Live-Cell Imaging of GFP in Plants
Adapted from “How to Study Gene Expression,” Chapter 7, in Arabidopsis: A Laboratory Manual (eds. Weigel and Glazebrook). Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, USA, 2002.Green Fluorescent Protein
The GFP from jellyfish Aequorea victoria is a very stable and relatively small protein of 238 amino acids.
GFP serves as a molecular marker that can be imaged dynamically in living cells, both in its native form and as a fusion to other proteins. For GFP imaging, plants present the challenge of autofluorescence from chlorophyll, lignified cell walls, vacuolar contents, and other cell materials, all of which can obscure the GFP signal. Maximizing the signal-to-noise ratio is a major concern, and careful consideration should be given to the choice of tissue imaged, GFP expression level, GFP spectral variant (e.g., mGFP5), and selection of filter sets that block as much autofluorescence as possible. Root tips are among the best Arabidopsis tissues for live-cell imaging as they lack chlorophyll, are transparent, and can be grown on a microscope stage.
Construct Design
Successful imaging of GFP fusions depends on adequate expression levels relative to background autofluorescence. GFP imaging is principally used for two purposes: subcellular localization and tissue-specific monitoring of reporter gene expression. For the first purpose, expression of a GFP fusion from a strong promoter such as the CaMV 35S promoter may be appropriate, and fluorescence intensity in these cases is less likely to be a problem. When using GFP for subcellular localization, it is desirable to determine whether the fusion protein will complement a mutation in the endogenous gene encoding the fusion partner. Several examples of mutant rescue with GFP fusion proteins have been reported in the literature. Unless related proteins have been analyzed, one cannot predict whether it is better to fuse GFP to the amino or carboxyl terminus; therefore, it is advisable to try out both amino- and carboxy-terminal fusions. Furthermore, including a linker consisting of six to 10 alanine residues can improve stability and/or folding of the fusion protein (Doyle and Botstein 1996).
Temporal or spatial expression patterns may be monitored using either native or modified GFP, or with translational fusions, in which GFP is inserted into a genomic clone containing both regulatory elements for transcriptional control of the gene of interest and its coding sequence. In the latter case, expression levels could pose a concern and limit detection of GFP fluorescence, especially in tissue with high autofluorescence such as leaves. One way to improve expression of a GFP reporter is to use genetic amplification. In this technique, the promoter of interest is fused to a heterologous trans-activator such as GAL4:VP16. GFP is then expressed under regulatory sequences that respond to these trans-activators. When GFP is used as a direct reporter for expression patterns (i.e., not as a fusion to an endogenous protein), one must be concerned that GFP can passively diffuse into the nucleus or to adjacent cells. For these reasons, it is advisable to use a version of GFP that is targeted to a specific compartment, such as the endoplasmic reticulum (Haseloff et al. 1997), or to increase the size of GFP (Grebenok et al. 1997; Crawford and Zambryski 2000).
GFP Imaging
GFP imaging can be performed on an inverted epifluorescence microscope or confocal laser-scanning microscope. For time-lapse imaging, set up the microscope such that the sample is exposed to the minimum amount of excitation required for imaging. This will prevent photobleaching of GFP as well as minimize cell damage and cell cycle arrest. For time-lapse imaging, the fluorescence source must be shuttered (most confocal microscopes have a time-lapse macro). If available, choose band pass filter sets (rather than long pass) appropriate for the GFP variant used (e.g., for imaging mGFP4 on a Zeiss 410 confocal microscope, a 488-nm excitation filter in combination with a 515-565-nm emission filter works well). For confocal imaging, use the minimum pinhole required for imaging to increase resolution.
Imaging of Roots
Place the sample chamber on the stage, and bring the sample into focus using the desired objective. It is important that the root tip remains in direct contact with the coverslip--this will optimize resolution and prevent changes in the focal plane during root growth. The sample chamber must be sealed to prevent movement caused by desiccation, and the sample should not be exposed to extremes of temperature that might inhibit normal growth. Optimize microscope settings on a practice sample. For time-lapse imaging over more than a few minutes, image emerging lateral roots that have not yet elongated significantly. Such lateral roots will not grow out of the field of view over time.
Imaging of Primordia on the Shoot Apex
For imaging of emerging primordia on the inflorescence apex, remove as many of the older flowers as possible. Short-day-grown plants, which have larger shoot apices, allow dissection down to the apical meristem of flowering plants. When removing the last floral buds, it is helpful to leave a few buds on one side of the apex intact to encourage the apex to lie on its side (see below). Vegetative meristems are a little harder to get to, but by peeling off one of the cotyledons from a young seedling, the entire apical region can be imaged. Mount freshly dissected material immediately in water or glycerol containing 1% Tween 20, with the meristem region as close as possible to the coverslip. If some floral buds remain on the apex, mount the apex in such a way that the intact buds face upward (away from the slide). When the coverslip is placed on top, these buds will be pushed aside, and the apex will come to lie on its side. This position is good for longitudinal views. 40X and 63X oil or water immersion objectives both work well for imaging.
Imaging of Hypocotyls and Leaves
Imaging epidermal cells of hypocotyls and leaves is quite easy since epidermal cells, apart from guard cells, do not contain chlorophyll. Imaging of mesophyll cells is more difficult, but not impossible if expression levels are high enough. For transverse views, use fresh hand-cut sections. Imaging of vascular and mesophyll tissues of the hypocotyl also works well.










