Fluorescent proteins have been available for several years, but new instruments that measure fluorescence have allowed researchers to use such proteins to study gene expression.

It's been about a decade since biological researchers introduced green fluorescent protein (GFP) into their molecular toolbox. Since then, this jellyfish-derived, molecular light bulb has made a huge impact on the way life scientists study cellular and molecular processes. Following its cloning in 1992, the gene gfp has been incorporated into a myriad of engineered plasmids. Genetic and spectral variants, including yellow fluorescent protein (YFP) and cyan fluorescent protein (CFP), as well as orthologues such as the sea coral protein DsRed, a red fluorescent protein (RFP), have been created.

Despite having different genetic sequences and spectral properties, all of these fluorescent proteins are extremely useful in studying cellular events such as localization of proteins to membranes and to cellular organelles. Fluorescent proteins have been useful for basic sciences such as cellular and developmental biology, as well as for the applied sciences, such as in drug discovery. New advances in instruments that measure fluorescence have expanded the utility of fluorescent proteins to include the study of gene expression. But why use fluorescent proteins to study gene expression when seemingly better methods exist? "We are using GFP in our model systems because there is no more convenient way to do it," says Kenneth Birnbaum, PhD, assistant professor of biology at New York University, New York. Birnbaum uses the fluorescent protein as a tool to do developmental genomics in the plant Arabidopsis thaliana. "The relevant scales for genomics will be at the cellular level," he says.

Although traditional tools of genomics, especially DNA microarrays, are useful for performing biochemical analyses of cellular events, they are not very useful for studying developmental biology, where the researcher may not have a purified population of specialized cell types readily available for the analysis. To study Arabidopsis root development, Birnbaum has taken advantage of the availability of GFP-expressing transgenic plants in which the protein is expressed within specific cells in the root.

The lab purifies these GFP-expressing cells from thousands of roots using a fluorescence-activated cell sorter (FACS), which sorts the cells based on whether or not they express GFP. RNA is then prepared from these cells and used in DNA microarray analysis to study genes involved in the development of the root. So, in this case, GFP is used to select a homogeneous population of specialized cells whose gene expression profile can be determined by DNA microarray analysis.

But what if a researcher wants to directly measure gene expression with a fluorescent protein? That has become possible as well. In fact, fluorescent proteins such as GFP have been used as direct transcriptional and translational reporters in living cells because they do not require fixation, enzymes, or substrates for generation of fluorescence. The fluorescence is clearly visible by fluorescence microscopy, and because of advances in the measurement of fluorescence, it is now quantifiable, even in whole organisms. Stephen Ekker, PhD, director of the Arnold and Mabel Beckman Center for Transposon Research at the University of Minnesota, uses fluorescent proteins as tools for in vivo imaging of anatomic structures and in vivo gene expression analysis in zebrafish. "Because the animal is transparent, the zebrafish allows you to do not only single, but multicolor imaging as well, so that you can label multiple structures simultaneously," Ekker says.

 Zebrafish fluorescently tagged with red fluorescent protein. Because the animal is transparent, multiple fluorescent protein tags can be visualized sumultaneously. (Source: Stephen Ekker, PhD)  
New confocal technology
The traditional fluorescence confocal microscope is a necessary instrument for all studies involving fluorescence-based imaging, but newer confocal technology has emerged for these applications. "Traditional confocals are not biocompatible for in vivo imaging studies because they will destroy the tissue that you are going to examine in most cases," says Ekker. Using fluorescent protein technology has become simpler and more productive for the Ekker laboratory since they began using the grid confocal systems. This new imaging technology allows the specimen to be examined for longer periods, such as in time-lapse experiments or for creating movies of cellular events, applications which were not possible with the traditional confocal microscopes. "As the new fluorescent technology comes online, the utility of the fluorescent proteins will go way up."

Although live cell imaging of anatomic structures is a large component of Ekker's research, he also conducts experiments to understand epigenetic regulation in zebrafish. By using GFP-containing transposons as reporter genes, Ekker is able to study the "position effect," a phenomenon in which the expression level of a randomly-integrated gene such as a transposon is dependent upon the specific genetic elements present near the site of integration. In other words, the "position effect" determines whether the gene is activated, attenuated, or repressed. Ekker says that epigenetic silencing fluorescent protein genes can occur as a consequence of the preponderance of CpG motifs, which can become methylated. "Believe it or not, if you look at the CpG content of the engineered vectors that you buy, the more engineered they are, the more CpG motifs they have….I am pretty sure that it is the CpG motifs that are causing the epigenetic silencing."

DNA promoters, as well as activator or repressor proteins, are necessary components of transcriptional regulatory systems, which are network-like in nature. Synthetic biologists study such transcriptional components by synthesizing hybrid transcriptional networks in which the modular effects of these components can be analyzed in a controlled system. James Collins, PhD, professor of biomedical engineering and codirector of the Center for BioDynamics at Boston University, is a synthetic biologist who creates synthetic gene networks with three goals in mind:
• to create biotech applications, including functional genomics and bioreactors;
• to test and validate mathematical models of transcriptional regulation;
• to explore basic networks in order to understand basic principles of biological control.

"Synthetic networks have been designed out of well-characterized biological components, such as promoters and genes, that we arrange in a fashion that they create a network or circuit of promoters, genes, and other components to produce a desired function," Collins says.

These systems are typically based on the well-defined transcriptional control systems in microorganisms, such as the lac operon in Escherichia coli which controls genes involved in the carbohydrate metabolic pathway, or the cI repressor system in bacteriophage ?. By creating a GFP reporter under the control of these specific promoters, general GFP fluorescence can be used as a "readout" to characterize the dynamics of the transcriptional network. Collins is particularly interested in understanding the variability in GFP expression which occurs as a consequence of the state of the network controllers, whether activator-only, repressor-only, or both (see figure below).

The use of fluorescent proteins to measure gene expression can circumvent the challenges that arise when studying genes expressed during in vivo bacterial infection. Researchers do not want to subject the sample to polymerase chain reaction amplification to quantify such expression for fear that disproportional amplification will yield inaccurate results. Ambrose Cheung, MD, professor of microbiology and immunology at Dartmouth University, Hanover, N.H., is using a plasmid-based GFP reporter system to identify virulence factors involved in human infection with the bacterium Staphylococcus aureus. "The gfp gene is normally expressed in mammalian cells quite well, but we adapted it so that this gene can be expressed in vivo in bacteria. So by optimizing expression in bacteria, then you can use the same tool to analyze gene expression in vivo," Cheung says.

Specifically, he creates strains of S. aureus in which each GFP expression cassette is under the transcriptional control of a different DNA promoter from a putative Staphylococcal virulence gene. By comparing the expression profile of in vitro and in vivo infection, he is able to identify virulence factors. For example, a higher level of expression of the GFP reporter in vivo than in vitro is usually a strong clue that the reporter is under the control of virulence factor promoter.

Measuring gene expression
Currently, GFP fluorescence can be detected in a variety of ways, including spectrofluorimetry, flow cytometry/FACS, fluorescence microscopy, and fluorometric imaging. Quantitative measurements of GFP fluorescence, however, are limited to spectrofluorimetry, quantitative microscopy, flow cytometry, and fluorometric imaging.

With spectrofluorimetry, the fluorescence of a sample is quantified using a specialized spectrophotometer. There have been multiple applications of spectrofluorimetry, especially for studying bacterial pathogenesis. Cheung uses spectrofluorimetry to measure changes in fluorescence in live bacterial samples using a 96-well microtiter format. "It is similar to a spectrophotometer except instead of measuring a single sample, it measures 96 samples at one time." According to Cheung, the other advantage of using this method is that it allows researchers to simultaneously measure multiple fluorescent protein types such as RFP, GFP, or YFP in a single sample by simply changing filters.
click the image to enlarge 
GFP can be used as a reporter of promoter activity. In this case, GFP is used as a reporter of repressed and activated promoter activity in E. coli. (Source: James Collins, PhD) 

Measurement of gene expression in real-time and in living cells is a challenge for cell and molecular biologists. However, advances in fluorescence microscopy, coupled with fluorescent protein technology, have allowed for the direct measurement of gene expression in a cell, tissue, or organism (such as a zebrafish) on a microscope slide without perturbation of the system. "We measure the actual light coming off the fish under the microscope using a camera," Ekker says. "One of the key features of fluorescent proteins is that they allow you to do live imaging when most other assays don't. It's live imaging, it's non-invasive, and can be performed multiple times." When measuring gene expression using in vivo imaging, validation is also necessary to achieve reproducible, meaningful results. "Fluorescence assays in animals in vivo are semi-quantitative, but pretty good." But to do more quantitative, more precise measurements of gene expression, Ekker performs Western Blotting or another analogous procedure.

Although fluorescent proteins are reliable tools for gene expression studies, it is still necessary to validate the specificity of their expression. Validation is a big part of Birnbaum's research with A. thaliana. He uses a GFP reporter system to test transgenic plant lines for which the specific cell type in the plant that expresses GFP is known. To control for false-positive expression of GFP, Birnbaum uses two other markers expressed on the specific cells in the plant root which presumably also express GFP. He then sorts these cells by FACS and determines the percentage of these sorted cells that also express GFP. "In general, the validation is 80% to 90% accurate," Birnbaum says, indicating that GFP is a good marker for sorting out specialized cells.

Other Challenges
A major limitation of using fluorescent proteins to measure gene expression is that, when measured, they have a lot of background fluorescence that can falsely elevate quantitative data. "In general, biological systems tend to have a lot more background in the blue end of the spectrum and tend to be more transparent in the red end of the spectrum," Ekker says. Although GFPs and RFPs are fundamentally different molecules with different activities, when RFPs are used for in vivo imaging they tend to have higher signal-to-noise ratios, causing them to emit brighter fluorescence.

Sensitivity of detection is also a big issue when using GFP reporters for quantitative assays. "GFP is only sensitive if it is produced in multiple copies," says Cheung, adding that GFP is not as sensitive when used to measure gene expression at the chromosomal level.

The stability of GFP can be an issue when looking at rapid gene expression events in real-time. "There are some applications where you want a fluorescent protein that does not degrade very quickly, and others where you want one that does degrade quickly," says Collins. "I think it would be interesting to have a site of fluorescent proteins of different colors—for example, green, yellow, and red—that are well-characterized and have tunable degradation properties and tunable production rates."

This article was published in G & P magazine: Vol. 6, No. 5, June, 2006, pp. G14-G16.