasmid stability and replication in bacterial hosts Agrobacterium tumefaciens and Escherichia coli.

Figure 3

Vector backbone integration in marker-free transformation events. Atlantic (H), Bintje (F) and Potae9 (W), were transformed with construct pBINAW2: Rpi-vnt1.1:Rpi-sto1. PCR analysis was performed using primers specific for tetA, trfA, NPTIII, ColE1, oriV and traJ to detect vector backbone integration. The plasmid pBINAW2:Rpi-vnt1.1:Rpi-sto1 was used as a positive control and the untransformed Atlantic as a negative control. Only the NPTIII primers amplified an a-specific fragment of similar size as shown here for the backbone free events W43-1 and W43-5 in untransformed Potae9. None of the other primers amplified an a-specific band in Potae9 or Bintje (data not shown) M: molecular weight marker.

Figure 4

Functional validation of cisgenic transformants by agroinfiltration and resistance assays. A. Avrvnt1- and Avrsto1-induced hypersensitive responses in cisgenic transformant H43-7 (Rpi-vnt1:Rpi-sto1 in Atlantic background). Avrvnt1 and Avrsto1 were infiltrated in cisgenic plants. A 1:1 mixture of R3a and Avr3a and pK7WG2 were infiltrated as positive and negative controls, respectively. B. Detached leaf assays for cisgenic transformant H43-7. Different isolates are shown in the middle. Cisgenic transformant are shown on the top and the wild type Atlantic on the bottom of the panel.