Understanding GFP chromophore biosynthesis: controlling backbone cyclization and modifying post-translational chemistry

The Aequorea victoria green fluorescent protein (GFP) undergoes a remarkable post-translational modification to create a chromophore out of its component amino acids S65, Y66, and G67. Here, we describe mutational experiments in GFP designed to convert this chromophore into a 4-methylidene-imidazole-5-one (MIO) moiety similar to the post-translational active-site electrophile of histidine ammonia lyase (HAL). Crystallographic structures of GFP variant S65A Y66S (GFPhal) and of four additional related site-directed mutants reveal an aromatic MIO moiety and mechanistic details of GFP chromophore formation and MIO biosynthesis. Specifically, the GFP scaffold promotes backbone cyclization by (1) favoring nucleophilic attack by close proximity alignment of the G67 amide lone pair with the pi orbital of the residue 65 carbonyl and (2) removing enthalpic barriers by eliminating inhibitory main-chain hydrogen bonds in the precursor state. GFP R96 appears to induce structural rearrangements important in aligning the molecular orbitals for ring cyclization, favor G67 nitrogen deprotonation through electrostatic interactions with the Y66 carbonyl, and stabilize the reduced enolate intermediate. Our structures and analysis also highlight negative design features of the wild-type GFP architecture, which favor chromophore formation by destabilizing alternative conformations of the chromophore tripeptide. By providing a molecular basis for understanding and controlling the driving force and protein chemistry of chromophore creation, this research has implications for expansion of the genetic code through engineering of modified amino acids.     

Protein Evolution via Amino Acid and Codon Elimination

Background

Global residue-specific amino acid mutagenesis can provide important biological insight and generate proteins with altered properties, but at the risk of protein misfolding. Further, targeted libraries are usually restricted to a handful of amino acids because there is an exponential correlation between the number of residues randomized and the size of the resulting ensemble. Using GFP as the model protein, we present a strategy, termed protein evolution via amino acid and codon elimination, through which simplified, native-like polypeptides encoded by a reduced genetic code were obtained via screening of reduced-size ensembles.

Methodology/Principal Findings

The strategy involves combining a sequential mutagenesis scheme to reduce library size with structurally stabilizing mutations, chaperone complementation, and reduced temperature of gene expression. In six steps, we eliminated a common buried residue, Phe, from the green fluorescent protein (GFP), while retaining activity. A GFP variant containing 11 Phe residues was used as starting scaffold to generate 10 separate variants in which each Phe was replaced individually (in one construct two adjacent Phe residues were changed simultaneously), while retaining varying levels of activity. Combination of these substitutions to generate a Phe-free variant of GFP abolished fluorescence. Combinatorial re-introduction of five Phe residues, based on the activities of the respective single amino acid replacements, was sufficient to restore GFP activity. Successive rounds of mutagenesis generated active GFP variants containing, three, two, and zero Phe residues. These GFPs all displayed progenitor-like fluorescence spectra, temperature-sensitive folding, a reduced structural stability and, for the least stable variants, a reduced steady state abundance.

Conclusions/Significance

The results provide strategies for the design of novel GFP reporters. The described approach offers a means to enable engineering of active proteins that lack certain amino acids, a key step towards expanding the functional repertoire of uniquely labeled proteins in synthetic biology.     

Deciphering the factors responsible for the stability of a GFP variant resistant to alkaline pH using molecular dynamics simulations

Charged amino acids having ionizable side chains play crucial roles in maintaining the solubility and stability of a protein. These charged amino acids are mostly exposed on protein surface and participate in electrostatic interactions with neighboring charged amino acids as well as with solvent. Therefore, the change in the solvent pH affects the protein stability in most cases. Previously, we reported a GFP variant, GFP14R having 14 surface lysines replaced with arginines, that showed enhanced stability under alkaline pH. Here, we analyzed the factors that contribute to the stability of the GFP14R under alkaline pH quantitatively using molecular dynamics simulations. Protonation state of the charged amino acids of GFP14R and control GFP under neutral pH and alkaline pH were modeled, and molecular dynamics simulations were performed. This comparative analysis revealed that the GFP14R with more arginine frequency on the surface maintained the stability under both pH conditions without much change in their salt-bridge interactions as well as the hydrogen bond interactions with solvent. On the other hand, these interactions were significantly reduced for the control GFP under alkaline pH due to the deprotonated lysine side chains. These results suggest that the advantageous property of arginine over lysine can be considered one of the parameter for the protein stability engineering under alkaline pH conditions.     

A new approach to gene mutation analysis using "GFP-Display"

The unique behavior of green fluorescent protein (GFP) on SDS-PAGE was applied to the detection of a single amino acid substitution in GFP-tagged polypeptides. This simple detection method using SDS/urea gels was designated GFP-display. The N-terminal 18 or 37 amino acids of K-Ras was used as a model GFP-tagged polypeptide. K-ras exon 1 was fused to a gfp cDNA at each end and expressed in Escherichia coli. Amino acid number 12 of K-Ras (wild type; Gly) was changed to Ser, Arg, Cys, Asp, Ala, or Val, and the mobility shift of the greenish fluorescent bands in the SDS/urea gel was analyzed. These mutants were easily detected by GFP-display; however, detection depended strongly on the urea concentration and electrophoresis temperature. Subsequently, GFP-display was applied to the 36 amino acids encoding human p53 exon 7. Amino acid number 248 (wild type; Arg) was changed to Gly, Trp, Gln, Pro, or Leu, and similar mobility shifts were observed. GFP-display could be coupled with an in vitro translation system. Fluorescent active GFP and GFP-Ras fusion proteins were synthesized within a few hours. GFP-display shows potential as a modern approach to gene mutation analysis at the protein level, and is a useful method for protein engineering studies.     

Deletional Protein Engineering Based on Stable Fold

Diversification of protein sequence-structure space is a major concern in protein engineering. Deletion mutagenesis can generate a protein sequence-structure space different from substitution mutagenesis mediated space, but it has not been widely used in protein engineering compared to substitution mutagenesis, because it causes a relatively huge range of structural perturbations of target proteins which often inactivates the proteins. In this study, we demonstrate that, using green fluorescent protein (GFP) as a model system, the drawback of the deletional protein engineering can be overcome by employing the protein structure with high stability. The systematic dissection of N-terminal, C-terminal and internal sequences of GFPs with two different stabilities showed that GFP with high stability (s-GFP), was more tolerant to the elimination of amino acids compared to a GFP with normal stability (n-GFP). The deletion studies of s-GFP enabled us to achieve three interesting variants viz. s-DL4, s-N14, and s-C225, which could not been obtained from n-GFP. The deletion of 191–196 loop sequences led to the variant s-DL4 that was expressed predominantly as insoluble form but mostly active. The s-N14 and s-C225 are the variants without the amino acid residues involving secondary structures around N- and C-terminals of GFP fold respectively, exhibiting comparable biophysical properties of the n-GFP. Structural analysis of the variants through computational modeling study gave a few structural insights that can explain the spectral properties of the variants. Our study suggests that the protein sequence-structure space of deletion mutants can be more efficiently explored by employing the protein structure with higher stability.     

Enhanced fluorescence resonance energy transfer between spectral variants of green fluorescent protein through zinc-site engineering

Although spectral variants of GFP should in theory be suited for fluorescence resonance energy transfer (FRET) and therefore suited for studies of protein-protein interactions, the unfavorable location of the fluorophore 15 A deep inside the GFP molecule has especially impaired this application. Here, metal-ion site engineering around the dimerization interface known from the X-ray structure of GFP is applied to the cyan and the yellow spectral variant of GFP to stabilize the heterodimeric form of these molecules and thereby increase FRET signaling. The FRET signal, determined as the ratio between the maximal emission for the yellow variant, 530 nm, and the cyan variant, 475 nm, during excitation of the cyan variant at 433 nm was increased up to 8-10-fold in the presence of 10(-4) M ZnCl2 by engineering of two symmetric metal-ion sites being either bidentate or tridentate. A similar increase in FRET signaling was however obtained in a pair of molecules in which a single bidentate metal-ion site was generated by introducing a zinc-binding residue in each of the two spectral variants of GFP and therefore creating an obligate heterodimeric pair. It is concluded that FRET signaling between spectral variants of GFP can be increased by stabilizing dimer formation and especially by favoring heterodimer formation in this case performed by metal-ion site engineering.     

Red fluorescent protein DsRed from Discosoma sp. as a reporter protein in higher plants

GFP from Aequorea victoria is a standard genetic marker widely used to visualize cellular events in a noninvasive manner. For simultaneous imaging of different processes, in vivo mutants of GFP with shifted wavelength spectra (e.g., blue fluorescent protein) are conventionally used. The recently reported red fluorescent protein from Discosoma sp., DsRed, represents a new marker that can be used together with GFP variants for multicolor imaging. DsRed is an interesting marker protein for use in plants because of its red-shifted wavelength spectrum that will avoid damaging cells and tissues by excitation light. In this report, we show that DsRed is an excellent marker in higher plants in spite of the interfering red autofluorescence of chlorophyll, which can be eliminated by using the appropriate filter sets. Transient expression of DsRed1-C1 and a soluble-modified, red-shifted GFP variant has been carried out both individually and jointly in the epidermal cells of three different Nicotiana species and Chenopodium quinoa, which gives rise to dual labeling in plants. For this purpose, a human codon-optimized variant of DsRed has been adopted for expression in plants. Moreover, the DsRed reporter gene was expressed by using a labeled plant viral vector derived from an infectious full-length clone of potato virus X.     

Misacylation of yeast amber suppressor tRNA(Tyr) by E. coli lysyl-tRNA synthetase and its effective repression by genetic engineering of the tRNA sequence

Through an exhaustive search for Escherichia coli aminoacyl-tRNA synthetase(s) responsible for the misacylation of yeast suppressor tRNA(Tyr), E. coli lysyl-tRNA synthetase was found to have a weak activity to aminoacylate yeast amber suppressor tRNA(Tyr) (CUA) with L-lysine. Since our protein-synthesizing system for site-specific incorporation of unnatural amino acids into proteins is based on the use of yeast suppressor tRNA(Tyr)/tyrosyl-tRNA synthetase (TyrRS) pair as the "carrier" of unusual amino acid in E. coli translation system, this misacylation must be repressed as low as possible. We have succeeded in effectively repressing the misacylation by changing several nucleotides in this tRNA by genetic engineering. This "optimized" tRNA together with our mutant TyrRS should serve as an efficient and faithful tool for site-specific incorporation of unnatural amino acids into proteins in a protein-synthesizing system in vitro or in vivo.     

YjeH Is a Novel Exporter of l-Methionine and Branched-Chain Amino Acids in Escherichia coli

Amino acid efflux transport systems have important physiological functions and play vital roles in the fermentative production of amino acids. However, no methionine exporter has yet been identified in Escherichia coli. In this study, we identified a novel amino acid exporter, YjeH, in E. coli. The yjeH overexpression strain exhibited high tolerance to the structural analogues of l-methionine and branched-chain amino acids, decreased intracellular amino acid levels, and enhanced export rates in the presence of a Met-Met, Leu-Leu, Ile-Ile, or Val-Val dipeptide, suggesting that YjeH functions as an exporter of l-methionine and the three branched-chain amino acids. The export of the four amino acids in the yjeH overexpression strain was competitively inhibited in relation to each other. The expression of yjeH was strongly induced by increasing cytoplasmic concentrations of substrate amino acids. Green fluorescent protein (GFP)-tagged YjeH was visualized by total internal reflection fluorescence microscopy to confirm the plasma membrane localization of YjeH. Phylogenetic analysis of transporters indicated that YjeH belongs to the amino acid efflux family of the amino acid/polyamine/organocation (APC) superfamily. Structural modeling revealed that YjeH has the typical “5 + 5” transmembrane α-helical segment (TMS) inverted-repeat fold of APC superfamily transporters, and its binding sites are strictly conserved. The enhanced capacity of l-methionine export by the overexpression of yjeH in an l-methionine-producing strain resulted in a 70% improvement in titer. This study supplements the transporter classification and provides a substantial basis for the application of the methionine exporter in metabolic engineering.     

Evaluation and biosynthetic incorporation of chlorotyrosine into recombinant proteins

Recently, non-canonical amino acids (NCAA) incorporation was developed to enhance the functional properties of proteins. Incorporation of NCAA containing chlorine atom is conceptually an attractive approach to prepare pharmacologically active substances, which is a difficult task since chlorine is bulky atom. In this study, we evaluated the efficiency and extent of in vivo incorporation of tyrosine analogue 3-chlorotyrosine [(3-Cl)Tyr] into the recombinant proteins GFP and GFPHS (highly stable GFP). The incorporation of (3-Cl)Tyr into GFP leads to dramatic reduction in the expression level of protein. On the other hand, the incorporation of (3-Cl)Tyr into GFPHS was expressed well as a soluble form. In addition we used bioinformatics tools for the analysis to explore the possible constraints in micro-environment of each natural amino acid residue to be replaced with chlorine atom accommodation into GFPHS. In conclusion, our approaches are reliable and straightforward way to enhance the translation of chlorinated amino acids into proteins.     

Efficient gene delivery and gene expression in zebrafish using the Sleeping Beauty transposon

We used the Tc1/mariner family transposable element Sleeping Beauty (SB) for transgenesis and long-term expression studies in the zebrafish (Danio rerio), a popular organism for clinical disease, vertebrate patterning, and cell biology applications. SB transposase enhanced the transgenesis and expression rate sixfold (from 5 to 31%) and more than doubled the total number of tagged chromosomes over standard, plasmid injection-based transgenesis methods. Molecular analysis of these loci demonstrated a precise integration of these elements into recipient chromosomes with genetic footprints diagnostic of transposition. GFP expression from transposase-mediated integrants was Mendelian through the eighth generation. A blue-shifted GFP variant (BFP) and a red fluorescent protein (DsRed) were also useful transgenesis markers, indicating that multiple reporters are practical for use with SB in zebrafish. We showed that SB is suitable for tissue-specific transgene applications using an abbreviated gamma-crystallin GFP cassette. Finally, we describe a general utility transposon vector for chromosomal engineering and molecular genetics experiments in zebrafish. Together, these data indicate that SB is an efficient tool for transgenesis and expression in zebrafish, and that the transposon will be useful for gene expression in cell biology applications as well as an insertional mutagen for gene discovery during development.     

A study on the effect of surface lysine to arginine mutagenesis on protein stability and structure using green fluorescent protein

Two positively charged basic amino acids, arginine and lysine, are mostly exposed to protein surface, and play important roles in protein stability by forming electrostatic interactions. In particular, the guanidinium group of arginine allows interactions in three possible directions, which enables arginine to form a larger number of electrostatic interactions compared to lysine. The higher pKa of the basic residue in arginine may also generate more stable ionic interactions than lysine. This paper reports an investigation whether the advantageous properties of arginine over lysine can be utilized to enhance protein stability. A variant of green fluorescent protein (GFP) was created by mutating the maximum possible number of lysine residues on the surface to arginines while retaining the activity. When the stability of the variant was examined under a range of denaturing conditions, the variant was relatively more stable compared to control GFP in the presence of chemical denaturants such as urea, alkaline pH and ionic detergents, but the thermal stability of the protein was not changed. The modeled structure of the variant indicated putative new salt bridges and hydrogen bond interactions that help improve the rigidity of the protein against different chemical denaturants. Structural analyses of the electrostatic interactions also confirmed that the geometric properties of the guanidinium group in arginine had such effects. On the other hand, the altered electrostatic interactions induced by the mutagenesis of surface lysines to arginines adversely affected protein folding, which decreased the productivity of the functional form of the variant. These results suggest that the surface lysine mutagenesis to arginines can be considered one of the parameters in protein stability engineering.     

Role of the alfalfa mosaic virus movement protein and coat protein in virus transport

The movement protein (MP) and coat protein (CP) encoded by Alfalfa mosaic virus (AMV) RNA 3 are both required for virus transport. RNA 3 vectors that expressed nonfused green fluorescent protein (GFP), MP:GPF fusions, or GFP:CP fusions were used to study the functioning of mutant MP and CP in protoplasts and plants. C-terminal deletions of up to 21 amino acids did not interfere with the function of the CP in cell-to-cell movement, although some of these mutations interfered with virion assembly. Deletion of the N-terminal 11 or C-terminal 45 amino acids did not interfere with the ability of MP to assemble into tubular structures on the protoplast surface. Additionally, N- or C-terminal deletions disrupted tubule formation. A GFP:CP fusion was targeted specifically into tubules consisting of a wild-type MP. All MP deletion mutants that showed cell-to-cell and systemic movement in plants were able to form tubular structures on the surface of protoplasts. Brome mosaic virus (BMV) MP did not support AMV transport. When the C-terminal 48 amino acids were replaced by the C-terminal 44 amino acids of the AMV MP, however, the BMV/AMV chimeric protein permitted wild-type levels of AMV transport. Apparently, the C terminus of the AMV MP, although dispensable for cell-to-cell movement, confers specificity to the transport process.     

Genetic incorporation of unnatural amino acids into proteins in mammalian cells

We developed a general approach that allows unnatural amino acids with diverse physicochemical and biological properties to be genetically encoded in mammalian cells. A mutant Escherichia coli aminoacyl-tRNA synthetase (aaRS) is first evolved in yeast to selectively aminoacylate its tRNA with the unnatural amino acid of interest. This mutant aaRS together with an amber suppressor tRNA from Bacillus stearothermophilus is then used to site-specifically incorporate the unnatural amino acid into a protein in mammalian cells in response to an amber nonsense codon. We independently incorporated six unnatural amino acids into GFP expressed in CHO cells with efficiencies up to 1 mug protein per 2 x 10(7) cells; mass spectrometry confirmed a high translational fidelity for the unnatural amino acid. This methodology should facilitate the introduction of biological probes into proteins for cellular studies and may ultimately facilitate the synthesis of therapeutic proteins containing unnatural amino acids in mammalian cells.     

The role of arginine 310 in catalysis and substrate specificity in xanthine dehydrogenase from Rhodobacter capsulatus

The rapid reaction kinetics of wild-type xanthine dehydrogenase from Rhodobacter capsulatus and variants at Arg-310 in the active site have been characterized for a variety of purine substrates. With xanthine as substrate, k(red) (the limiting rate of enzyme reduction by substrate at high [S]) decreased approximately 20-fold in an R310K variant and 2 x 10(4)-fold in an R310M variant. Although Arg-310 lies on the opposite end of the substrate from the C-8 position that becomes hydroxylated, its interaction with substrate still contributed approximately 4.5 kcal/mol toward transition state stabilization. The other purines examined fell into two distinct groups: members of the first were effectively hydroxylated by the wild-type enzyme but were strongly affected by the exchange of Arg-310 to methionine (with a reduction in k(red) greater than 10(3)), whereas members of the second were much less effectively hydroxylated by wild-type enzyme but also much less significantly affected by the amino acid exchanges (with a reduction in k(red) less than 50-fold). The effect was such that the 4000-fold range in k(red) seen with wild-type enzyme was reduced to a mere 4-fold in the R310M variant. The data are consistent with a model in which "good" substrates are bound "correctly" in the active site in an orientation that allows Arg-310 to stabilize the transition state for the first step of the overall reaction via an electrostatic interaction at the C-6 position, thereby accelerating the reaction rate. On the other hand, "poor" substrates bound upside down relative to this "correct" orientation. In so doing, they are unable to avail themselves of the additional catalytic power provided by Arg-310 in wild-type enzyme but, for this reason, are significantly less affected by mutations at this position. The kinetic data thus provide a picture of the specific manner in which the physiological substrate xanthine is oriented in the active site relative to Arg-310 and how this residue is used catalytically to accelerate the reaction rate (rather than simply bind substrate) despite being remote from the position that is hydroxylated.     

Photoreactive bicyclic amino acids as substrates for mutant Escherichia coli phenylalanyl-tRNA synthetases

Unnatural amino acids carrying reactive groups that can be selectively activated under non-invasive biologically benign conditions are of interest in protein engineering as biological tools for the analysis of protein-protein and protein-nucleic acids interactions. The double ring system phenylalanine analogues benzofuranylalanine and benzotriazolylalanine were synthesized, and their photolability was tested by UV irradiation at 254, 320, and 365 nm. Although both showed photo reactivity, benzofuranylalanine appeared as the most promising compound because this amino acid was activated by UVA (long wavelength) irradiation. These amino acids were also tested for in vitro charging of tRNA(Phe) and for protein mutagenesis via the phenylalanyl-tRNA synthetase variant alphaA294G that is able to facilitate in vivo protein synthesis using a range of para-substituted phenylalanine analogues. The results demonstrate that benzofuranylalanine, but not benzotriazolylalanine, is a substrate for phenylalanine tRNA synthetase alphaA294G, and matrix-assisted laser desorption ionization time-of-flight analysis showed it to be incorporated into a model protein with high efficiency. The in vivo incorporation into a target protein of a bicyclic phenylalanine analogue, as described here, demonstrates the applicability of phenylalanine tRNA synthetase variants in expanding the scope of protein engineering.     

Identification of an outer segment targeting signal in the COOH terminus of rhodopsin using transgenic Xenopus laevis

Mislocalization of the photopigment rhodopsin may be involved in the pathology of certain inherited retinal degenerative diseases. Here, we have elucidated rhodopsin's targeting signal which is responsible for its polarized distribution to the rod outer segment (ROS). Various green fluorescent protein (GFP)/rhodopsin COOH-terminal fusion proteins were expressed specifically in the major red rod photoreceptors of transgenic Xenopus laevis under the control of the Xenopus opsin promoter. The fusion proteins were targeted to membranes via lipid modifications (palmitoylation and myristoylation) as opposed to membrane spanning domains. Membrane association was found to be necessary but not sufficient for efficient ROS localization. A GFP fusion protein containing only the cytoplasmic COOH-terminal 44 amino acids of Xenopus rhodopsin localized exclusively to ROS membranes. Chimeras between rhodopsin and alpha adrenergic receptor COOH-terminal sequences further refined rhodopsin's ROS localization signal to its distal eight amino acids. Mutations/deletions of this region resulted in partial delocalization of the fusion proteins to rod inner segment (RIS) membranes. The targeting and transport of endogenous wild-type rhodopsin was unaffected by the presence of mislocalized GFP fusion proteins.