An Analysis of Agrobacterium tumefaciens Attachment to Leaf Tissues and Subsequent Plant Transformation

Plant-based expression systems offer a number of benefits compared to traditional microbial expression systems for producing recombinant proteins, including the ability to perform certain post-translational protein modifications and the potential to reduce energy inputs due to their autotrophic nature. While most commercially available transgenic crops are stable transformants expressing genes that confer agronomic traits to the plant, the possibility exists for high-level plant-based production of recombinant proteins that have inherent value not just to growers, but to consumers and other industries directly. Transient expression of transgenes in planta can provide higher levels of recombinant protein in a fraction of the time needed to generate stably transformed plants. Furthermore, the rapid nature of transient expression facilitates recombinant protein production in harvested plant tissues. The bacterium Agrobacterium tumefaciens is commonly used to induce transient expression in plant tissue via its endogenous ability to transform plants. However, certain aspects of the A. tumefaciens infection process are not well understood. Specifically, the fate of the bacteria after entering leaf tissue is largely unknown in part due to lack of knowledge regarding what factors affect the attachment of bacteria to leaf cells. Understanding how bacteria distribute themselves within leaf tissue, infiltration factors that influence bacterial attachment, and the relationship between attachment levels and in planta transgene expression may provide avenues for improving transformation efficiency and expression levels.;To quantify A. tumefaciens attached to leaf cells, a rapid in situ assay was developed. The use of fluorescent stains was investigated for labeling bacteria in lieu of transforming cells to express a fluorescent protein. Several vital dyes and nucleic acid stains were tested for their ability to yield a detectable signal against a leaf tissue background using fluorescent confocal microscopy. The fluorescent nucleic acid stain Syto16 was found to perform best. Furthermore, Syto16 was found to have no significant affect on the viability or virulence of A. tumefaciens, as indicated by plate counting and agroinfiltration assays. Subsequently, a protocol was developed for quantifying bound bacteria in micrographs of sections of infiltrated leaf tissue, yielding measurements of attached bacteria per volume leaf tissue.;The in situ bacterial attachment assay was used to study A. tumefaciens attachment in response to infiltration factors including bacterial density, vacuum intensity, and surfactant concentration in both lettuce and switchgrass harvested leaf tissue. While vacuum level influenced the volume of bacterial suspension that entered leaf tissue, no significant effect on bacterial attachment levels was observed in either lettuce or switchgrass in response to varying vacuum intensity levels. For lettuce, surfactant concentration in the bacterial suspension had a positive effect on both the volume of suspension that entered leaf tissue and the amount of bacteria that attached to leaf cells. Alternately, addition of surfactant did not correspond to larger volumes of suspension infiltrated into switchgrass leaf tissue and there was no effect on bacterial binding. For both lettuce and switchgrass, bacterial adhesion was most sensitive to changes in the density at which the bacteria were infiltrated into leaf tissue. Levels of attached bacteria in lettuce leaves versus the density of infiltrated bacteria exhibited a saturation trend. Likewise, there was a positive relationship between bacterial density and attachment rates in switchgrass, although the relationship was linear with no indication of saturation at the highest bacterial density. In light of its dominant effect on adhesion levels, infiltrated bacterial density was varied and in planta transient expression levels of an agroinfiltrated beta-glucuronidase gene were measured in lettuce and switchgrass leaves. For the range of bacterial densities tested in lettuce, lower bacterial densities yielded a constant expression level. However, expression levels fell dramatically at the highest observed attachment level. These results show that infiltration parameters can be used to control Agrobacterium attachment density and that attachment levels can influence transgene transient expression levels in lettuce. However, the results also reveal that optimizing expression levels is not simply a matter of promoting maximal bacterial attachment, as attachment levels beyond a threshold correspond to plant responses that ultimately lower transgene transient expression levels. Alternately, two methods for measuring beta-glucuronidase activity in agroinfiltrated switchgrass demonstrated that in planta transgene transient expression levels remain low regardless of Agrobacterium attachment levels. Spectrophotometric activity assays performed on leaf extracts revealed no significant beta-glucuronidase activity while staining for beta-glucuronidase activity showed extremely low-levels of activity signified by microscopic stained regions. The lack of high-level transgene transient expression in response to altered Agrobacterium attachment levels suggests that recalcitrance of switchgrass to agroinfiltration does not lie at the level of Agrobacterium attachment.