Structure-based engineering of streptavidin monomer with a reduced biotin dissociation rate.

We recently reported the engineering of monomeric streptavidin, mSA, corresponding to one subunit of wild type (wt) streptavidin tetramer. The monomer was designed by homology modeling, in which the streptavidin and rhizavidin sequences were combined to engineer a high affinity binding pocket containing residues from a single subunit only. Although mSA is stable and binds biotin with nanomolar affinity, its fast off rate (koff ) creates practical challenges during applications. We obtained a 1.9 Å crystal structure of mSA bound to biotin to understand their interaction in detail, and used the structure to introduce targeted mutations to improve its binding kinetics. To this end, we compared mSA to shwanavidin, which contains a hydrophobic lid containing F43 in the binding pocket and binds biotin tightly. However, the T48F mutation in mSA, which introduces a comparable hydrophobic lid, only resulted in a modest 20-40% improvement in the measured koff . On the other hand, introducing the S25H mutation near the bicyclic ring of bound biotin increased the dissociation half life (t½ ) from 11 to 83 min at 20°C. Molecular dynamics (MD) simulations suggest that H25 stabilizes the binding loop L3,4 by interacting with A47, and protects key intermolecular hydrogen bonds by limiting solvent entry into the binding pocket. Concurrent T48F or T48W mutation clashes with H25 and partially abrogates the beneficial effects of H25. Taken together, this study suggests that stabilization of the binding loop and solvation of the binding pocket are important determinants of the dissociation kinetics in mSA.

Stable, high-affinity streptavidin monomer for protein labeling and monovalent biotin detection.

The coupling between the quaternary structure, stability and function of streptavidin makes it difficult to engineer a stable, high affinity monomer for biotechnology applications. For example, the binding pocket of streptavidin tetramer is comprised of residues from multiple subunits, which cannot be replicated in a single domain protein. However, rhizavidin from Rhizobium etli was recently shown to bind biotin with high affinity as a dimer without the hydrophobic tryptophan lid donated by an adjacent subunit. In particular, the binding site of rhizavidin uses residues from a single subunit to interact with bound biotin. We therefore postulated that replacing the binding site residues of streptavidin monomer with corresponding rhizavidin residues would lead to the design of a high affinity monomer useful for biotechnology applications. Here, we report the construction and characterization of a structural monomer, mSA, which combines the streptavidin and rhizavidin sequences to achieve optimized biophysical properties. First, the biotin affinity of mSA (K(d) = 2.8 nM) is the highest among nontetrameric streptavidin, allowing sensitive monovalent detection of biotinylated ligands. The monomer also has significantly higher stability (T(m) = 59.8 °C) and solubility than all other previously engineered monomers to ensure the molecule remains folded and functional during its application. Using fluorescence correlation spectroscopy, we show that mSA binds biotinylated targets as a monomer. We also show that the molecule can be used as a genetic tag to introduce biotin binding capability to a heterologous protein. For example, recombinantly fusing the monomer to a cell surface receptor allows direct labeling and imaging of transfected cells using biotinylated fluorophores. A stable and functional streptavidin monomer, such as mSA, should be a useful reagent for designing novel detection systems based on monovalent biotin interaction.

Engineered streptavidin monomer and dimer with improved stability and function.

Although streptavidin's high affinity for biotin has made it a widely used and studied binding protein and labeling tool, its tetrameric structure may interfere with some assays. A streptavidin mutant with a simpler quaternary structure would demonstrate a molecular-level understanding of its structural organization and lead to the development of a novel molecular reagent. However, modulating the tetrameric structure without disrupting biotin binding has been extremely difficult. In this study, we describe the design of a stable monomer that binds biotin both in vitro and in vivo. To this end, we constructed and characterized monomers containing rationally designed mutations. The mutations improved the stability of the monomer (increase in T(m) from 31 to 47 °C) as well as its affinity (increase in K(d) from 123 to 38 nM). We also used the stability-improved monomer to construct a dimer consisting of two streptavidin subunits that interact across the dimer-dimer interface, which we call the A/D dimer. The biotin binding pocket is conserved between the tetramer and the A/D dimer, and therefore, the dimer is expected to have a significantly higher affinity than the monomer. The affinity of the dimer (K(d) = 17 nM) is higher than that of the monomer but is still many orders of magnitude lower than that of the wild-type tetramer, which suggests there are other factors important for high-affinity biotin binding. We show that the engineered streptavidin monomer and dimer can selectively bind biotinylated targets in vivo by labeling the cells displaying biotinylated receptors. Therefore, the designed mutants may be useful in novel applications as well as in future studies in elucidating the role of oligomerization in streptavidin function.

Disulfide trapping of protein complexes on the yeast surface.

Protein complexes are common in nature and play important roles in biology, but studying the quaternary structure formation in vitro is challenging since it involves lengthy and expensive biochemical steps. There are frequent technical difficulties as well with the sensitivity and resolution of the assays. In this regard, a technique that can analyze protein-protein interactions in high throughput would be a useful experimental tool. Here, we introduce a combination of yeast display and disulfide trapping that we refer to as stabilization of transient and unstable complexes by engineered disulfide (STUCKED) that can be used to detect the formation of a broad spectrum of protein complexes on the yeast surface using fluorescence labeling. The technique uses an engineered intersubunit disulfide to covalently crosslink the subunits of a complex, so that the disulfide-trapped complex can be displayed on the yeast surface for detection and analysis. Transient protein complexes are difficult to display on the yeast surface, since they may dissociate before they can be detected due to a long induction period in yeast. To this end, we show that three different quaternary structures with the subunit dissociation constant K(d) approximately 0.5-20 microM, the antibody variable domain (Fv), the IL-8 dimer, and the p53-MDM2 complex, cannot be displayed on the yeast surface as a noncovalent complex. However, when we introduce an interchain disulfide between the subunits, all three systems are efficiently displayed on the yeast surface, showing that disulfide trapping can help display protein complexes that cannot be displayed otherwise. We also demonstrate that a disulfide forms only between the subunits that interact specifically, the displayed complexes exhibit functional characteristics that are expected of wt proteins, the mutations that decrease the affinity of subunit interaction also reduce the display efficiency, and most of the disulfide stabilized complexes are formed within the secretory pathway during export to the surface. Disulfide crosslinking is therefore a convenient way to study weak protein association in the context of yeast display.

Mapping the dynamics and nanoscale organization of synaptic adhesion proteins using monomeric streptavidin.

The advent of super-resolution imaging (SRI) has created a need for optimized labelling strategies. We present a new method relying on fluorophore-conjugated monomeric streptavidin (mSA) to label membrane proteins carrying a short, enzymatically biotinylated tag, compatible with SRI techniques including uPAINT, STED and dSTORM. We demonstrate efficient and specific labelling of target proteins in confined intercellular and organotypic tissues, with reduced steric hindrance and no crosslinking compared with multivalent probes. We use mSA to decipher the dynamics and nanoscale organization of the synaptic adhesion molecules neurexin-1ß, neuroligin-1 (Nlg1) and leucine-rich-repeat transmembrane protein 2 (LRRTM2) in a dual-colour configuration with GFP nanobody, and show that these proteins are diffusionally trapped at synapses where they form apposed trans-synaptic adhesive structures. Furthermore, Nlg1 is dynamic, disperse and sensitive to synaptic stimulation, whereas LRRTM2 is organized in compact and stable nanodomains. Thus, mSA is a versatile tool to image membrane proteins at high resolution in complex live environments, providing novel information about the nano-organization of biological structures.

Biotin-assisted folding of streptavidin on the yeast surface.

Yeast surface display allows heterologously expressed proteins to be targeted to the exterior of the cell wall and thus has a potential as a biotechnology platform. In this study, we report the successful display of functional streptavidin on the yeast surface. Streptavidin binds the small molecule biotin with high affinity (K(d) ≈ 10(-14)M) and is used widely in applications that require stable noncovalent interaction, including immobilization of biotinylated compounds on a solid surface. As such, engineering functional streptavidin on the yeast surface may find novel uses in future biotechnology applications. Although the molecule does not require any post-translational modification, streptavidin is difficult to fold in bacteria. We show that Saccharomyces cerevisiae can fold the protein correctly if induced at 20°C. Contrary to a previous report, coexpression of anchored and soluble streptavidin subunits is not necessary, as expressing the anchored subunit alone is sufficient to form a functional complex. For unstable monomer mutants, however, addition of free biotin during protein induction is necessary to display a functional molecule, suggesting that biotin helps the monomer fold. To show that surface displayed streptavidin can be used to immobilize other biomolecules, we used it to capture biotinylated antibody, which is then used to immunoprecipitate a protein target.

Strain improvement of recombinant Escherichia coli for efficient production of plant flavonoids.

Plant flavonoid polyphenols continue to find increasing pharmaceutical and nutraceutical applications; however their isolation, especially of pure compounds, from plant material remains an underlying challenge. In the past Escherichia coli, one of the most well-characterized microorganisms, has been utilized as a recombinant host for protein expression and heterologous biosynthesis of small molecules. However, in many cases the expressed protein activities and biosynthetic efficiency are greatly limited by the host cellular properties, such as precursor and cofactor availability and protein or product tolerance. In the present work, we developed E. coli strains capable of high-level flavonoid synthesis through traditional metabolic engineering techniques. In addition to grafting the plant biosynthetic pathways, the methods included engineering of an alternative carbon assimilation pathway and the inhibition of competitive reaction pathways in order to increase intracellular flavonoid backbone precursors and cofactors. With this strategy, we report the production of plant-specific flavanones up to 700 mg/L and anthocyanins up to 113 mg/L from phenylpropanoic acid and flavan-3-ol precursors, respectively. These results demonstrated the efficient and scalable production of plant flavonoids from E. coli for pharmaceutical and nutraceutical applications.

Engineering central metabolic pathways for high-level flavonoid production in Escherichia coli.

The identification of optimal genotypes that result in improved production of recombinant metabolites remains an engineering conundrum. In the present work, various strategies to reengineer central metabolism in Escherichia coli were explored for robust synthesis of flavanones, the common precursors of plant flavonoid secondary metabolites. Augmentation of the intracellular malonyl coenzyme A (malonyl-CoA) pool through the coordinated overexpression of four acetyl-CoA carboxylase (ACC) subunits from Photorhabdus luminescens (PlACC) under a constitutive promoter resulted in an increase in flavanone production up to 576%. Exploration of macromolecule complexes to optimize metabolic efficiency demonstrated that auxiliary expression of PlACC with biotin ligase from the same species (BirAPl) further elevated flavanone synthesis up to 1,166%. However, the coexpression of PlACC with Escherichia coli BirA (BirAEc) caused a marked decrease in flavanone production. Activity improvement was reconstituted with the coexpression of PlACC with a chimeric BirA consisting of the N terminus of BirAEc and the C terminus of BirAPl. In another approach, high levels of flavanone synthesis were achieved through the amplification of acetate assimilation pathways combined with the overexpression of ACC. Overall, the metabolic engineering of central metabolic pathways described in the present work increased the production of pinocembrin, naringenin, and eriodictyol in 36 h up to 1,379%, 183%, and 373%, respectively, over production with the strains expressing only the flavonoid pathway, which corresponded to 429 mg/liter, 119 mg/liter, and 52 mg/liter, respectively.

Expression of a soluble flavone synthase allows the biosynthesis of phytoestrogen derivatives in Escherichia coli.

Flavones are plant secondary metabolites with potent pharmacological properties. We report the functional expression of FSI, a flavonoid 2-oxoglutarate-dependent dioxygenase-encoding flavone synthase from parsley in Escherichia coli. This expression allows the biosynthesis of various flavones from phenylpropanoid acids in recombinant E. coli strains simultaneously expressing five plant-specific flavone biosynthetic genes. The gene ensemble consists of 4CL-2 (4-coumarate:CoA ligase) and FSI (flavone synthase I) from parsley, chsA (chalcone synthase) and chiA (chalcone isomerase) from Petunia hybrida, and OMT1A (7-O-methyltransferase) from peppermint. After a 24-h cultivation, the recombinant E. coli produces significant amounts of apigenin (415 microg/l), luteolin (10 microg/l), and genkwanin (208 microg/l). The majority of the flavone products are excreted in the culture media; however, 25% is contained within the cells. The metabolic engineering strategy presented demonstrates that plant-specific flavones are successfully produced in E. coli for the first time by incorporating a soluble flavone synthase confined only in Apiaceae.

Investigation of two distinct flavone synthases for plant-specific flavone biosynthesis in Saccharomyces cerevisiae.

Flavones are plant secondary metabolites that have wide pharmaceutical and nutraceutical applications. We previously constructed a recombinant flavanone pathway by expressing in Saccharomyces cerevisiae a four-step recombinant pathway that consists of cinnamate-4 hydroxylase, 4-coumaroyl:coenzyme A ligase, chalcone synthase, and chalcone isomerase. In the present work, the biosynthesis of flavones by two distinct flavone synthases was evaluated by introducing a soluble flavone synthase I (FSI) and a membrane-bound flavone synthase II (FSII) into the flavanone-producing recombinant yeast strain. The resulting recombinant strains were able to convert various phenylpropanoid acid precursors into the flavone molecules chrysin, apigenin, and luteolin, and the intermediate flavanones pinocembrin, naringenin, and eriodictyol accumulated in the medium. Improvement of flavone biosynthesis was achieved by overexpressing the yeast P450 reductase CPR1 in the FSII-expressing recombinant strain and by using acetate rather than glucose or raffinose as the carbon source. Overall, the FSI-expressing recombinant strain produced 50% more apigenin and six times less naringenin than the FSII-expressing recombinant strain when p-coumaric acid was used as a precursor phenylpropanoid acid. Further experiments indicated that unlike luteolin, the 5,7,4'-trihydroxyflavone apigenin inhibits flavanone biosynthesis in vivo in a nonlinear, dose-dependent manner.