Within leaf variation is the largest source of variation in agroinfiltration of Nicotiana benthamiana

Abstract

Background: Transient gene expression utilizing syringe agroinfiltration offers a simple and efficient technique for different transgenic applications. Leaves of Nicotiana benthamiana show reliable and high transformation efficiency, but in quantitative assays also a certain degree of variation. We used a nested design in our agroinfiltration experiments to dissect the sources of this variation.

Results: An intron containing firefly luciferase gene was used as a reporter for agroinfiltration. A number of 6 week old tobacco plants were infiltrated for their top leaves, several samples were punched from the leaves after 2 days of transient expression, and protein extracts from the samples were repeatedly measured for luciferase activity. Interestingly, most of the variation was due to differences between the sampling spots in the leaves, the next important source being the different leaves on each plant. Variation between similar experiments, between plants and between repetitive measurements of the extracts could be easily minimized.

Conclusions: Efforts and expenditure of agroinfiltration experiments can be optimized when sources of variation are known. In summary, infiltrate more plants but less leaves, sample more positions on the leaf but run only few technical replicates.

Keywords: Agrobacterium meditated transformation; Agroinfiltration; Estradiol induction; Luciferase expression; Nested design.

Fig. 1

Design of the nested infiltration experiments. In the first experiment, three leaves of two Nicotiana benthamiana plants were infiltrated with agrobacterium, each leaf was sampled from five positions and luciferase activity of each sample extract was measured five times. The same procedure was repeated for four different promoters or inducer treatments driving the luciferase reporter

Fig. 2

Promoter activity in agroinfiltrated Nicotiana benthamiana leaves. The estradiol inducible XVE/LexA cassette was measured with (+) and without (−) induction. Error bars show standard deviation of all measurements

Fig. 3

Components of variance in the agroinfiltration experiment. In the first experiment (a), the variance caused by the different promoters is excluded. In the second experiment (b) none of the observed variance could be addressed to the three plant individuals within one agroinfiltration subexperiment, or its three repetitions. In both experiments, largest variation occurred between the sample disks punched from infiltrated leaves