Critical parameters for robust Agrobacterium-mediated transient transformation and quantitative promoter assays in Catharanthus roseus seedlings

Abstract

Agrobacterium-mediated transient expression methods are widely used to study gene function in both model and non-model plants. Using a dual-luciferase assay, we quantified the effect of Agrobacterium-infiltration parameters on the transient transformation efficiency of Catharanthus roseus seedlings. We showed that transformation efficiency is highly sensitive to seedling developmental state and a pre- and post-infiltration dark incubation and is less sensitive to the Agrobacterium growth stage. For example, 5 versus 6 days of germination in the dark increased seedling transformation efficiency by seven- to eight-fold while a dark incubation pre- and post-infiltration increased transformation efficiency by five- to 13-fold. Agrobacterium in exponential compared with stationary phase increased transformation efficiency by two-fold. Finally, we quantified the variation in our Agrobacterium-infiltration method in replicate infiltrations and experiments. Within a given experiment, significant differences of up to 2.6-fold in raw firefly luciferase (FLUC) and raw Renilla luciferase (RLUC) luminescence occurred in replicate infiltrations. These differences were significantly reduced when FLUC was normalized to RLUC values, highlighting the utility of including a reference reporter to minimize false positives. Including a second experimental replicate further reduced the potential for false positives. This optimization and quantitative validation of Agrobacterium infiltration in C. roseus seedlings will facilitate the study of this important medicinal plant and will expand the application of Agrobacterium-mediated transformation methods in other plant species.

Keywords: Agrobacterium tumefaciens; Agrobacterium‐mediated transformation; Catharanthus roseus; luciferase; transient expression; vacuum infiltration.

FIGURE 1

Seedling developmental state was critical for high transformation efficiency. (a) Seedlings germinated in the dark for 5 days were less than 1 cm in height with only the radicle and apical hook developed. In contrast, seedlings germinated in the dark for 6–8 days (our previous method) were 2–3 cm in height with cotyledons emerged. (b) After growth in light for 3 days, seedlings that had germinated in the dark for 5 days were shorter than seedlings germinated for 6 days. (c) Seedlings germinated in the dark for 5 days yielded higher luminescence, indicating higher transformation efficiency, than seedlings germinated for 6 days. (d) Seedlings germinated in the dark for 5 days and then grown in the light for 2 days had higher transformation efficiency than those grown in the light for 3 days. Transformation efficiency was measured by luminescence from firefly luciferase (FLUC) driven by the pMAS (c) or pD4H promoter (d) and luminescence from Renilla luciferase (RLUC) driven by the pNOS promoter. Each data point or biological replicate is a pool of two seedlings, N = 10. ****p < .0001, ***p < .001, **p < .01 according to an unpaired two‐tailed Student's t test on log‐transformed luminescence. Box plots represent the 25th and 75th percentile with a line marking the median. Whiskers extend to the minimum and maximum.

FIGURE 2

Variations in Agrobacterium growth conditions did not critically impact transformation efficiency. (a) Agrobacteria in exponential growth phase (OD₆₀₀ = 1.0) increased transformation efficiency by about two‐fold compared with Agrobacteria in stationary phase (OD₆₀₀ = 3.0). Transformation efficiency was measured by luminescence from firefly luciferase driven by the pMAS promoter (FLUC) and luminescence from Renilla luciferase driven by the pNOS promoter (RLUC). (b) The number of days that Agrobacteria was grown on solid media did not affect transformation efficiency. Glycerol stocks of Agrobacteria were streaked onto solid LB media containing selective antibiotics and were incubated at 25°C for 2 or 3 days before a streak of colonies was used to inoculate a liquid culture used for vacuum infiltration. (c) Visual comparison of streaked Agrobacteria on solid LB grown for 2 versus 3 days. Individual colonies were not visible after 2 days of growth. (d) Three different methods of cultivating Agrobacterium were comparable for transformation efficiency. Liquid Agrobacterium cultures were inoculated either directly from a glycerol stock or were first grown on solid media for 3 days before starting cultures either from a single colony or a streak of colonies. Each data point or biological replicate is a pool of two seedlings, N = 10. **p < .01, *p < .05 according to an unpaired two‐tailed Student's t test (a,b) or a one‐way ANOVA (d) on log‐transformed luminescence. Box plots represent the 25th and 75th percentile with a line marking the median. Whiskers extend to the minimum and maximum.

FIGURE 3

Dark incubations before and after infiltration were critical for high transformation efficiency. (a) Incubating the seedlings in the dark for about 18 h prior to infiltration (pre‐infiltration dark incubation) significantly increased transformation efficiency compared with no pre‐infiltration dark incubation. (b) Incubating the seedlings in the dark for 2 days rather than overnight post‐infiltration did not impact transformation efficiency. (c) Incubating the seedlings in the dark for about 18 h after infiltration (overnight post‐infiltration dark incubation) significantly increased transformation efficiency compared with no post‐infiltration dark incubation. Transformation efficiency was measured by luminescence from firefly luciferase driven by the light‐responsive pD4H promoter (FLUC) and luminescence from Renilla luciferase driven by the constitutive pNOS promoter (RLUC). Each data point or biological replicate is a pool of two seedlings, N = 10. ****p < .0001, ***p < .001 according to an unpaired two‐tailed Student's t test on log‐transformed luminescence. Box plots represent the 25th and 75th percentile with a line marking the median. Whiskers extend to the minimum and maximum.

FIGURE 4

Normalization of EASI experiments to the RLUC reference reporter and to the control condition effectively reduced variability between replicate conditions within and between experiments. (a) In two independent experiments, four replicate conditions (cond.) were infiltrated with the same agrobacterium strains: (1) A reporter strain containing the D4H promoter (pD4H) driving firefly luciferase (FLUC) and the pNOS promoter driving Renilla luciferase (RLUC); (2) An effector strain expressing GUS. Each data point or biological replicate is a pool of two seedlings, N = 10 per experiment. (b,c) In both experiments, raw FLUC and RLUC luminescence values differed significantly between the four replicates. (d) Normalization of FLUC to RLUC removed variation between conditions within an experiment but did not remove variability between experiments. (e) Expressing normalized FLUC/RLUC values relative to Condition 1 of each experiment and taking the natural logarithm effectively removed the variability between experiments. % of total variation and p values are the result of a full‐factorial two‐way ANOVA on log‐transformed luminescence comparing the effects of replicate conditions and experimental repeats. Box plots represent the 25th and 75th percentile with a line marking the median. Whiskers extend to the minimum and maximum.