Thioredoxin fusions increase folding of single chain Fv antibodies in the cytoplasm of Escherichia coli: evidence that chaperone activity is the prime effect of thioredoxin

In this study we investigate the effect of thioredoxin (Trx1) protein fusions in the production, oxidation, and folding of single chain Fv (scFv) antibodies in the cytoplasm of Escherichia coli. We analyze the expression levels, solubility, disulfide-bond formation, and antigen-binding properties of Trx1-scFv fusions in E. coli wild-type cells and isogenic strains carrying mutations in the glutathione oxidoreductase (gor) and/or thioredoxin reductase (trxB) genes. We compare the Trx1-scFv fusions with other reported systems for production of scFv in the cytoplasm of E. coli, including protein fusions to the maltose-binding protein. In addition, we analyze the effect of co-expressing a signal-sequence-less derivative of the periplasmic chaperone and disulfide-bond isomerase DsbC (DeltassDsbC), which has been shown to act as a chaperone for scFvs in the cytoplasm. The results reported here demonstrate that Trx1 fusions produce the highest expression level and induce the correct folding of scFvs even in the absence of DeltassDsbC in the cytoplasm of E. coli trxB gor cells. The disulfide bridges of Trx1-scFv fusions were formed correctly in E. coli trxB gor cells, but not in trxB single mutants. Antigen-binding assays showed that Trx1 has only a minor influence in the affinity of the scFv, indicating that Trx1-scFv fusions can be used without removal of the Trx1 moiety. In addition, we proved that a Trx1"AGPA" variant, having its catalytic cysteine residues mutated to alanine, was fully capable of assisting the folding of the fused scFvs. Taken together, our data indicate that the Trx1 moiety acts largely as an intramolecular protein chaperone, not as a disulfide bond catalyst, inducing the correct folding of scFvs in the cytoplasm of E. coli trxB gor cells.     

Macromolecular crowding in the Escherichia coli periplasm maintains alpha-synuclein disorder

The natively disordered protein alpha-synuclein is the primary component of Lewy bodies, the cellular hallmark of Parkinson's disease. Most studies of this protein are performed in dilute solution, but its biologically relevant role is performed in the crowded environment inside cells. We addressed the effects of macromolecular crowding on alpha-synuclein by combining NMR data acquired in living Escherichia coli with in vitro NMR data. The crowded environment in the E.coli periplasm prevents a conformational change that is detected at 35 degrees C in dilute solution. This change is associated with an increase in hydrodynamic radius and the formation of secondary structure in the N-terminal 100 amino acid residues. By preventing this temperature-induced conformational change, crowding in the E.coli periplasm stabilizes the disordered monomer. We obtain the same stabilization in vitro upon crowding alpha-synuclein with 300 g/l of bovine serum albumin, indicating that crowding alone is sufficient to stabilize the disordered, monomeric protein. Two disease-associated variants (A30P and A53T) behave in the same way in both dilute solution and in the E.coli periplasm. These data reveal the importance of approaching the effects of macromolecular crowding on a case-by-case basis. Additionally, our work shows that discrete structured protein conformations may not be achieved by alpha-synuclein inside cells, implicating the commonly overlooked aspect of macromolecular crowding as a possible factor in the etiology of Parkinson's disease.     

Interactions of the Escherichia coli DnaB helicase hexamer with the replication factor the DnaC protein. Effect of nucleotide cofactors and the ssDNA on protein-protein interactions and the topology of the complex

Quantitative studies of interactions between the Escherichia coli replication factor DnaC protein and the DnaB helicase have been performed using sedimentation velocity and fluorescence energy transfer techniques. The applied novel analysis of the sedimentation data allows us to construct thermodynamic rigorous binding isotherms without any assumption as to the relationship between the observed molecular property of the complexes formed, the average sedimentation coefficient, or the degree of binding. Experiments have been performed with the fluorescein-modified DnaB helicase, which allows an exclusive monitoring of the DnaB-DnaC complex formation. The DnaC binding to the unmodified helicase has been characterized in competition experiments. The data establish that, in the presence of the ATP analog AMP-PNP, or ADP, a maximum of six DnaC monomers bind cooperatively to the DnaB hexamer. The positive cooperative interactions are limited to the two neighboring DnaC molecules. Analyses using a statistical thermodynamic hexagon model indicate that, under the solution conditions examined, the affinity is characterized by the intrinsic binding constant K=1.4(+/-0.5)x10(5)M(-1) and cooperativity parameter sigma=21+/-5. These data suggest strongly that the DnaC-DnaB complex exists in vivo as a mixture of complexes with a different number of bound DnaC molecules, although the complex with six DnaC molecules bound dominates the distribution. The DnaC nucleotide-binding site is not involved in the stabilization of the complex. Moreover, the hydrolysis of NTP bound to the helicase or the DnaC is not required for the release of the DnaC protein from the complex. The single-stranded DNA (ssDNA) bound to the helicase does not affect the DnaC protein binding. However, in the presence of the DNA, there is a significant difference in the energetics and structure of the ternary complex, DnaC-DnaB-ssDNA, formed in the presence of AMP-PNP as compared to ADP. The topology of the ternary complex DnaC-DnaB-ssDNA has been determined using the fluorescence energy transfer method. In solution, the DnaC protein-binding site is located on the large 33 kDa domain of the DnaB helicase. The significance of the results in the functioning of the DnaB helicase-DnaC protein complex is discussed.     

Both temperature and medium composition regulate RNase E processing efficiency of the rpsO mRNA coding for ribosomal protein S15 of Escherichia coli

Cleavage by RNase E is believed to be the rate-limiting step in the degradation of many RNAs. These cleavages are modulated by 5' end-phosphorylation, folding and translation of the mRNA in question. Here, we present data suggesting that these cleavages are also regulated by environmental conditions. We report that rpsO mRNA, 15 minutes after a shift to 44 degrees C, is stabilized in cells grown in minimal medium. This stabilization is correlated with a reduction in the efficiency of the RNase E cleavage which initiates its decay. We also observe the appearance of RNA fragments previously detected following RNase E inactivation and a defect in the adaptation of RNase E concentration. These observations, coupled to the fact that RNase E overproduction slightly reduces the accumulation of the rpsO mRNA, suggest that this stabilization is caused in part by a limitation in RNase E concentration. An increase in the steady-state level of rpsT mRNA is also observed following a shift to 44 degrees C in minimal medium; however, processing of the 9 S rRNA precursor is not affected under these conditions. We thus propose that RNase E concentration changes in the cell in response to environmental conditions and that these changes can selectively affect the processing and the stability of individual mRNAs. Our data also indicate that the efficiency of cleavage of the rpsO mRNA by RNase E is modified by other factor(s) which remain to be identified.     

A structure-based anatomy of the E.coli metabolome

The Escherichia coli metabolome has been characterised using the two-dimensional structures of 745 metabolites, obtained from the EcoCyc and KEGG databases. Physicochemical properties of the metabolome have been calculated to provide an overview of this set of cognate ligands. A library of fragments commonly found among these molecules has been employed to reveal the main constituents of metabolites, and to assist a broad classification of the metabolome into biochemically relevant classes. Fragment-based fingerprints reveal the metabolome as a continuum in the two-dimensional structural space, where clusters of molecules sharing similar scaffolds can be identified, but are generally overlapping. Nucleotide, carbohydrate and amino acid-like molecules are the most prominent, but at high levels of similarity, a more detailed classification is possible. Classification schemes for the metabolome are a promising tool for understanding the chemical diversity of the metabolome. When used in conjunction with existing classifications of the proteome, they can help to elucidate the binding preferences and promiscuity of proteins and their cognate substrates.     

Production of a functional single-chain Fv protein in transgenic tobacco root exudates

Nicotiana tabacum was transformed with a gene encoding anti-PreS1 of hepatitis B surface antigen single-chain Fv antibody (scFv) and bearing an N-terminal endoplasmic reticulum protein signal peptide sequence. The scFv antibody protein was continuously secreted from the transgenic tobacco roots into a simple hydroponic medium at 630 to 760 ng g^–1 dry wt root day^–1. The antibody was about 2% of the total secreted protein and still possessed antigen-binding activity.     

A ligand-centric analysis of the diversity and evolution of protein-ligand relationships in E.coli

As enzymes evolve and diverge from common ancestor sequences, they often keep their overall reaction chemistry but specialize in the binding of different cognate ligands. This study borrows methods for the computational assessment of 2D similarity of small molecules from the field of chemoinformatics, to examine the extent of structure conservation of cognate ligands binding to similar proteins. Proteins from 87 structural superfamilies from Escherichia coli form the core dataset, which is extended using homologues with functional assignments from any organism. We find that correlation of the substrate similarity with protein similarity (measured by either sequence-based or structure-based scores) can only be clearly established for very similar proteins. At low sequence identities, the superfamily to which a protein belongs can give helpful clues to its function, and more importantly, the confidence attached to such clues is superfamily-dependent. Our data indicate that only a few superfamilies show great substrate diversity, and that most exhibit conservation of at least part of the structural scaffold of the substrate.     

Feedback control of ribosome function in Escherichia coli

We have previously proposed that the rate of ribosome function during balanced growth in E. coli, expressed as the rate of peptide chain elongation, is adjusted by a feedback mechanism: whenever that rate is submaximal (i.e. below 22 amino acid residues polymerized per active ribosome at 37 degrees C), the feedback signal ppGpp is generated by an activation of the ppGpp synthetase expressed from the spoT gene. The accumulation of ppGpp reduces the synthesis of additional ribosomes and thereby reduces the consumption of amino acids which, in turn, allows the remaining ribosomes to function at a higher rate. Here we have described with supporting evidence the proposed feedback loop in greater detail and provided a mathematical analysis which predicts that the SpoT ppGpp synthetase activity should be highest when the ribosomes function at their half-maximal rate. This prediction is consistent with reported observations and is independent of the particular (unknown) mechanism by which the rate of translation controls the ppGpp synthetase activity of SpoT.     

抗汉坦病毒核蛋白单链抗体的构建与表达

汉坦病毒核蛋白在病毒各个基因组片段核糖核蛋白(RNP)复合物的形成,以及RNP复合物装配入病毒颗粒中发挥重要作用。体外研究表明,核蛋白与病毒RNA的特异性结合结构域位于第175—217个氨基酸残基,羧基端其它部分为非特异性结合结构域。为了在细胞内病毒复制的正常过程中研究核蛋白的功能,应用噬菌体表面呈现技术,分别从鼠杂交瘤细胞和人外周血淋巴细胞天然抗体库中,筛选出两株不同抗原结合位点的抗汉坦病毒核蛋白抗体L13 F3和H34Fab抗体,并在大肠杆菌中进行表达。L13 F3 Fab抗体的抗原结合位点位于核蛋白氨基端25—65个氨基酸之间,H34的抗原结合位点位于核蛋白的羧基端的一半。分别使重链可变区羧基端与轻链可变区氨基端通过9个氨基酸寡肽连接起来,构建成单链抗体,并在大肠杆菌内进行表达。通过酶联免疫吸附试验(EL-SIA)、间接免疫荧光试验(IFA)和Western blot检测,确定所构建单链抗体与母抗体的抗原结合特性无差别。单链抗体分子量小,易于操作并保留了全部的抗原结合活性,是在细胞内探索汉坦病毒核蛋白功能的良好工具。     

Eclipse-synchrony relationship in Escherichia coli strains with mutations affecting sequestration, initiation of replication and superhelicity of the bacterial chromosome

Initiation of replication from oriC on the Escherichia coli chromosomes occurs once and only once per generation at the same cell mass per origin. During rapid growth there are overlapping replication cycles, and initiation occurs synchronously at two or more copies of oriC. Since the bacterial growth can vary over a wide range (from three divisions per hour to 2.5 hours or more per division) the frequency of initiation should change in coordination with bacterial growth. Prevention of reinitiation from a newly replicated origin by temporary sequestration of the hemi-methylated GATC-sites in the origin region provides the molecular/genetic basis for the maintenance of the eclipse period between two successive rounds of replication. Sequestration is also believed to be responsible for initiation synchrony, since inactivation of either the seqA or the dam gene abolishes synchrony while drastically reducing the eclipse. In this work, we attempted to examine the functional relationship(s) between the eclipse period and the synchrony of initiation in E.coli strains by direct measurements of these parameters by density-shift centrifugation and flow-cytometric analyses, respectively. The eclipse period, measured as a fraction of DNA-duplication times, varied continuously from 0.6 for the wild-type E.coli K12 to 0.1 for strains with mutations in seqA, dam, dnaA, topA and gyr genes (all of which have been shown to cause asynchrony) and their various combinations. The asynchrony index, a quantitative indicator for the loss of synchrony of initiation, changed from low (synchronous) to high (asynchronous) values in a step-function-like relationship with the eclipse. An eclipse period of approximately 0.5 generation time appeared to be the critical value for the switch from synchronous to asynchronous initiation.     

Unzipping mechanism of the double-stranded DNA unwinding by a hexameric helicase: the effect of the 3' arm and the stability of the dsDNA on the unwinding activity of the Escherichia coli DnaB helicase

The effect of two structural elements of a replication DNA fork substrate, the length of the 3' arm of the fork and the stability of the double-stranded DNA (dsDNA) part, on the kinetics of the dsDNA unwinding by the Escherichia coli hexameric helicase DnaB protein has been examined under single turnover conditions using the rapid quench-flow technique. The length of the 3' arm of the replication fork, i.e. the number of nucleotides in the arm, is a major structural factor that controls the unwinding rate and processivity of the helicase. The data show the existence of an optimal length of the 3' arm where there is the highest unwinding rate and processivity, indicating that during the unwinding process, the helicase transiently interacts with the 3' arm at a specific distance on the arm with respect to the duplex part of the DNA. Moreover, the area on the enzyme that engages in interactions has also a discrete size. For DNA substrates with the 3' arm containing 14, or less, nucleotide residues, the DnaB helicase becomes a completely distributive enzyme. However, the 3' arm is not a "specific activating cofactor" in the unwinding reaction. Rather, the 3' arm plays a role as a mechanical fulcrum for the enzyme, necessary to provide support for the advancing large helicase molecule on the opposite strand of the DNA. Binding of ATP is necessary to engage the 3' arm with the DnaB helicase, but it does not change the initial distribution of complexes of the enzyme with the DNA fork substrate. Stability of the dsDNA has a significant effect on the unwinding rate and processivity. The unwinding rate constant is a decreasing linear function of the fractional content of GC base-pairs in the dsDNA, indicating that the activation of the unwinding step is proportional to the stability of the nucleic acid.     

抑制E.coli的SOS应答对恩诺沙星抗药性的影响

通过敲除SOS应答启动蛋白基因rec A,探讨SOS应答对E.coli恩诺沙星抗药性的影响,并体外评价Rec A抑制剂和恩诺沙星联用对细菌协同抑制作用的影响.利用Red重组系统,构建E.coli ATCC 25922的rec A缺失菌株E.coli ATCC 25922/?rec A;在恩诺沙星压力下,利用荧光定量PCR测定SOS应答系统相关基因rec A和umu C表达量的变化.用微量肉汤稀释法测定恩诺沙星等常用抗生素对两个菌株的MIC变化;利用梯度平板法测定恩诺沙星对两个菌株抗药性变异的影响;合成Rec A抑制剂,并评估其与恩诺沙星联合抑制E.coli生长及其抗药性的作用.结果表明,E.coli ATCC 25922/?rec A菌株对恩诺沙星的最低抑菌浓度值降低至原始菌株的1/8;经药物处理后,在梯度平板上,rec A缺失菌株较野生型不易产生抗药性;荧光定量PCR表明,rec A缺失菌株或在Rec A抑制剂作用下,SOS应答系统受到一定的抑制.敲除rec A,使菌株对恩诺沙星的抗药性和抗药率均明显降低;Rec A抑制剂在一定程度上能抑制SOS应答,起到协同抑菌作用.     

Unzipping mechanism of the double-stranded DNA unwinding by a hexameric helicase: quantitative analysis of the rate of the dsDNA unwinding, processivity and kinetic step-size of the Escherichia coli DnaB helicase using rapid quench-flow method

Kinetics of the double-stranded (ds) DNA unwinding by the Escherichia coli replicative helicase DnaB protein has been examined under single-turnover conditions using the chemical quench-flow technique. The unwinding reaction proceeds through an initial conformational transition followed by the unwinding catalytic steps and the release of the single-stranded (ss) DNA. Analyses of the reaction as a function of the number of base-pairs in the dsDNA reveal that the number of catalytic steps is not strictly proportional to the length of the dsDNA. As the helicase approaches the end of the substrate, the remaining approximately 11 bp of the DNA melts without catalytic participation of the enzyme. The kinetic step-size of the DnaB helicase, i.e. the number of the base-pairs unwound in a single catalytic step is only 1.4(+/- 0.2). The low value of the step-size indicates that the helicase unwinds a single base-pair in a single catalytic step. Thus, the DnaB helicase unzips the dsDNA in a reverse process to the zipping mechanism of the non-enzymatic double helix formation. The protein is a fast helicase that at 25 degrees C unwinds approximately 291 bp/s, much faster than previously thought, and the unwinding rate can be much higher at higher temperatures. However, the ATP-state of the enzyme has an increased dissociation rate, resulting in only a moderate unwinding processivity, P = 0.89(+/- 0.03), little dependent on the temperature. The conformational transition of the DnaB helicase-DNA complex, preceding the unwinding, is an intrinsic transition of the enzyme from the stationary conformation to the ATP-state of the helicase.     

Combination of Dsb coexpression and an addition of sorbitol markedly enhanced soluble expression of single-chain Fv in Escherichia coli

Many eukaryotic proteins have been produced successfully in Escherichia coli. However, not every gene can be expressed efficiently in this organism. Most proteins, especially those with multiple disulfide bonds, have been shown to form insoluble protein or inclusion body in E. coli. An inactive form of protein would require an in vitro refolding step to regain biological functions. In this study, we described the system for soluble expression of a single-chain variable fragment (scFv) against hepatocellular carcinoma (Hep27scFv) by coexpressing Dsb protein and enhancing with medium additives. The results revealed that overexpression of DsbABCD protein showed marked effect on the soluble production of Hep27scFv, presumably facilitating correct folding. The optimal condition for soluble scFv expression could be obtained by adding 0.5M sorbitol to the culture medium. The competitive enzyme-linked immunosorbent assay (ELISA) indicated that soluble scFv expressed by our method retains binding activity toward the same epitope on a hepatocellular carcinoma cell line (HCC-S102) recognized by intact antibody (Ab) (Hep27 Mab). Here, we report an effective method for soluble expression of scFv in E. coli by the Dsb coexpression system with the addition of sorbitol medium additive. This method might be applicable for high-yield soluble expression of proteins with multiple disulfide bonds.     

Efficient disinfection of Escherichia coli in water by silver loaded alumina

The catalytic inactivation of Escherichia coli (E. coli) in water by silver loaded alumina as catalyst was investigated. Ag/Al₂O₃ and AgCl/Al₂O₃ catalysts exhibited high bactericidal activity at room temperature in water with no need for any light or electrical power input. Dissolved oxygen which can be catalyzed to reactive oxygen species (ROS) was found to be essential for the strong bactericidal activities of the catalysts. Decomposition of the cell wall leading to leakage of the intracellular component and the complete lysis of the whole cell were directly observed by transmission electron microscopy (TEM). The resultant change in cell permeability was confirmed by potassium ion leakage. The different morphological changes between E. coli cells treated with the catalysts and Ag⁺ were also observed. The formation of ROS involved in the bactericidal process by AgCl/Al₂O₃ was confirmed by addition of catalase and OH scavenger. Higher temperature and pH value were found to have positive effect on the bactericidal activity of AgCl/Al₂O₃. All these results indicated that the bactericidal effect of the catalyst was a synergic action of ROS and Ag⁺, not an additive one. A possible mechanism is proposed.     

Soluble Expression of a Neo2/15-Conjugated Single Chain Fv against PD-L1 in Escherichia coli

Immunocytokines, antibody-cytokine fusion proteins, have the potential to improve the therapeutic index of cytokines by delivering the cytokine to the site of localized tumor cells using antibodies. In this study, we produced a recombinant anti-programmed death-ligand 1 (PD-L1) scFv, an antibody fragment against PD-L1 combined with a Neo2/15, which is an engineered interleukin with superior function using an E. coli expression system. We expressed the fusion protein in a soluble form and purified it, resulting in high yield and purity. The high PD-L1-binding efficiency of the fusion protein was confirmed via enzyme-linked immunosorbent assay, suggesting the application of this immunocytokine as a cancer-related therapeutic agent.     

Phage display reveals multiple contact sites between FhuA, an outer membrane receptor of Escherichia coli, and TonB

The ferric hydroxamate uptake receptor FhuA from Escherichia coli transports siderophores across the outer membrane (OM). TonB-ExbB-ExbD transduces energy from the cytoplasmic membrane to the OM by contacts between TonB and OM receptors that contain the Ton box, a consensus sequence near the N terminus. Although the Ton box is a region of known contact between OM receptors and TonB, our biophysical studies established that TonB binds to FhuA through multiple regions of interaction. Panning of phage-displayed random peptide libraries (Ph.D.-12, Ph.D.-C7C) against TonB identified peptide sequences that specifically interact with TonB. Analyses of these sequences using the Receptor Ligand Contacts (RELIC) suite of programs revealed clusters of multiply aligned peptides that mapped to FhuA. These clusters localized to a continuous periplasm-accessible surface: Ton box/switch helix; cork domain/beta1 strand; and periplasmic turn 8. Guided by such matches, synthetic oligonucleotides corresponding to DNA sequences identical to fhuA were fused to malE; peptides corresponding to the above regions were displayed at the N terminus of E.coli maltose-binding protein (MBP). Purified FhuA peptides fused to MBP bound specifically to TonB by ELISA. Furthermore, they competed with ligand-loaded FhuA for binding to TonB. RELIC also identified clusters of multiply aligned peptides corresponding to the Ton box regions in BtuB, FepA, and FecA; to periplasmic turn 8 in BtuB and FecA; and to periplasmic turns 1 and 2 in FepA. These experimental outcomes identify specific molecular contacts made between TonB and OM receptors that extend beyond the well-characterized Ton box.     

Tetramerization of the LexA repressor in solution: implications for gene regulation of the E.coli SOS system at acidic pH

Structural changes on LexA repressor promoted by acidic pH have been investigated. Intense protein aggregation occurred around pH 4.0 but was not detected at pH values lower than pH 3.5. The center of spectral mass of the Trp increased 400 cm(-1) at pH 2.5 relatively to pH 7.2, an indication that LexA has undergone structural reorganization but not denaturation. The Trp fluorescence polarization of LexA at pH 2.5 indicated that its hydrodynamic volume was larger than its dimer at pH 7.2. 4,4'-Dianilino-1,1'-binaphthyl-5,5'- disulfonic acid (bis-ANS) experiments suggested that the residues in the hydrophobic clefts already present at the LexA structure at neutral pH had higher affinity to it at pH 2.5. A 100 kDa band corresponding to a tetramer was obtained when LexA was subject to pore-limiting native polyacrylamide gel electrophoresis at this pH. The existence of this tetrameric state was also confirmed by small angle X-ray scattering (SAXS) analysis at pH 2.5. 1D 1H NMR experiments suggested that it was composed of a mixture of folded and unfolded regions. Although 14,000-fold less stable than the dimeric LexA, it showed a tetramer-monomer dissociation at pH 2.5 from the hydrostatic pressure and urea curves. Albeit with half of the affinity obtained at pH 7.2 (Kaff of 170 nM), tetrameric LexA remained capable of binding recA operator sequence at pH 2.5. Moreover, different from the absence of binding to the negative control polyGC at neutral pH, LexA bound to this sequence with a Kaff value of 1415 nM at pH 2.5. A binding stoichiometry experiment at both pH 7.2 and pH 2.5 showed a [monomeric LexA]/[recA operator] ratio of 2:1. These results are discussed in relation to the activation of the Escherichia coli SOS regulon in response to environmental conditions resulting in acidic intracellular pH. Furthermore, oligomerization of LexA is proposed to be a possible regulation mechanism of this regulon.     

De novo folding of GFP fusion proteins: high efficiency in eukaryotes but not in bacteria

Eukaryotic genomes encode a considerably higher fraction of multi-domain proteins than their prokaryotic counterparts. It has been postulated that efficient co-translational and sequential domain folding has facilitated the explosive evolution of multi-domain proteins in eukaryotes by the recombination of pre-existent domains. Here, we tested whether eukaryotes and bacteria differ generally in the folding efficiency of multi-domain proteins generated by domain recombination. To this end, we compared the folding behavior of a series of recombinant proteins comprised of green fluorescent protein (GFP) fused to four different robustly folding proteins through six different linkers upon expression in Escherichia coli and the yeast Saccharomyces cerevisiae. We found that, unlike yeast, bacteria are remarkably inefficient at folding these fusion proteins, even at comparable levels of expression. In vitro and in vivo folding experiments demonstrate that the GFP domain imposes significant constraints on de novo folding of its fusion partners in bacteria, consistent with a largely post-translational folding mechanism. This behavior may result from an interference of GFP with adjacent domains during folding due to the particular topology of the beta-barrel GFP structure. By following the accumulation of enzymatic activity, we found that the rate of appearance of correctly folded fusion protein per ribosome is indeed considerably higher in yeast than in bacteria.