Flow cytometric analysis of transgene expression in higher plants: Green fluorescent protein

Publisher Summary The green fluorescent protein (GFP) of Aequorea victoria provides a unique means to study living cells, because transgenic expression of the coding sequence is sufficient to produce fluorescence within the cells of many different pro- and eukaryotic species. The availability of GFP as a transgenic marker has spurred interest in methodologies and instruments that can accurately measure cellular fluorescence emission, of which flow cytometry is one. Flow cytometric instrumentation is designed to examine the optical properties of suspensions of single cells or cellular homogenates as they are constrained to pass rapidly in a fluid stream through focused illumination. The individual cells or subcellular particles scatter light and, if they contain fluorochromes, also absorb light that is subsequently reemitted in the form of fluorescence. These signals are detected using sensitive photomultiplier tubes (PMTs) and photodiodes screened by appropriate wavelength-selective filters. After conversion to voltage-versus-time pulse waveforms, the intensity values are digitized and stored for analysis or further processing; the values are stored either on a cell-by-cell basis (“listmode”), or in the form of frequency distributions. Flow cytometric data are then presented in the form of multidimensional population distributions, which are graphically represented as histograms. This method of data accumulation and display, because it preserves differences between cells, is different from that commonly encountered in biochemical determinations, which provide only population means. Flow cytometry therefore allows identification of subpopulations of cells having different optical characteristics; these typically are observed as individual clusters within two-dimensional representations. Flow cytometry is also inherently quantitative, in that the intensities of the individual light scatter and fluorescent signals are accurately recorded. Either linear or logarithmic scales can be employed, the latter being preferred when signals of large dynamic ranges are encountered.