Green Fluorescent Protein

Green fluorescent protein (GFP) is a bioluminescent polypeptide consisting of 238 residues isolated from the body of Aequorea victoria jellyfish.^[1] GFP converts the blue chemiluminescent of aequorin in the jellyfish into green fluorescent light.^[2] It remains unclear why these jellyfish use fluorescence, why green is better than blue, or why they produce a separate protein for green fluorescence as opposed to simply mutating the present aequorin to shift its wavelength,^[3] but in the laboratory, GFP can be incorporated into a variety of biological systems in order to function as a marker protein. Since its discovery in 1962, GFP has become a significant contributor to the research of monitoring gene expression, localization, mobility, traffic, interactions between various membrane and cytoplasmic proteins, as well as many others.See also Green Fluorescent Protein: Research Tool.^[4]

History

Aequorea victoria was first discovered and investigated for its bioluminescence by Frank Johnson, who invited Osamu Shimomura to work with him in on a small island not far from British Columbia, where the jellyfish is abundant.^[5] Found off the west coast of the United States between British Columbia and central California,^[6] the jellyfish was considered a local phenomenon as it would drift in and out of the harbors.^[5]

Mice with GFP inserted into their genomes for neurology studies.

Shimomura was originally looking only to isolate the blue luminescent protein of Aequorea victoria, traditionally thought to be luciferase, but it would soon become apparent that the glow was in fact due to aequorin, a substance related, but slightly varying from luciferase.^[4][5] However, the light emitted from aequorin still differed from the light emitted from the wild jellyfish. This quandary led to the discovery of the green fluorescent protein responsible for this disparity, but sufficient amounts of the protein could not be collected for study until 1979. The journey to discover the nature of GFP had begun.^[5]

In the 1990’s, Douglas Prasher, Frank Predergast, and co-workers successfully cloned the gene that encoded for GFP. Martin Chalfie further pursued this line of work and was eventually able to express GFP in heterologous systems such as E. coli and C. elegans. Chalfie’s research provided the first evidence that GFP was unique as it did not require the presence of any exogenous substance or cofactor for fluorescence.^[4] The lack for the need for a cofactor proved that the cloned GFP gene contained all the information necessary for posttranslational synthesis of the chromophore. ^[3]

Roger Tsien and co-workers were intrigued by the absence of a necessary cofactor and began to research the structure of GFP and how it relates to its fluorescence. They discovered that a helix within the beta barrel structure of GFP actually contained a fluorophore responsible for fluorescence. In researching its structure, they were able to develop GFP derivatives with improved fluorescence and photo-stability. Shimomura, Chalfie, and Tsien were each recognized for their work involving GFP with the Nobel Prize in 2008.^[4] In the time since the work of these three researchers, GFP has been successfully expressed and utilized in bacteria, yeast, slime mold, plants, drosophila fruit flies, zebra-fish, and mammalian cells.^[2] Below, mice have had GFP inserted into their genomes for studies in neurology.

Structure

Primary & Secondary Structure

Green fluorescent protein () is a 21 kDa protein consisting of 238 residues strung together^[7] to form a of five α-helices and one eleven-stranded β-pleated sheet,^[1] where each strand contains nine to thirteen residues each.^[8] (To view the primary and secondary structure of GFP, go to .) These β-strands display an almost “seamless symmetry” in which only two of the strands vary in structural content.^[9] This β-sheet conforms itself through regular hydrogen bonding into a β-barrel.^[2] In GFP, the structure is so regular that of water molecules (red) can be seen following the structure of the barrel.^[9] Together with the α-helices at either end of the molecule, a nearly perfect cylinder is produced, 42Å long and 24Å in diameter,^[8] creating what is referred to as a “β-can” formation.^[9] The short helical segments at either end of the cylinder form “caps” to further protect the interior of the β-barrel.^[9] Overall stability is maintained by this β-can structure, helping to resist unfolding from heat and other denaturants.^[2]

One can be found running through the central axis of the β-barrel,^[4] roughly to the symmetry axis of the barrel.^[8] This helix is extremely important as it contains the fluorophore responsible for fluorescence.^[2][4] This α-helix in particular is highly stabilized by the many that are made with each strand of the barrel.^[10]

.

The Chromophore

The () of GFP is located at the center of the β-barrel with a wild-type excitation peak of 395 nm, and a minor peak at 475 nm (about three times less intense^[3]) ^{[2][11][8][9]} with extinction coefficients of approximately 30,000 and 7,000 M^-1 cm^-1, respectively.^[2][9] Interestingly, the Aequorea victoria jellyfish utilizes the smaller of the two excitation peaks as pure aequorin emits a light of 470 nm.^[3] The relative amplitudes of these two excitation peaks can vary depending on environmental factors and previous illumination.^[8] For example, continued excitation leads to a diminution of the 395 nm excitation peak with a reciprocal amplification of the 475 nm peak.^[9] Regardless of absorption, the chromophore of GFP emits light of 508 nm.^{[2][11][8][9]}

Three amino residues in the central α-helix constitute the fluorophore of GFP: Ser⁶⁵Tyr⁶⁶Gly⁶⁷ (see below). Tsien et al. discovered that this tri-peptide sequence is post-translationally modified by internal cyclization and oxidation^[4] to produce a structure.^[2] Studies with E. coli proposed a sequential mechanism for the formation of the fluorophore that was initiated by a rapid cyclization between Ser⁶⁵ and Gly⁶⁷ to form an imidazolin-5-one intermediate.^[2] This rapid cyclization is carried out via nucleophilic attack of the amino group from Gly⁶⁷ on the carbonyl group of Ser⁶⁵ to form a five-membered ring. The loss of water then forms the imidazolin-5-one intermediate.^[11] Cyclization is succeeded by a much slower rate-limiting oxygenation of the Tyr⁶⁶ hydroxybenzyl side chain by atmospheric oxygen (No fluorescence was seen in anaerobically grown E. coli.), resulting in the 4-(p-hydroxybenzylidene)-imidazolidin-5-one stucture.^[2][11][9] The double bond that results from this series of reactions results in the linkage of the two π-systems of the rings, forming a larger conjugated system essential for fluorophore stability. ^[12]

The process is completely auto-catalytic such that there are no known co-factors or enzymatic components required.^[2] Despite the stability of the final product, while the chromophore is forming, the environmental temperature cannot drop below 30°C or the yield of viable GFP will decrease substantially.^[2][9] This, of course, is not an issue for the protein in nature as the jellyfish is unlikely to encounter waters of this degree in the Pacific Northwest.^[3] Such a temperature sensitivity is only relevant during formation as the stability of the final product is maintained through a network of close contacts surrounding the fluorophore.^[2] This, however, can and has been used in pulse-chase experiments in which the GFP-expressing cells are exposed to varying temperatures in place of labeled vs. unlabeled trials.^[3]

As the central α-helix is not located directly in the center of the β-barrel, cavities of differing area exist on either side of the chromophore. The larger cavity, consisting of about 135 Å,^[8] does not open out to the bulk solvent, but rather houses .^[8][13] Had this space not been occupied, it would be expected to considerably destabilize the protein as a whole. The hydrogen bonding created by the presence of the water molecules, however, helps to link the buried of Glu²²² and Gln⁶⁹ that would otherwise be actively polar.^[8] Therefore, the water molecules are extremely important in establishing a hydrogen bonding network about the chromophor.^[14]

The opposite side of the chromophore, however, is within close proximity of several aromatic and polar side chains. Several between the surrounding residues and the chromophore are present including: hydrogen bonds of His¹⁴⁸, Thr²⁰³, and Ser²⁰⁵ with the phenolic hydroxyl of Tyr⁶⁶; Arg⁹⁶ and Gln⁹⁴ with the carbonyl of the imidazolidinone ring; and hydrogen bonds of Glu²²² with the side chain of Thr⁶⁵. Additional hydrogen bonding in the area around the chromophore helps to stabilize Arg⁹⁶ in the protonated form, which suggests the presence of a partial negative charge on the carbonyl oxygen of the imidazolidinone ring in the deprotonated fluorophore.^[8] Arg⁹⁶ and Gln⁹⁴ in turn help to steady the imidazolidone.^[2] Therefore, it is thought that Arg⁹⁶ is essential for the formation of the fluorophore by catalyzing the initial ring closure.^[8] Tyr¹⁴⁵ provides a stabilizing

^[15] with the benzyl ring of the chromophore.^[8] The stability provided by the internal polar interactions are further augmented by the surrounding β-barrel.

The β-barrel provides a highly constrained environment that protects the chromophore from the bulk solvent,^[4] nearly creating the atmosphere of a vacuum.^[14] This is most likely responsible for the small Stoke’s shift, or the small wavelength difference between excitation and emission.^[8]

.

Findings show that fluorescence will not occur from a naked chromophore, but rather requires the protection of the β-can structure.^[11] However, in crystallum GFP will exhibit a nearly identical fluorescence spectrum and lifetime when compared with aqueous GFP. These two elements point to a fluorescence that is not inherent to the isolated fluorophore,^[2][9] but rather from the auto-catalytic cyclization of the polypeptide sequence Ser⁶⁵Tyr⁶⁶Gly⁶⁷ and subsequent oxidation of Tyr⁶⁶.^[9] However, this sequence is found in many proteins - why does GFP fluoresce? According to Phillips (1997), fluorophore formation is due to the close proximity of the backbone atoms between Ser⁶⁵. and Gly⁶⁷ gained through a lack of sterical hindrance by the hydrogen atom side chain of glycine. In fact, no functional fluorescent proteins have been found in which any other amino acid other than glycine was found at position 67. Even so, there are still proteins that have this specific sequence, therefore, there must be another inherent property to GFP that is still left misunderstood.^[9]

This quandary led Phillips to study the acid/base chemistry catalyzing the initial cyclization of the chromophore. He found that Arg⁹⁶ actually acts as a

by withdrawing electrons through hydrogen bonding with the carbonyl oxygen of Ser⁶⁵ to activate the carbonyl carbon for nucleophilic attack by the amide nitrogen of Gly⁶⁷. This mechanism was further supported by ab initio calculations, as well as database searches of similar compounds and protein sequences. Through acid/base chemistry, the chromophore is stabilized by resonance.^[9] Femtosecond Raman spectroscopy has been used to map the alteration of the structure close the chromophore during excited-state protein transfer and shown that chromophore wagging is orchestrated by the protein environment.^[16]

Mutant Studies

Many mutant green fluorescent proteins have been developed in order to further understand the structure and mechanism of the fluorophore. The first mutagenesis studies simply of the amino acid sequence (. NOTE: The structure represented here is already truncated at the carbonyl terminus). Shortening the polypeptide by more than seven amino acids from either terminus lead to a total loss of fluorescence, as well as a complete failure to absorb light at the traditional wavelengths. This is most likely due to the structure of the protein. The last seven amino acid residues of the carboxyl terminus are roughly disordered, and thus do not interfere with the overall structure. After seven residues, however, the capping α-helix structure is disrupted, leading to an unstable or unformed chromophore. The is less understood, but the same principle still applies even though the β-barrel does not begin until residue ten or eleven.^[2]

Point mutations have also been extensively studied in order to examine their effects on the chromophore. In general, most point mutations lead to a diminished excitation, especially in regions of the sequence adjacent to the fluorophore or those that interact with the fluorophore. An exception to this trend is the Ser⁶⁵Thr⁶⁶ mutant (normal Ser⁶⁵Tyr⁶⁶), which actually increases fluorescence intensity, although the reason is unclear.^[2]

An interesting mutation discovered by Ormo et al. (1996) was the Thr⁶⁵Tyr⁶⁶Gly⁶⁷ mutant, which produces an α-helical conformation in the chromophore opposed to the normal conformation, which is nearly perpendicular to the helical axis, due to its interaction with Arg⁹⁶. This further supports the idea that Arg⁹⁶ is an important factor in the structural arrangement required for cyclization, perhaps by promoting the attack of Gly⁶⁷ on the carbonyl carbon of Thr⁶⁵.^[8]

In high protein concentrations, GFP has been found to dimerize under the influence of high ionic strength between the two monomers. In Aequorea victoria, the aequorin is able to bind to the (1gfl), but not the monomer. Therefore, dimerization is a very important structural feature in terms of its function, as it also assists the GFP to absorb energy at the excitation wavelength of aequorin even though GFP has only a “modest” extinction coefficient. As a result, dimers, and often even higher (1w7s), are predominant protein populations within the jellyfish.^[11]

.

Using GFP as a Research Tool

A description of some of the ways GFP is being used as a tool in research is at Green_Fluorescent_Protein:_Research_Tool.