The amazing green fluorescent protein

Events leading to the cloning and expression
of its gene and reflections on its impact

Only Mother Nature could construct a molecule whose fluorescence quantum yield approaches 100 percent when dissolved in water. This characteristic of green fluorescent protein is the reason its gene has revolutionized cell biology. The use of the GFP gene is responsible for advancing our knowledge of mechanisms in many areas of cell biology, such as gene expression, cell division, cytoskeletal organization, vesicle trafficking and neurotransmission. Moreover, only during a time when a project was supported, as a matter of course, because it asked an interesting, fundamental question about the natural world (in this case, “How do marine invertebrates emit light?”) would GFP have been discovered.
Many of the details leading to the cloning of the GFP gene from the laboratory led by Milt Cormier (one of the authors) never have been reported. Because of the importance of the GFP gene, we feel that these details may be of interest to the scientific community.

From enzymology to molecular biology


Cormier began his graduate work at the University of Texas at Austin under the guidance of Lester Reed and obtained his Ph.D. at the Oak Ridge National Laboratory in Tennessee as a fellow of the Oak Ridge Institute of Nuclear Studies. While at Oak Ridge, (see Cormier et al and Bernard Strehler et al) discovered two of the components required for light emission in luminous bacteria. During his stay at Oak Ridge, Cormier had the pleasure of meeting many well-known scientists from various parts of the world as a result of the famous Gatlinburg Conferences held each year during this time.
Starting in the late 1950s, the goal of the research program in the Cormier lab at the University of Georgia was to understand the biochemistry and biophysics of light emission in bioluminescent marine invertebrates, with the major focus initially on the sea pansy Renilla reniformis, which is an anthozoan soft coral common along the Georgia coast.
During the 1970s, Bill Ward was a postdoc in the lab. He is now a professor at Rutgers University in New Jersey. Harry Charbonneau was a graduate student at the time and is now a professor at Purdue University in Indiana. Rick McCann (also an author) was a technician in the lab and is now a professor at Mercer University School of Medicine in Georgia.
Over several summers, the three of them went to the University of Washington Marine Laboratory in Friday Harbor to collect the bioluminescent jellyfish Aequorea victoria. In that period, thousands of jellyfish were collected and processed, and the extracts were frozen on dry ice for transportation back to the laboratory in Georgia.
After moving to Rutgers, Ward, who continued to collect Aequorea at Friday Harbor alongside the Cormier group, focused his research on the structure and function of Aequorea GFP after characterizing Renilla GFP at the University of Georgia. Charbonneau, who was by then a postdoctoral fellow with Tom Vanaman at Duke University, was determining the amino acid sequence of the Ca2+-activated photoprotein aequorin using protein purified by McCann at UGA.
By the late 1970s, members of the Cormier laboratory had isolated and characterized the three major proteins involved in bioluminescence in Renilla reniformis: luciferase, luciferin-binding protein and GFP. It became apparent that we would never be able to isolate sufficient amounts of these Renilla proteins in order to study their structure–function relationships required for bioluminescence. We had to take a different approach.
There were by then two examples of the cloning of genes in higher organisms. One was the cloning of the gene coding for human insulin. So Cormier decided to change his lab from an enzymology lab to a molecular biology lab. Since Charbonneau had made significant progress in determining the amino-acid sequence of aequorin, an attempt to clone the aequorin gene seemed logical.
This was before the facile cloning of your favorite gene was a routine procedure in every lab: no polymerase chain reaction; no automated DNA sequencers; no commercially available plasmids with multiple cloning sites; no cloning kits; no BLAST, or basic local alignment search tool. From the partial amino-acid sequence of aequorin, we were able to derive oligonucleotide probes that were used subsequently to identify putative aequorin clones.
At about this time, the National Science Foundation grant that supported the Cormier lab was up for renewal, so Cormier submitted a new grant proposal to support the cloning work. For the first time in 25 years, his funding request was denied. Fortunately, Cormier had a contact at a major pharmaceutical firm who seemed interested in the project. After he presented a seminar to the company, it offered generous support for the cloning work.
At that point, Cormier began looking for a molecular biologist who could help clone the aequorin gene. Doug Prasher, then a postdoc in the UGA genetics department, was interested and agreed to join the Cormier lab in the early 1980s. By the time Prasher arrived, everything was in place for molecular biology, including some frozen Aequorea tissue. That summer Prasher and McCann went to Friday Harbor to collect more jellyfish and, ultimately, construct an Aequorea cDNA library.
By the autumn of 1984, Prasher felt that he had isolated the aequorin gene based on the hybridization of the aequorin-specific oligonucleotides to several clonal isolates, but he could not verify this, because he was having difficulty in expressing the gene. We had a conversation about this problem.
McCann suggested Prasher might be expressing aequorin at a low level even from pBR322, which was an early cloning vector in which inserts were cloned into either ampR or tetR genes but not an expression plasmid, and that this could be measured in E. coli extracts, given that it is possible to detect sub-attomole (10-18 mole) levels of aequorin.
We suggested Prasher look for expression using a bioluminescence assay used routinely in the lab. The very first try produced so much light that the luminometer became saturated. There was jubilation in the lab. We knew then that we had expression of aequorin. That paper was published in 1985.

Aequorea victoria
Aequorea victoria. Image courtesy of Sierra Blakely - Wikimedia Commons  

Cloning of the GFP gene


Upon completion of our work on aequorin, Cormier suggested that Prasher try to clone the GFP gene, since we already had a cDNA library from Aequorea. Furthermore, Ward was willing to furnish us with partial amino-acid sequence data. Prasher agreed and was successful in isolating a GFP clone. When the gene was sequenced, we realized that the clone represented 70 percent of the coding sequence.
Since Prasher could not identify the full-length gene in that cDNA library, it was obvious that additional collections of Aequorea were required. At this point, Prasher obtained a position at the Woods Hole Oceanographic Institution, but he and Cormier agreed to continue their collaboration on the cloning of GFP.
Cormier was running out of research funds again by then, so he applied to the NSF. Once again, the grant was turned down. This loss of funding forced the closure of his lab. Cormier subsequently retired but insisted that two assistant professors be hired to replace him. That was done. He also remained available while Prasher continued his work on GFP. Fortunately, Prasher obtained independent funding in 1989. An additional collection of Aequorea was made, and the full-length gene was isolated. That work was published in 1992.
Based on the protein sequence of GFP, Frank Prendergast, a professor at the Mayo Medical School in Minnesota who earlier had published a paper on the characterization of Aequorea GFP, predicted the likely GFP chromophore structure.
Prasher then turned his attention to the expression of GFP.
After making a number of attempts to express the gene, he phoned Cormier about the difficulty he was having. Cormier assured Prasher that he would figure it out. However, Prasher’s research position at WHOI was ending. (Had Cormier known this, he would have urged Prasher to return to Georgia to complete his work on the expression of GFP.)
Rather than let the project languish, Prasher gave the gene to Martin Chalfie and Roger Tsien in 1992 upon their request. The rest, as they say, is history. Chalfie’s lab figured out how to express the GFP gene shortly after receiving it. They then used the GFP gene to study gene expression in living cells. This work was published in 1994.
Tsien subsequently designed variants of GFP that fluoresce in the various colors we now use in virtually all of cell biology. The Nobel Prize committee credited Osamu Shimomura with the discovery of GFP in Aequorea and gave him, Chalfie and Tsien the Nobel in chemistry in 2008. The seminal contributions of Prasher to this work were not forgotten by Chalfie and Tsien, however, and that part of the story was covered in a long article in the magazine Discover in 2011.
But there is a larger point about the work that led to the discovery of GFP and the ongoing revolution in cell biology that has been facilitated by this fascinating molecule. All of the foundational research in the Cormier laboratory and that of the others who worked on bioluminescence in coelenterates was supported by the NSF, the National Institutes of Health and a precursor to the Department of Energy largely because it addressed a fundamental question straight out of natural history: How do these organisms emit light?
The simple answer is that one protein with enzymatic activity (luciferase, aequorin) oxidizes a reduced substrate (luciferin, coelenterazine) to produce blue light. In the organism, however, the energy from the excited state of the substrate is transferred nonradiatively with high efficiency to GFP, which then emits the green light seen when living Renilla or Aequorea are stimulated. This is the same green color seen when the original Aequorea GFP, expressed as a recombinant protein, is excited by blue light in a fluorescence microscope.

Multiple applications in cell biology


Native aequorin was first injected into living cells and used as a calcium indicator by E.B. Ridgway and C.C. Ashley in 1967. Today, aequorin-expression vectors are used to measure calcium transients in animal, plant and fungal cells. Shortly after aequorin was cloned in the Cormier laboratory and following on the work of : John Matthews, now a professor at the University of Mississippi, Walt Lorenz and McCann cloned and expressed Renilla luciferase. This luciferase, in tandem with firefly luciferase, which has a completely different substrate, is now used widely as a reporter for gene expression in cells of all types. The use of GFP in all its colors is limited only by the imagination of cell biologists. For example, fluorescence resonance energy transfer, known as FRET, between GFP variants of different colors can be used to measure the distance between molecules in living cells.
Although we outline here events that occurred in the Cormier lab, Frederick Tsuji also has written an informative article from his perspective regarding the history and cloning of the GFP gene. All investigators who took part in this exciting endeavor to clone and express the GFP gene and the other proteins responsible for coelenterate bioluminescence should take heart in knowing that they were part of the effort that resulted in a revolution in cell biology.
Moreover, as outlined recently by Joram Piatigorsky, those responsible for deciding which research gets funded, from policymakers to members of review panels, should remember that the answers to scientific questions cannot be known in advance and that these questions and answers often lead to advances in scientific knowledge and scientific practice that are as revolutionary as they are unimaginable and unpredictable.

Milton J. Cormier
Richard O. McCann
Milton J. Cormier (milton.cormier@ is a distinguished research professor emeritus at the University of Georgia.

Richard O. McCann (mccann_ro@ is an associate professor at Mercer University School of Medicine and director of the graduate program in biomedical sciences.

comments powered by Disqus
found= true2901