Light-activated reassembly of split green fluorescent protein

Truncated green fluorescent protein (GFP) with the 11th β-strand removed is potentially interesting for bioconjugation, imaging, and the preparation of semisynthetic proteins with novel spectroscopic or functional properties. Surprisingly, the truncated GFP generated by removing the 11th strand, once refolded, does not reassemble with a synthetic peptide corresponding to strand 11 but does reassemble following light activation. The mechanism of this process has been studied in detail by absorption, fluorescence, and Raman spectroscopy. The chromophore in this refolded truncated GFP is found to be in the trans configuration. Upon exposure to light a photostationary state is formed between the trans and cis conformations of the chromophore, and only truncated GFP with the cis configuration of the chromophore binds the peptide. A kinetic model describing the light-activated reassembly of this split GFP is discussed. This unique light-driven reassembly is potentially useful for controlling protein-protein interactions.     

Detecting protein-protein interactions with a green fluorescent protein fragment reassembly trap: scope and mechanism

Identification of protein binding partners is one of the key challenges of proteomics. We recently introduced a screen for detecting protein-protein interactions based on reassembly of dissected fragments of green fluorescent protein fused to interacting peptides. Here, we present a set of comaintained Escherichia coli plasmids for the facile subcloning of fusions to the green fluorescent protein fragments. Using a library of antiparallel leucine zippers, we have shown that the screen can detect very weak interactions (K(D) approximately 1 mM). In vitro kinetics show that the reassembly reaction is essentially irreversible, suggesting that the screen may be useful for detecting transient interactions. Finally, we used the screen to discriminate cognate from noncognate protein-ligand interactions for tetratricopeptide repeat domains. These experiments demonstrate the general utility of the screen for larger proteins and elucidate mechanistic details to guide the further use of this screen in proteomic analysis. Additionally, this work gives insight into the positional inequivalence of stabilizing interactions in antiparallel coiled coils.     

Deconstructing green fluorescent protein

Green fluorescent protein (GFP) has been reassembled from two pieces, a large fragment 214 amino acids in length that is produced recombinantly (GFP 1-10) and a short synthetic peptide corresponding to the 11th stave of the beta-barrel that is 16 amino acids long (synthetic GFP 11), following a system developed by Waldo and co-workers (Cabantous, S.; et al. Nat. Biotechnol. 2005, 23, 102-7) as an in vivo probe for protein association and folding. We demonstrate that the reassembled protein has identical absorption and excited-state proton transfer dynamics as a whole protein of the identical sequence. We show that the reassembled protein can be taken apart and the peptide replaced with a different synthetic peptide designed to perturb the chromophore absorption. Thus, this semisynthetic reassembly process offers a general route for studying the assembly of the beta-barrel as well as the introduction of unnatural amino acids.     

Fluorophores related to the green fluorescent protein

Imidazolin-5-one derivatives and isosteres (oxazolinones, butenolides, and pyrrolinones) of the 4-hydroxybenzylidene-imidazolinone chromophore of the GFP have been synthesized and their photophysical properties have been investigated.     

Light-driven decarboxylation of wild-type green fluorescent protein

The response of wild-type GFP to UV and visible light was investigated using steady state absorption, fluorescence, and Raman spectroscopies. As reported previously [van Thor, Nat. Struct. Biol. 2002, 9, 37-41], irradiation of GFP results in decarboxylation of E222. Here it is reported that the rate of the light-driven decarboxylation reaction strongly depends on the excitation wavelength, decreasing in the order 254 nm > 280 nm > 476 nm. The relative efficiencies of decarboxylation are explained in terms of the Kolbe-type mechanism in which the excited state of the chromophore acts as an oxidant by accepting an electron from E222. Specifically, it is proposed that 254 nm excitation populates the S2 (or higher) excited state of the chromophore, whereas 404 and 476 nm excitation populate the S1 excited state of neutral and anionic forms, respectively, and that the relative oxidizing power of the three excited states controls the rate of the decarboxylation reaction. In addition, the role of W57 in the photophysics of GFP has been probed by mutating this residue to phenylalanine. These studies reveal that while W57 does not affect decarboxylation, this residue is involved in resonance energy transfer with the chromophore, thereby partially explaining the green fluorescence observed upon UV irradiation of wild-type GFP. Finally, comparison of Raman spectra obtained from nonilluminated and decarboxylated forms of wild-type GFP has provided further vibrational band assignments for neutral and anionic forms of the chromophore within the protein. In addition, these spectra provide valuable insight into the specific interactions between the protein and the chromophore that control the optical properties of wild-type GFP.     

The role of the protein matrix in green fluorescent protein fluorescence

In the ground state of the highly conjugated green fluorescent protein (GFP), the chromophore should be planar. However, numerous crystal structures of GFP and GFP-like proteins have been reported with slightly twisted chromophores. We have previously shown that the protein cavity surrounding the chromophore in wild-type GFP is not complementary with a planar chromophore. This study shows that the crystal structure of wild-type GFP is not an anomaly: most of the GFP and GFP-like proteins in the protein databank have a protein matrix that is not complementary with a planar chromophore. When the pi-conjugation across the ethylenic bridge of the chromophore is removed the protein matrix will significantly twist the freely rotating chromophore from the relatively planar structures found in the crystal structures. The possible consequences of this nonplanar deformation on the photophysics of GFP are discussed. A volume analysis of the cis-trans-isomerization of HBDI, a GFP chromophore model compound, reveals that its hula-twist motion is volume conserving. This means that, if the GFP chromophore or GFP chromophore model compounds undergo a cis-trans-isomerization in a volume-constricting medium, such as a protein matrix or viscous liquid, it will probably isomerize by means of a HT-type motion.     

Excited-state structure determination of the green fluorescent protein chromophore

Ultrafast polarization-sensitive infrared (IR) spectroscopy of the C=O stretching mode of the chromophore of the green fluorescent protein reveals a near complete twisting around the ethylenic bridge between the phenolate and imidazolidinone groups upon electronic excitation, hinting at a decisive role of this motion in the efficient internal conversion process.     

Structural events in the photocycle of green fluorescent protein

Picosecond time-resolved mid-infrared absorption changes of the wild type green fluorescent protein from Aequorea victoria are reported on structural events during the photocycle. Concomitant with rapid H/D transfer following excitation of the neutral A state at 400 nm, a transient signal at 1721/1711 cm(-1) (H/D) developed belonging to protonated glutamate 222, which was definitively assigned using the E222D mutant from the altered proton-transfer kinetics to aspartate in addition to the altered band position and intensity in the spectra. A transient at 1697 cm(-1), assigned to a structural perturbation of glutamine 69, had a H/D kinetic isotope effect of >32, showing the conformational dynamics to be sensitive to the active site H/D vibrations. The kinetic data up to 2 ns after excitation in the 1250-1800 cm(-1) region in D2O provided 10 and 75 ps time constants for the excited-state deuteron transfer and the associated A1* - A1 and A2* - A2 difference spectra and showed the radiative intermediate I state vibrations and the transient accumulation of the long-lived ground-state intermediate I2. Assignments of chromophore modes for the A1, A2, and I2 ground states are proposed on the basis of published model compound studies (Esposito, A. P.; Schellenberg, P.; Parson, W. W.; Reid, P. J. J. Mol. Struct. 2001, 569, 25 and He, X.; Bell, A. F.; Tonge, P. J. J. Phys. Chem. B 2002, 106, 6056). Tentative assignments for the singlet-state intermediates A1*, A2*, and I* are discussed. An unexpected and unassigned band that may be a C=C chromophore vibration was observed in the ground state (1665 cm(-1)) as well as in all photocycle intermediates. Optical dumping of the transient I population was achieved using an additional 532 nm pulse and the directly obtained I2 - I* difference spectrum was highly similar to the I2 - I* photocycle spectrum. The pump-dump-probe spectrum differed from the pump-probe photocycle difference spectrum with respect to the intensity of the phenol 1 mode at 1578 cm(-1), suggesting stronger delocalization of the negative charge onto the phenolic ring of the anionic chromophore in the dumped I2 state. Indication for structural heterogeneity of the chromophore, Glu 222, and the chromophore environment was found in the two parallel proton-transfer reactions and their distinct associated ground- and intermediate-state vibrations. Vibrational spectral markers at 1695 cm(-1) assigned to Gln 69, at 1631 cm(-1) belonging to a C=C mode, and at 1354 cm(-1) belonging to a phenolate vibration further indicated the I2 and I* states to be unrelaxed.     

Fluorescence resonance energy transfer between green fluorescent protein and doxorubicin enabled by DNA nanotechnology

DNA nanotechnology is a rapidly growing research area, where DNA may be used for wide range of applications such as construction of nanodevices serving for large scale of diverse purposes. Likewise a panel of various purified fluorescent proteins is investigated for their ability to emit their typical fluorescence spectra under influence of particular excitation. Hence these proteins may form ideal donor molecules for assembly of fluorescence resonance emission transfer (FRET) constructions. To extend the application possibilities of fluorescent proteins, while using DNA nanotechnology, we developed nanoconstruction comprising green fluorescent protein (GFP) bound onto surface of surface active nanomaghemite and functionalized with gold nanoparticles. We took advantage of natural affinity between gold and thiol moieties, which were modified to bind DNA fragment. Finally we enclosed doxorubicin into fullerene cages. Doxorubicin intercalated in DNA fragment bound on the particles and thus we were able to connect these parts together. Because GFP behaved as a donor and doxorubicin as an acceptor using excitation wavelength for GFP (395 nm) in emission wavelength of doxorubicin (590 nm) FRET was observed. This nanoconstruction may serve as a double‐labeled transporter of doxorubicin guided by force of external magnetic force owing to the presence of nanomaghemite. Further nanomaghemite offers the possibility of using this technology for thermotherapy.     

Unnatural amino acid mutagenesis of green fluorescent protein

Unnatural amino acid mutagenesis has been used to selectively substitute tyrosine 66 of green fluorescent protein (GFP) with five novel amino acids: p-amino-L-phenylalanine, p-methoxy-L-phenylalanine, p-iodo-L-phenylalanine, p-bromo-L-phenylalanine, and L-3-(2-naphthyl)alanine. The absorbance and emission maxima of the resulting mutant GFPs span the range from 375 to 435 nm and 428 to 498 nm, respectively. The spectral properties of the mutant GFPs, including the absorbance and fluorescence maxima and quantum yields, correlate with the structural and electronic properties of the substituents on the amino acids.     

Bond selection in the photoisomerization reaction of anionic green fluorescent protein and kindling fluorescent protein chromophore models

The chromophores of the most widely known fluorescent proteins (FPs) are derivatives of a core p-hydroxybenzylidene-imidazolinon-5-one (HBI) motif, which usually occurs as a phenolate anion. Double bond photoisomerization of the exocyclic bridge of HBI is widely held to be an important internal conversion mechanism for FP chromophores. Herein we describe the ground and excited-state electronic structures and potential energy surfaces of two model chromophores: 4- p-hydroxybenzylidiene-1,2-dimethyl-imidazolin-5-one anion (HBDI), representing green FPs (GFPs), and 2-acetyl-4-hydroxybenylidene-1-methyl-imidazolin-5-one anion (AHBMI), representing kindling FPs (KFPs). These chromophores differ by a single substitution, but we observe qualitative differences in the potential energy surfaces which indicate inversion of bond selection in the photoisomerization reaction. Bond selection is also modulated by whether the reaction proceeds from a Z or an E conformation. These configurations correspond to fluorescent and nonfluorescent states of structurally characterized FPs, including some which can be reversibly switched by specific illumination regimes. We explain the difference in bond selectivity via substituent stabilization effects on a common set of charge-localized chemical structures. Different combinations of these structures give rise to both optically active (planar) and twisted intramolecular charge-transfer (TICT) states of the molecules. We offer a prediction of the gas-phase absorption of AHBMI, which has not yet been measured. We offer a hypothesis to explain the unusual fluorescence of AHBMI in DMF solution, as well as an experimental proposal to test our hypothesis.     

Structural chemistry of a green fluorescent protein Zn biosensor

We designed a green fluorescent protein mutant (BFPms1) that preferentially binds Zn(II) (enhancing fluorescence intensity) and Cu(II) (quenching fluorescence) directly to a chromophore ligand that resembles a dipyrrole unit of a porphyrin. Crystallographic structure determination of apo, Zn(II)-bound, and Cu(II)-bound BFPms1 to better than 1.5 A resolution allowed us to refine metal centers without geometric restraints, to calculate experimental standard uncertainty errors for bond lengths and angles, and to model thermal displacement parameters anisotropically. The BFPms1 Zn(II) site (KD = 50 muM) displays distorted trigonal bipyrimidal geometry, with Zn(II) binding to Glu222, to a water molecule, and tridentate to the chromophore ligand. In contrast, the BFPms1 Cu(II) site (KD = 24 muM) exhibits square planar geometry similar to metalated porphyrins, with Cu(II) binding to the chromophore chelate and Glu222. The apo structure reveals a large electropositive region near the designed metal insertion channel, suggesting a basis for the measured metal cation binding kinetics. The preorganized tridentate ligand is accommodated in both coordination geometries by a 0.4 A difference between the Zn and Cu positions and by distinct rearrangements of Glu222. The highly accurate metal ligand bond lengths reveal different protonation states for the same oxygen bound to Zn vs Cu, with implications for the observed metal ion specificity. Crystallographic anisotropic thermal factor analysis validates metal ion rigidification of the chromophore in enhancement of fluorescence intensity upon Zn(II) binding. Thus, our high-resolution structures reveal how structure-based design has effectively linked selective metal binding to changes in fluorescent properties. Furthermore, this protein Zn(II) biosensor provides a prototype suitable for further optimization by directed evolution to generate metalloprotein variants with desirable physical or biochemical properties.     

Continuous separation of green fluorescent protein by annular chromatography

The concept of annular chromatography was tested by separation of a real protein solution used in biotechnology. Green fluorescent protein was expressed in S. cerevisiae and the extract was continuously separated by a pressurized annular chromatograph packed with a Superdex 200 prep grade size-exclusion chromatography medium. Purity was checked by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, Western blotting and fluorescent intensity. The continuous mode was compared to batchwise operation. Under the assumption that equivalent fractions are collected, both modes are comparable.     

Hidden electronic excited state of enhanced green fluorescent protein

Precise two-photon absorption spectra of the green fluorescent protein (GFP) and the mutants sapphire-GFP (T203I) and enhanced GFP (S65T/F64L), as well as a model compound for the chromophore, 4'-hydroxybenzylidene-2,3-dimethylimidazolinone (HBDI) were measured by multiplex two-photon absorption spectroscopy. The observed TPA bands of the anionic forms of enhanced GFP and HBDI were significantly shifted to the higher energy compared with the lowest-energy bands in one-photon absorption spectra. This result indicated the existence of a hidden electronic excited state in the vicinity of the lowest excited singlet (S1) state of the anionic form of the GFP chromophore, which is the origin of the blue shift of the two-photon absorption spectra as well as two-photon fluorescence excitation spectra.     

Purification of recombinant green fluorescent protein by three-phase partitioning

The technique of three-phase partitioning (TPP) was used to purify the green fluorescent protein (GFP) in a single step. TPP uses a combination of ammonium sulphate and tert-butanol to precipitate proteins from their crude extracts. In the first round of TPP with 20% ammonium sulphate saturation at the ratio of crude to tert-butanol 1:1 (v/v), most of the GFP remains in the lower aqueous phase. When subjected to a second round of TPP with 60% ammonium sulphate saturation at the ratio of crude to tert-butanol 1:2 (v/v) gives 78% recovery of GFP with a 20-fold purification. The sodium dodecyl sulphate-polyacrylamide gel electrophoretic (SDS-PAGE) analysis of purified preparation shows single band. The fluorescence excitation and emission spectra agreed with values reported in literature.     

Reaction progress of chromophore biogenesis in green fluorescent protein

The mature form of green fluorescent protein (GFP) is generated by a spontaneous self-modification process that is essentially irreversible. A key step in chromophore biosynthesis involves slow air oxidation of an intermediate species, in which the backbone atoms of residues 65-67 have condensed to form a five-membered heterocycle. We have investigated the kinetics of hydrogen peroxide evolution during in vitro GFP maturation and found that the H2O2 coproduct is generated prior to the acquisition of green fluorescence at a stoichiometry of 1:1 (peroxide/chromophore). The experimental progress curves were computer-fitted to a three-step mechanism, in which the first step proceeds with a time constant of 1.5 (+/-1.1) min and includes protein folding and peptide cyclization. Kinetic data obtained by HPLC analysis support a rapid cyclization reaction that can be reversed upon acid denaturation. The second step proceeds with a time constant of 34.0 (+/-1.5) min and entails rate-limiting protein oxidation, as supported by a mass loss of 2 Da observed for tryptic peptides derived from species that accumulate during the reaction. The final step in GFP maturation proceeds with a time constant of 10.6 (+/-1.2) min, suggesting that this step may contribute to overall rate retardation. We propose that under highly aerobic conditions, the dominant reaction path follows a cyclization-oxidation-dehydration mechanism, in which dehydration of the heterocycle is facilitated by slow proton abstraction from the Tyr66 beta-carbon. In combination, the results presented here suggest a role for molecular oxygen in trapping the cyclized form of GFP.     

Labeling effects on the isoelectric point of green fluorescent protein.

We studied the effects of fluorescent labeling on the isoelectric points (pI values) of proteins using capillary isoelectric focusing with laser-induced fluorescence detection (cIEF-LIF). Specifically, we labeled green fluorescent protein (GFP) from the jellyfish Aequorea victoria with the fluorogenic dye 3-(2-furoyl)quinoline-2-carboxaldehyde (FQ). cIEF-LIF was used to monitor the native fluorescence of GFP and showed pI changes in GFP's FQ-labeled products. Multiple labeling of GFP with FQ produced a series of products with pI values shifted towards a low pH. We verified cIEF-LIF results with traditional slab gel IEF. Our cIEF-LIF technique can routinely detect 10(-11) M of FQ-labeled protein, whereas traditional slab gel IEF with silver stain detection gives detection limits of 10(-7) M in the same samples.     

Slow exchange in the chromophore of a green fluorescent protein variant

Green fluorescent protein and its mutants have become valuable tools in molecular biology. They also provide systems rich in photophysical and photochemical phenomena of which an understanding is important for the development of new and optimized variants of GFP. Surprisingly, not a single NMR study has been reported on GFPs until now, possibly because of their high tendency to aggregate. Here, we report the (19)F nuclear magnetic resonance (NMR) studies on mutants of the green fluorescent protein (GFP) and cyan fluorescent protein (CFP) labeled with fluorinated tryptophans that enabled the detection of slow molecular motions in these proteins. The concerted use of dynamic NMR and (19)F relaxation measurements, supported by temperature, concentration- and folding-dependent experiments provides direct evidence for the existence of a slow exchange process between two different conformational states of CFP. (19)F NMR relaxation and line shape analysis indicate that the time scale of exchange between these states is in the range of 1.2-1.4 ms. Thermodynamic analysis revealed a difference in enthalpy (Delta)H(0) = (18.2 +/- 3.8) kJ/mol and entropy T(Delta)S(0) = (19.6 +/- 1.2) kJ/mol at T = 303 K for the two states involved in the exchange process, indicating an entropy-enthalpy compensation. The free energy of activation was estimated to be approximately 60 kJ/mol. Exchange between two conformations, either of the chromophore itself or more likely of the closely related histidine 148, is suggested to be the structural process underlying the conformational mobility of GFPs. The possibility to generate a series of single-atom exchanges ("atomic mutations") like H --> F in this study offers a useful approach for characterizing and quantifying dynamic processes in proteins by NMR.