Protein Engineering Design and Selection - recent issues

Structural insight into the binding mode between the targeting domain of ALE-1 (92AA) and pentaglycine of peptidoglycan

Hirakawa, H., Akita, H., Fujiwara, T., Sugai, M., Kuhara, S. — 2009-06-19

ALE-1 is a glycylglycine endopeptidase that selectively targets and lyses Staphylococcus aureus, and is expected to be a next generation antibacterial agent because of its substrate specificity to pathogenic bacteria. It has a central catalytic domain and a targeting domain called 92AA. 92AA has been shown to recognize pentaglycine, but the molecular mechanism by which it recognizes and interacts with pentaglycine has not been elucidated. To predict the binding modes of pentaglycine is important for estimating the catalytic reaction mechanism of ALE-1. In the present study, we characterized the binding cleft of 92AA by a computational method and modeled the complexes formed between 92AA and the pentaglycine of peptidoglycan by a binding simulation. In addition, we performed precise simulations of the molecular dynamics by which the complexes identify the amino acid residues interacting with the pentaglycine. We also experimentally constructed mutants in which the amino acid residues present in the binding cleft were changed by site-directed mutagenesis and assessed their ability to bind to peptidoglycan by ELISA. Based on the results of these analyses, we proposed a mode of binding between 92AA and the pentaglycine of peptidoglycan, and modeled the energetically stable complexes between 92AA and the pentaglycine.

Creating lactose phosphorylase enzymes by directed evolution of cellobiose phosphorylase

2009-06-19

Disaccharide phosphorylases are interesting enzymes for the production of sugar phosphates from cheap starting materials and for the synthesis of novel glycosides. Cellobiose phosphorylase (CP) from Cellulomonas uda was subjected to directed evolution in order to create enzyme variants with significantly increased lactose phosphorylase (LP) activity, useful for the production of -d-galactose 1-phosphate. In a first round, random mutagenesis was performed on part of the CP gene and the resultant library was selected on minimal lactose medium. One clone containing six amino acid mutations was found with increased LP activity compared with the wild-type CP enzyme. The negative and neutral mutations were eliminated by site-directed mutagenesis and the resultant enzyme variant containing two amino acid substitutions (T508A/N667T) showed more LP activity than the parent mutant. Saturation mutagenesis of the beneficial sites and screening for improved mutants allowed us to identify the T508I/N667A mutant which has 7.5 times higher specific activity on lactose than the wild-type. The kinetic parameters of the mutants were determined and showed that the increased LP activity was caused by a higher k_cat value. This is the first report of an engineered CP with modified substrate specificity.

Distributions of enzyme residues yielding mutants with improved substrate specificities from two different directed evolution strategies

Paramesvaran, J., Hibbert, E. G., Russell, A. J., Dalby, P. A. — 2009-06-19

A previous study of random mutations, mostly introduced by error-prone PCR (EPPCR) or DNA shuffling (DS), demonstrated that those closer to the enzyme active site were more effective than distant ones at improving enzyme activity, substrate specificity or enantioselectivity. Since then, many studies have taken advantage of this observation by targeting site-directed saturation mutagenesis (SDSM) to residues closer to or within enzyme active sites. Here, we have analysed a set of SDSM studies, in parallel to a similar set from EPPCR/DS, to determine whether the greater range of amino-acid types accessible by SDSM affects the distances at which the most effective sites occur. We have also analysed the relative effectiveness for obtaining beneficial mutants of residues with different degrees of natural sequence variation, as determined by their sequence entropy which is related to sequence conservation. These analyses attempt to answer the question—how well focused have targeted mutagenesis strategies been? We also compared two different sets of active-site atoms from which to measure distances and found that the inclusion of catalytic, substrate and cofactor atoms refined the analysis compared to using a single key catalytic atom. Using this definition, we found that EPPCR/DS is not effective for altering substrate specificity at sites that are within 5 Å of the active-site atoms. In contrast, SDSM is most effective when targeted to residues at <5–6 Å from the catalytic, substrate or cofactor atom, and also for residues with intermediate sequence entropies. Furthermore, SDSM is capable of altering substrate specificity at highly and completely conserved residues in the active site. The results suggest ways in which directed evolution by SDSM could be improved for greater efficiency in terms of reducing the library sizes required to obtain beneficial mutations that alter substrate specificity.

Directed evolution of Candida antarctica lipase A using an episomaly replicating yeast plasmid

Sandstrom, A. G., Engstrom, K., Nyhlen, J., Kasrayan, A., Backvall, J.-E. — 2009-06-19

We herein report the first directed evolution of Candida antarctica lipase A (CalA), employing a combinatorial active-site saturation test (CAST). Wild-type CalA has a modest E-value of 5.1 in kinetic resolution of 4-nitrophenyl 2-methylheptanoate. Enzyme variants were expressed in Pichia pastoris by using the episomal vector pBGP1 which allowed efficient secretory expression of the lipase. Iterative rounds of CASTing yielded variants with good selectivity toward both the (S)- and the (R)-enantiomer. The best obtained enzyme variants had E-values of 52 (S) and 27 (R).

Stabilising the DNA-binding domain of p53 by rational design of its hydrophobic core

Khoo, K. H., Joerger, A. C., Freund, S. M.V., Fersht, A. R. — 2009-06-19

The core domain of the tumour suppressor p53 is of inherently low thermodynamic stability and also low kinetic stability, which leads to rapid irreversible denaturation. Some oncogenic mutations of p53 act by just making the core domain thermosensitive, and so it is the target of novel anti-cancer drugs that bind to and stabilise the protein. Increasing the stability of the unstable core domain has also been crucial for biophysical and structural studies, in which a stabilised quadruple mutant (QM) is currently used. We generated an even more stabilised hexamutant (HM) by making two additional substitutions, Y236F and T253I, to the QM. The residues are found in the more stable paralogs p63 and p73 and stabilise the wild-type p53 core domain. We solved the structure of the HM core domain by X-ray crystallography at 1.75 Å resolution. It has minimal structural changes from QM that affect the packing of hydrophobic core residues of the β-sandwich. The full-length HM was also fully functional in DNA binding. HM was more stable than QM at 37°C. Anomalies in biophysics and spectroscopy in urea-mediated denaturation curves of HM implied the accumulation of a folding intermediate, which may be related to those detected in kinetic experiments. The two additional mutations over-stabilise an unfolding intermediate. These results should be taken into consideration in drug design strategies for increasing the stability of temperature-sensitive mutants of p53.

A structural model for the HAT domain of Utp6 incorporating bioinformatics and genetics

Champion, E. A., Kundrat, L., Regan, L., Baserga, S. J. — 2009-06-19

The half-a-tetratricopeptide (HAT) repeat motif is of interest because it is found exclusively in proteins that are involved in RNA metabolism. Little is known about structure–function relationships in this class of repeat motif. Here, we present the results of a combined bioinformatics, modeling and mutagenesis study of the HAT domain of Utp6. We have derived a new HAT consensus, delineated its structure-defining residues and, by homology modeling, have placed these residues in a structural context. By considering only HAT motifs from Utp6 proteins, we identified residues that are shared by, and unique to, only this subset of HAT motifs, suggesting a key functional role. Employing both random and directed mutagenesis of the HAT domain in yeast Utp6, we have identified residues whose mutation results in loss of function. By examining these residues in the context of the homology model, we have delineated those that act by perturbing structure and those that more likely have a direct effect on function. Importantly, the residues we predict to have a direct effect on function map together on the tertiary structure, thus defining a potential functional interaction surface.

Prediction and classification of chemokines and their receptors

Lata, S., Raghava, G.P.S. — 2009-06-19

Chemokines are low molecular mass cytokine-like proteins that orchestrate myriads of immune functions like leukocyte trafficking, T cell differentiation, angiogenesis, hematopeosis and mast cell degranulation. Chemokines also play a role as HIV-1 inhibitor and act as potent natural adjuvant in antitumor immunotherapy. Receptors for these molecules are all seven-pass transmembrane G-protein-coupled receptors that are intimately involved with chemokines in a wide array of physiological and pathological conditions. These receptors also have a major role as co-receptors for HIV-1 entry into target cells. Therefore, chemokine receptors have proven to be excellent targets for small molecule in pharmaceutical industry. The immense importance of chemokines and their receptors motivated us to develop a support vector machine-based method ChemoPred to predict this important class of proteins and further classify them into subfamilies. ChemoPred is capable of predicting chemokines and chemokine receptors with an accuracy of 95.08% and 92.19%, respectively. The overall accuracy of classification of chemokines into three subfamilies was 96.00% and that of chemokine receptors into three families was 92.87%. The server ChemoPred is freely available at www.imtech.res.in/raghava/chemopred.

Mining the genome sequence for novel enzyme activity: characterisation of an unusual member of the hormone-sensitive lipase family of esterases from the genome of Streptomyces coelicolor A3 (2)

Soror, S. H., Rao, R., Cullum, J. — 2009-06-01

The sequence of the estA gene (locus SCO 7131) of S. coelicolor A3 (2) suggested that it might differ in substrate specificity from other characterised members of the hormone-sensitive lipase (HSL) family of lipases and esterases. This difference may be attributed to the unique substitutions within the conserved motifs of the family. There was no homologue with any other lipase or esterase to estA in the chromosome of S. avermitilis or other streptomyces species and the sequence showed differences in the conserved motifs from the characterised members of the HSL family. The gene was cloned and expressed as a His-tagged protein in Escherichia coli. The purified enzyme was an esterase, which hydrolyzed the acetate ester of p-nitrophenol, but had little activity on esters with longer side chains, unlike the other characterised bacterial members of the HSL family which showed optimal activity against caproate (C₆) esters. Site-directed mutagenesis was used to change two amino acids to the consensus for the HSL family. This increased the activity against butyrate and caproate esters. The changes also affected thermostability: in one case increasing stability and in the other case reducing it. A profile was constructed for the HSL family and used to detect 119 members in the protein database. The location of conserved amino acid motifs in a 3-D homology model of the enzyme identified further members of the family with unusual amino acid replacements.

The positively charged C-terminal region of the inactivating Shaker B peptide binds to the potassium channel KcsA

Neira, J. L. — 2009-06-01

K⁺ channels are universally involved in electrical activity in muscles and nerves, and also in regulating salt and water transport in tissues implicated in metabolism. The prokaryotic KcsA K⁺ channel has become a structural archetype for the pore domain of voltage-dependent channels. The binding of the inactivating peptide from the eukaryotic Shaker B K⁺ channel (ShB peptide) to either asolectin-reconstituted or DDM-solubilised KcsA has been shown to occur mainly through the hydrophobic region of the peptide (namely, residues Val4, Tyr8, Leu7 and Leu10). In this work, we studied the binding of a deletion variant of the ShB peptide, where the first 11 residues, and then, the hydrophobic region, have been removed ((1–11)ShB). The aim of this work is to elucidate whether binding to KcsA can also occur through the highly charged C-terminal region of ShB peptide. The STD-NMR experiments indicate that there is binding of (1–11)ShB to either asolectin-reconstituted or DDM-solubilised KcsA. The protons showing the largest effects are those of the side-chain of His16, and probably, the backbone amide protons of both Lys18 and Lys19. These results indicate that the hydrophobic residues in ShB peptide are not necessary to ensure binding to the channel, and then, binding to KcsA is also driven by electrostatic interactions.

Molecular dynamics studies on the interactions of PTP1B with inhibitors: from the first phosphate-binding site to the second one

Wang, J.-F., Gong, K., Wei, D.-Q., Li, Y.-X., Chou, K.-C. — 2009-06-01

Protein tyrosine phosphatases 1B (PTP1B) is a major negative regulator of both insulin and leptin signaling pathways. In view of this, it becomes an important target for drug development against cancers, diabetes and obesity. The aim of the current study is to use the long time-scale molecular dynamics (MD) simulations to investigate the structural and dynamic factors that cause its inhibition by INTA and INTB, the two most potent and highly selective PTP1B inhibitors known so far. In order to investigate the mode of collective motions that is vitally important to the biological function, the covariance matrix of C atoms was introduced for performing the dynamic analysis of the inhibition systems. It has been observed that the conformational and dynamic features of WPD-Loop, R-Loop and S-Loop play a key role in providing a smooth entrance for the inhibitors moving into the binding pocket as well as a favorable microenvironment to stabilize them. Furthermore, the hydrogen bonding networks formed around the active site with INTA and INTB may be the main reason of why the inhibition of PTP1B by the two ligands is so potent and selective. All these findings might provide useful insights for developing novel and effective drugs to treat cancer, diabetes and obesity.

Selection and characterization of DARPins specific for the neurotensin receptor 1

Milovnik, P., Ferrari, D., Sarkar, C. A., Pluckthun, A. — 2009-06-01

We describe here the selection and characterization of designed ankyrin repeat proteins (DARPins) that bind specifically to the rat neurotensin receptor 1 (NTR1), a G-protein coupled receptor (GPCR). The selection procedure using ribosome display and the initial clone analysis required <10 µg of detergent-solubilized, purified NTR1. Complex formation with solubilized GPCR was demonstrated by ELISA and size-exclusion chromatography; additionally, the GPCR could be detected in native membranes of mammalian cells using fluorescence microscopy. The main binding epitope in the GPCR lies within the 33 amino acids following the seventh transmembrane segment, which comprise the putative helix 8, and additional binding interactions are possibly contributed by the cytoplasmic loop 3, thus constituting a discontinuous epitope. Since the selected binders recognize the GPCR both in detergent-solubilized and in membrane-embedded forms, they will be potentially useful both in co-crystallization trials and for signal transduction experiments.

Engineering the glycosaminoglycan-binding affinity, kinetics and oligomerization behavior of RANTES: a tool for generating chemokine-based glycosaminoglycan antagonists

Brandner, B., Rek, A., Diedrichs-Mohring, M., Wildner, G., Kungl, A. J. — 2009-06-01

Binding to glycosaminoglycans (GAGs) is a necessary prerequisite for the biological activity of the proinflammatory chemokine RANTES in vivo. We have applied protein engineering methods to modulate equilibrium-binding affinity as well as binding kinetics of RANTES towards its GAG ligand which also altered the chemokine’s oligomerization behavior. Out of 10 mutants, A22K and H23K were chosen for further in vitro and in vivo characterization because their stability was comparable with wild-type (wt) RANTES. In chemical cross-linking experiments, A22K gave higher and H23K lower molecular weight aggregates compared with wtRANTES as shown on SDS–PAGE. All mutants contained an N-terminal methionine residue, a well-described G-protein-coupled receptor (GPCR) antagonistic modification, which resulted in the mutants’ inability to induce monocyte chemotaxis. In surface plasmon resonance experiments using immobilized heparan sulfate (HS) and physiological buffer conditions, Met-RANTES exhibited a significantly longer residual time on the GAG chip compared with the other RANTES variants. In Scatchard plot analysis, RANTES gave a bi-phasic, bell-shaped curve suggesting ‘creation’ of ligand-binding sites on the protein during HS interaction. This was not observed in the mutants’ Scatchard plots which gave K_d values of 317.5 and 44.5 nM for the A22K and H23K mutants, respectively. The mutants were subsequently tested for their inhibitory effect in a rat model of autoimmune uveitis where only H23K exhibited a transient improvement of the clinical disease score. H23K is therefore proposed to be a GPCR-inactive GAG antagonist which displaces the wt chemokine from its natural HS-proteoglycan co-receptor. The protein engineering approach presented here opens new ways for the treatment of RANTES-related diseases.

Site-directed mutagenesis to probe catalysis by a Thermobifida fusca {beta}-1,3-glucanase (Lam81A)

McGrath, C. E., Vuong, T. V., Wilson, D. B. — 2009-06-01

Thermobifida fuscaLam81A is a single domain family-81 β-1,3-endoglucanase, but no structure is known for this family. Site-directed mutagenesis of 14 conserved residues chosen from sequence alignments was used to identify those with critical roles in catalysis, binding or substrate specificity. Mutant enzymes were assayed for their ability to bind and hydrolyze substrates with various glycosyl linkages. Residues D422, E499 and E503 were candidates for the catalytic acid or catalytic base, and E499 was shown to be the catalytic base by azide rescue. F425 was shown to have a major role in substrate binding possibly mediated by aromatic ring stacking with the sugar substrate. In addition, mutation of D424 to histidine altered the substrate specificity by increasing the rate of cleavage of mixed-linkage β-glucan and carboxymethyl-cellulose, 60- and 16-fold, respectively, over the wild-type enzyme.

Message from the Editors

Bradbury, A. — 2009-06-01

The C-terminal domain of the HIV-1 Vif protein is natively unfolded in its unbound state

Reingewertz, T. H., Benyamini, H., Lebendiker, M., Shalev, D. E., Friedler, A. — 2009-04-13

The human immunodeficiency virus type-1 (HIV-1) Vif protein neutralizes the cellular defense mechanism against the virus. The C-terminal domain of Vif (CTD, residues 141–192) mediates many of its interactions. Full-length Vif is difficult to purify in large amounts, hence the only available structure of Vif is of residues 140–155 within the ElonginBC complex. Other structural information, derived from modeling and indirect experiments, indicates that the Vif CTD may be unstructured. Here, we chemically synthesized the Vif CTD using pseudo-proline-building blocks, studied its solution structure in the unbound state using biophysical techniques and found that it is unstructured under physiological conditions. The circular dichroism (CD) spectrum of Vif CTD showed a pattern of random coil with residual helical structure. The ¹⁵N-HSQC nuclear magnetic resonance (NMR) spectrum was characteristic of natively unfolded peptides. Vif CTD eluted from an analytical gel filtration column earlier than expected, indicating an extended conformation. Disorder predictions found the CTD to be unstructured, in agreement with our experimental results. CD experiments showed that Vif CTD underwent a conformational change upon interacting with membrane-mimicking DPC micelles, but not upon binding to a peptide derived from its binding region in ElonginC. Our results provide direct evidence for the unfolded structure of the free Vif CTD and indicate that it may gain structure upon binding its natural ligands.

Detection of the protein dimers, multiple monomeric states and hydrated forms of Plasmodium falciparum triosephosphate isomerase in the gas phase

2009-04-13

Dimeric and monomeric forms of the enzyme triosephosphate isomerase (TIM) from Plasmodium falciparum (Pf) have been detected under conditions of nanoflow by electrospray mass spectrometry. The dimer (M = 55 663 Da) exhibits a narrow charge state distribution with intense peaks limited to values of 18⁺ to 21⁺, maximal intensity being observed for charge states 19⁺ and 20⁺. A monomeric species with a charge state distribution ranging from 11⁺ to 16⁺ is also observed, which may be assigned to folded dissociated subunits. Complete dimer dissociation results under normal electrospray condition. The effects of solution pH and source temperature have been investigated. The observation of four distinct charge state distributions which may be assigned to a dimer, folded monomer, partially folded monomer and unfolded monomer is reported. Circular dichromism and fluorescence studies of Pf TIM at low pH support the retention of substantial secondary and tertiary structures. Satellite peaks in mass spectra corresponding to hydrated species are also observed and isotope shift upon deuteration is demonstrated. The analysis of all available independent crystal structures of Pf TIM and TIMs from other organisms permits identification of structurally conserved water molecules. Hydration observed in the dimer and folded monomeric forms in the gas phase may correspond to these conserved sites.

Humanized-monoclonal antibody against heterologous Leptospira infection

2009-04-13

Patients with leptospirosis are commonly treated with antibiotics. Jarisch–Herxheimer reaction caused by toxic bacterial substances massively released as a result of the antibiotic mediated-bacterial lysis occurs in some patients which may aggravate the existing severe clinical manifestations. In this study, a humanized-murine single-chain monoclonal antibody (HuScFv) was produced and tested as an alternative of antibiotics for treatment of leptospirosis. Complementary DNA was prepared from total RNA of a murine hybridoma clone secreting monoclonal antibody (MAb) specific to LipL32 of pathogenic Leptospira spp. The MAb had therapeutic efficacy in Leptospira challenged hamsters. The VH and VL coding sequences were amplified using the cDNA as a template. The sequences were linked to form a single-chain variable murine DNA fragment (muscFv). CDR sequences of the muscFv were grafted onto the best matching human VH and VL immunoglobulin frameworks. After cloning of the humanized murine DNA sequences (huscFv) into a phagemid vector and the vector was introduced into competent Escherichia coli, the HuScFv was produced. On the same weight basis, the HuScFv possessed equal neutralizing activities to the murine ScFv counterpart against heterologous Leptospira-mediated hemolysis in vitro and rescued hamsters from a heterologous Leptospira lethal challenge. The HuScFv antibody has high therapeutic potential as an alternative to antibiotics for human leptospirosis, especially for drug hypersensitive patients.

Directed evolution of an extremely stable fluorescent protein

Kiss, C., Temirov, J., Chasteen, L., Waldo, G. S., Bradbury, A. R.M. — 2009-04-13

In this paper we describe the evolution of eCGP123, an extremely stable green fluorescent protein based on a previously described fluorescent protein created by consensus engineering (CGP: consensus green protein). eCGP123 could not be denatured by a standard thermal melt, preserved almost full fluorescence after overnight incubation at 80°C and possessed a free energy of denaturation of 12.4 kcal/mol. It was created from CGP by a recursive process involving the sequential introduction of three destabilizing heterologous inserts, evolution to overcome the destabilization and finally ‘removal’ of the destabilizing insert by gene synthesis. We believe that this approach may be generally applicable to the stabilization of other proteins.

Engineering an ultra-stable affinity reagent based on Top7

2009-04-13

Antibodies are widely used for diagnostic and therapeutic applications because of their sensitive and specific recognition of a wide range of targets; however, their application is limited by their structural complexity. More demanding applications require greater stability than can be achieved by immunoglobulin-based reagents. Highly stable, protein-based affinity reagents are being investigated for this role with the goal of identifying a suitable scaffold that can attain specificity and sensitivity similar to that of antibodies while performing under conditions where antibodies fail. We have engineered Top7—a highly stable, computationally designed protein—to specifically bind human CD4 by inserting a peptide sequence derived from a CD4-specific antibody. Molecular dynamics simulations were used to evaluate the structural effect of the peptide insertion at a specific site within Top7 and suggest that this Top7 variant retains conformational stability over 100°C. This engineered protein specifically binds CD4 and, consistent with simulations, is extremely resistant to thermal and chemical denaturation—retaining its secondary structure up to at least 95°C and requiring 6 M guanidine to completely unfold. This CD4-specific protein demonstrates the functionality of Top7 as a viable scaffold for use as a general affinity reagent which could serve as a robust and inexpensive alternative to antibodies.