Computational design of a specific heavy chain/κ light chain interface for expressing fully IgG bispecific antibodies.

The use of bispecific antibodies (BsAbs) to treat human diseases is on the rise. Increasingly complex and powerful therapeutic mechanisms made possible by BsAbs are spurring innovation of novel BsAb formats and methods for their production. The long-lived in vivo pharmacokinetics, optimal biophysical properties and potential effector functions of natural IgG monoclonal (and monospecific) antibodies has resulted in a push to generate fully IgG BsAb formats with the same quaternary structure as monoclonal IgGs. The production of fully IgG BsAbs is challenging because of the highly heterogeneous pairing of heavy chains (HCs) and light chains (LCs) when produced in mammalian cells with two IgG HCs and two LCs. A solution to the HC heterodimerization aspect of IgG BsAb production was first discovered two decades ago; however, addressing the LC mispairing issue has remained intractable until recently. Here, we use computational and rational engineering to develop novel designs to the HC/LC pairing issue, and particularly for κ LCs. Crystal structures of these designs highlight the interactions that provide HC/LC specificity. We produce and characterize multiple fully IgG BsAbs using these novel designs. We demonstrate the importance of specificity engineering in both the variable and constant domains to achieve robust HC/LC specificity within all the BsAbs. These solutions facilitate the production of fully IgG BsAbs for clinical use.

Engineering and Application of LOV2-Based Photoswitches.

Cellular optogenetic switches, a novel class of biological tools, have improved our understanding of biological phenomena that were previously intractable. While the design and engineering of these proteins has historically varied, they are all based on borrowed elements from plant and bacterial photoreceptors. In general terms, each of the optogenetic switches designed to date exploits the endogenous light-induced change in photoreceptor conformation while repurposing its effect to target a different biological phenomenon. We focus on the well-characterized light-oxygen-voltage 2 (LOV2) domain from Avena sativa phototropin 1 as our cornerstone for design. While the function of the LOV2 domain in the context of the phototropin protein is not fully elucidated, its thorough biophysical characterization as an isolated domain has created a strong foundation for engineering of photoswitches. In this chapter, we examine the biophysical characteristics of the LOV2 domain that may be exploited to produce an optogenetic switch and summarize previous design efforts to provide guidelines for an effective design. Furthermore, we provide protocols for assays including fluorescence polarization, phage display, and microscopy that are optimized for validating, improving, and using newly designed photoswitches.

Design of structurally distinct proteins using strategies inspired by evolution.

Natural recombination combines pieces of preexisting proteins to create new tertiary structures and functions. We describe a computational protocol, called SEWING, which is inspired by this process and builds new proteins from connected or disconnected pieces of existing structures. Helical proteins designed with SEWING contain structural features absent from other de novo designed proteins and, in some cases, remain folded at more than 100°C. High-resolution structures of the designed proteins CA01 and DA05R1 were solved by x-ray crystallography (2.2 angstrom resolution) and nuclear magnetic resonance, respectively, and there was excellent agreement with the design models. This method provides a new strategy to rapidly create large numbers of diverse and designable protein scaffolds.

Adaptation to high salinity in poplar involves changes in xylem anatomy and auxin physiology.

To investigate the physiological basis of salt adaptation in poplar, we compared the effect of salt stress on wood anatomy and auxin physiology of the salt-resistant Populus euphratica and salt-sensitive Populus x canescens. Both poplar species showed decreases in vessel lumina associated with increases in wall strength in response to salt, however, in P. euphratica at three-fold higher salt concentrations than in P. x canescens. The predicted hydraulic conductivity of the wood formed under salt stress decreased in P. x canescens, while in P. euphratica, no significant effects of salt on conductivity and transpiration were observed. The concentration of free indole-3-acetic acid (IAA) decreased under salt stress in the xylem of both poplar species, but to a larger extent in P. x canescens than in P. euphratica. Only salt-treated P. euphratica exhibited an increase in IAA-conjugates in the xylem. Genes homologous to the auxin-amidohydrolase ILL3 were isolated from the xylems of P. euphratica and P. x canescens. For functional analysis, the auxin-amidohydrolase from P. x canescens was overexpressed in Arabidopsis. Transgenic Arabidopsis plants were more resistant to salt stress than the wild-type plants. Increased sensitivity of the transgenic Arabidopsis to IAA-Leu showed that the encoded hydrolase used IAA-Leu as a substrate. These results suggest that poplar can use IAA-amidoconjugates in the stem as a source of auxin to balance the effects of salt stress on auxin physiology.

Conversion of monomeric protein L to an obligate dimer by computational protein design.

Protein L consists of a single alpha-helix packed on a four-stranded beta-sheet formed by two symmetrically opposed beta-hairpins. We use a computer-based protein design procedure to stabilize a domain-swapped dimer of protein L in which the second beta-turn straightens and the C-terminal strand inserts into the beta-sheet of the partner. The designed obligate dimer contains three mutations (A52V, N53P, and G55A) and has a dissociation constant of approximately 700 pM, which is comparable to the dissociation constant of many naturally occurring protein dimers. The structure of the dimer has been determined by x-ray crystallography and is close to the in silico model.

Computer-based redesign of a protein folding pathway.

A fundamental test of our current understanding of protein folding is to rationally redesign protein folding pathways. We use a computer-based design strategy to switch the folding pathway of protein G, which normally involves formation of the second, but not the first, beta-turn at the rate limiting step in folding. Backbone conformations and amino acid sequences that maximize the interaction density in the first beta-hairpin were identified, and two variants containing 11 amino acid replacements were found to be approximately 4 kcal mol-1 more stable than wild type protein G. Kinetic studies show that the redesigned proteins fold approximately 100 x faster than wild type protein and that the first beta-turn is formed and the second disrupted at the rate limiting step in folding.

Native protein sequences are close to optimal for their structures.

How large is the volume of sequence space that is compatible with a given protein structure? Starting from random sequences, low free energy sequences were generated for 108 protein backbone structures by using a Monte Carlo optimization procedure and a free energy function based primarily on Lennard-Jones packing interactions and the Lazaridis-Karplus implicit solvation model. Remarkably, in the designed sequences 51% of the core residues and 27% of all residues were identical to the amino acids in the corresponding positions in the native sequences. The lowest free energy sequences obtained for ensembles of native-like backbone structures were also similar to the native sequence. Furthermore, both the individual residue frequencies and the covariances between pairs of positions observed in the very large SH3 domain family were recapitulated in core sequences designed for SH3 domain structures. Taken together, these results suggest that the volume of sequence space optimal for a protein structure is surprisingly restricted to a region around the native sequence.

Effects of varying the local propensity to form secondary structure on the stability and folding kinetics of a rapid folding mixed alpha/beta protein: characterization of a truncation mutant of the N-terminal domain of the ribosomal protein L9.

The N-terminal domain of the ribosomal protein L9 forms a split betaalphabeta structure with a long C-terminal helix. The folding transitions of a 56 residue version of this protein have previously been characterized, here we report the results of a study of a truncation mutant corresponding to residues 1-51. The 51 residue protein adopts the same fold as the 56 residue protein as judged by CD and two-dimensional NMR, but it is less stable as judged by chemical and thermal denaturation experiments. Studies with synthetic peptides demonstrate that the C-terminal helix of the 51 residue version has very little propensity to fold in isolation in contrast to the C-terminal helix of the 56 residue variant. The folding rates of the two proteins, as measured by stopped-flow fluorescence, are essentially identical, indicating that formation of local structure in the C-terminal helix is not involved in the rate-limiting step of folding.

Folding of the multidomain ribosomal protein L9: the two domains fold independently with remarkably different rates.

The folding and unfolding behavior of the multidomain ribosomal protein L9 from Bacillus stearothermophilus was studied by a novel combination of stopped-flow fluorescence and nuclear magnetic resonance (NMR) spectroscopy. One-dimensional 1H spectra acquired at various temperatures show that the C-terminal domain unfolds at a lower temperature than the N-terminal domain (Tm = 67 degrees C for the C-terminal domain, 80 degrees C for the N-terminal domain). NMR line-shape analysis was used to determine the folding and unfolding rates for the N-terminal domain. At 72 degrees C, the folding rate constant equals 2980 s-1 and the unfolding rate constant equals 640 s-1. For the C-terminal domain, saturation transfer experiments performed at 69 degrees C were used to determine the folding rate constant, 3.3 s-1, and the unfolding rate constant, 9.0 s-1. Stopped-flow fluorescence experiments detected two resolved phases: a fast phase for the N-terminal domain and a slow phase for the C-terminal domain. The folding and unfolding rate constants determined by stopped-flow fluorescence are 760 s-1 and 0.36 s-1, respectively, for the N-terminal domain at 25 degrees C and 3.0 s-1 and 0.0025 s-1 for the C-terminal domain. The Chevron plots for both domains show a V-shaped curve that is indicative of two-state folding. The measured folding rate constants for the N-terminal domain in the intact protein are very similar to the values determined for the isolated N-terminal domain, demonstrating that the folding kinetics of this domain is not affected by the rest of the protein. The remarkably different rate constants between the N- and C-terminal domains suggest that the two domains can fold and unfold independently. The folding behavior of L9 argues that extremely rapid folding is not necessarily functionally important.

pKa values and the pH dependent stability of the N-terminal domain of L9 as probes of electrostatic interactions in the denatured state. Differentiation between local and nonlocal interactions.

pKa values were measured for the 6 carboxylates in the N-terminal domain of L9 (NTL9) by following NMR chemical shifts as a function of pH. The contribution of each carboxylate to the pH dependent stability of NTL9 was estimated by comparing the pKa values for the native and denatured state of the protein. A set of peptides with sequences derived from NTL9 were used to model the denatured state. In the protein fragments, the pKa values measured for the aspartates varied between 3.8 and 4.1 and the pKa values measured for the glutamates varied between 4.1 and 4.6. These results indicate that the local sequence can significantly influence pKa values in the denatured state and highlight the difficulties in using standard pKa values derived from small compounds. Calculations based on the measured pKa values suggest that the free energy of unfolding of NTL9 should decrease by 4.4 kcal mol-1 when the pH is lowered from 6 to 2. In contrast, urea and thermal denaturation experiments indicate that the stability of the protein decreases by only 2.6 kcal mol-1 when the carboxylates are protonated. This discrepancy indicates that the protein fragments are not a complete representation of the denatured state and that nonlocal sequence effects perturb the pKa's in the denatured state. Increasing the salt concentration from 100 to 750 mM NaCl removes the discrepancy between the stabilities derived from denaturation experiments and the stability changes calculated from the pKa values. At high concentrations of salt there is also less variation of the pKa values measured in the protein fragments. Our results argue that in the denatured state of NTL9 there are electrostatic interactions between groups both local and nonlocal in primary sequence.

Global analysis of the thermal and chemical denaturation of the N-terminal domain of the ribosomal protein L9 in H2O and D2O. Determination of the thermodynamic parameters, deltaH(o), deltaS(o), and deltaC(o)p and evaluation of solvent isotope effects.

The stability of the N-terminal domain of the ribosomal protein L9, NTL9, from Bacillus stearothermophilus has been monitored by circular dichroism at various temperatures and chemical denaturant concentrations in H2O and D2O. The basic thermodynamic parameters for the unfolding reaction, deltaH(o), deltaS(o), and deltaC(o)p, were determined by global analysis of temperature and denaturant effects on stability. The data were well fit by a model that assumes stability varies linearly with denaturant concentration and that uses the Gibbs-Helmholtz equation to model changes in stability with temperature. The results obtained from the global analysis are consistent with information obtained from individual thermal and chemical denaturations. NTL9 has a maximum stability of 3.78 +/- 0.25 kcal mol(-1) at 14 degrees C. DeltaH(o)(25 degrees C) for protein unfolding equals 9.9 +/- 0.7 kcal mol(-1) and TdeltaS(o)++(25 degrees C) equals 6.2 +/- 0.6 kcal mol(-1). DeltaC(o)p equals 0.53 +/- 0.06 kcal mol(-1) deg(-1). There is a small increase in stability when D2O is substituted for H2O. Based on the results from global analysis, NTL9 is 1.06 +/- 0.60 kcal mol(-1) more stable in D2O at 25 degrees C and Tm is increased by 5.8 +/- 3.6 degrees C in D2O. Based on the results from individual denaturation experiments, NTL9 is 0.68 +/- 0.68 kcal mol(-1) more stable in D2O at 25 degrees C and Tm is increased by 3.5 +/- 2.1 degrees C in D2O. Within experimental error there are no changes in deltaH(o) (25 degrees C) when D2O is substituted for H2O.

Amide proton exchange measurements as a probe of the stability and dynamics of the N-terminal domain of the ribosomal protein L9: comparison with the intact protein.

Amide H/D exchange rates have been measured for the N-terminal domain of the ribosomal protein L9, residues 1-56. The rates were measured at pD 3.91, 5.03, and 5.37. At pD 5.37, 18 amides exchange slowly enough to give reliable rate measurements. At pD 3.91, seven additional residues could be followed. The exchange is shown to occur by the EX2 mechanism for all conditions studied. The rates for the N-terminal domain are very similar to those previously measured for the corresponding region in the full-length protein (Lillemoen J et al., 1997, J Mol Biol 268:482-493). In particular, the rates for the residues that we have shown to exchange via global unfolding in the N-terminal domain agree within the experimental error with the rates measured by Hoffman and coworkers, suggesting that the structure of the domain is preserved in isolation and that the stability of the isolated domain is comparable to the stability of this domain in intact L9.

Cooperative folding of a protein mini domain: the peripheral subunit-binding domain of the pyruvate dehydrogenase multienzyme complex.

The peripheral subunit-binding domain from the dihydrolipoamide acetyltransferase (E2) component of the pyruvate dehydrogenase multienzyme complex from Bacillus stearothermophilus is stably folded, despite its short sequence of only 43 amino acid residues. A 41 residue peptide derived from this domain, psbd41, undergoes a cooperative thermal unfolding transition with a tm of 54 degrees C. This three-helix protein is monomeric as judged by ultracentrifugation and concentration-dependent CD measurements. Peptides corresponding to the individual helices are largely unstructured both alone and in combination, indicating that the unusual stability of this protein does not arise solely from unusually stable alpha-helices. Chemical denaturation by guanidine hydrochloride is also cooperative with a delta GH2O of 3.1 kcal mol-1 at pH 8.0 and 25 degrees C. The chemical denaturation is broad with an m-value of 760 cal mol-1 M-1. psbd41 contains a buried aspartate residue at position 34 that may provide stability and specificity to the fold. A mutant peptide, psbd41Asn was synthesized in which the buried aspartate residue was mutated to asparagine. This peptide still folds cooperatively and it is monomeric, but is much less thermostable than the wild-type with a tm of only 31 degrees C. Chemical denaturations at 4 degrees C give an m-value of 740 cal mol-1 M-1, similar to the wild-type, but the stability delta GH2O is only 1.4 kcal mol-1. Both the wild-type and the mutant unfold at extremes of pH, but at 4 degrees C psbd41Asn is folded over a narrower pH range than the wild-type. Although the mutant unfolds cooperatively by thermal and by chemical denaturation, its NMR spectrum is significantly broader than that of the wild-type and it binds ANS. These results show that Asp34 is vital for the stability and specificity of this structure, the second smallest natural sequence known to fold in the absence of disulfide bonds or metal or ligand-binding sites.

Structure and stability of the N-terminal domain of the ribosomal protein L9: evidence for rapid two-state folding.

The N-terminal domain, residues 1-56, of the ribosomal protein L9 has been chemically synthesized. The isolated domain is monomeric as judged by analytical ultracentrifugation and concentration-dependent CD. Complete 1H chemical shift assignments were obtained using standard methods. 2D-NMR experiments show that the isolated domain adopts the same structure as seen in the full-length protein. It consists of a three-stranded antiparallel beta-sheet sandwiched between two helixes. Thermal and urea unfolding transitions are cooperative, and the unfolding curves generated from different experimental techniques, 1D-NMR, far-UV CD, near-UV CD, and fluorescence, are superimposable. These results suggest that the protein folds by a two-state mechanism. The thermal midpoint of folding is 77 +/- 2 degrees C at pD 8.0, and the domain has a delta G degree folding = 2.8 +/- 0.8 kcal/mol at 40 degrees C, pH 7.0. Near the thermal midpoint of the unfolding transition, the 1D-NMR peaks are significantly broadened, indicating that folding is occurring on the intermediate exchange time scale. The rate of folding was determined by fitting the NMR spectra to a two-state chemical exchange model. Similar folding rates were measured for Phe 5, located in the first beta-strand, and for Tyr 25, located in the short helix between strands two and three. The domain folds extremely rapidly with a folding rate constant of 2000 s-1 near the midpoint of the equilibrium thermal unfolding transition.

Global analysis of the effects of temperature and denaturant on the folding and unfolding kinetics of the N-terminal domain of the protein L9.

The folding and unfolding kinetics of the N-terminal domain of the ribosomal protein L9 have been measured at temperatures between 7 and 85 degrees C and between 0 and 6 M guanidine deuterium chloride. Stopped-flow fluorescence was used to measure rates below 55 degrees C and NMR lineshape analysis was used above 55 degrees C. The amplitudes and rate profiles of the stopped-flow fluorescence experiments are consistent with a two-state folding mechanism, and plots of ln(k) versus guanidine deuterium chloride concentration show the classic v-shape indicative of two-state folding. There is no roll over in the plots when the experiments are repeated in the presence of 400 mM sodium sulfate. Temperature and denaturant effects were fit simultaneously to the simple model k=D exp(-DeltaG*/RT) where DeltaG* represents the change in apparent free energy between the transition state and the folded or unfolded state and D represents the maximum possible folding speed. DeltaG* is assumed to vary linearly with denaturant concentration and the Gibbs-Helmholtz equation is used to model stability changes with temperature. Approximately 60% of the surface area buried upon folding is buried in the transition state as evidenced by changes in the heat capacity and m value between the unfolded state and the transition state. The equilibrium thermodynamic parameters, DeltaCp degrees, m and DeltaG degrees, all agree with the values calculated from the kinetic experiments, providing additional evidence that folding is two-state. The folding rates at 0 M guanidine hydrochloride show a non-Arrhenius temperature dependence typical of globular proteins. When the folding rates are examined along constant DeltaG degrees/T contours they display an Arrhenius temperature dependence with a slope of -8600 K. This indicates that for this system, the non-Arrhenius temperature dependence of folding can be accounted for by the anomalous temperature dependence of the interactions which stabilize proteins.

An exceptionally stable helix from the ribosomal protein L9: implications for protein folding and stability.

The ribosomal protein L9 has an unusual structure comprising two compact globular domains connected by a 34 residue alpha-helix. The middle 17 residues of the helix are exposed to solvent while the first seven pack against and form part of the N-terminal domain, and the last ten form part of the C-terminal domain. Here we report results which show that a peptide corresponding to the central helix of L9 is monomeric in aqueous solution and >85% helical at 1 degrees C and 68(+/-7)% helical at 25 degrees C. This is considerably more helical than any other protein fragment studied to date. Another peptide corresponding to the middle 17 residues of the helix is monomeric and is 41(+/-4)% helical at 1 degrees C. Because the central helix has high intrinsic stability the globular N and C-terminal domains will likely be stabilized by their interactions with the helix. Therefore, the stability of the two terminal domains should not be completely independent because both domains gain stability from a shared structural element, the central helix. Also, the ability of the central helix to form native-like structure in isolation highlights a potential role for the helix in the early stages of the folding process.

Calcium binding peptides from alpha-lactalbumin: implications for protein folding and stability.

The calcium binding protein alpha-lactalbumin folds via a molten globule intermediate. Calcium does not bind strongly to the unfolded protein or the molten globule, but does bind to the transition state between the molten globule and the native protein. Of interest are the structures formed in the transition state that promote calcium binding. To study the importance of local secondary structure on calcium binding, we have synthesized two peptides corresponding to the calcium binding site that include the flanking C-helix and 3(10)-helix. The first peptide, elbow-A, consists of residues 72-100 from bovine alpha-lactalbumin, but with Cys 73, Cys 77, and Cys 91 replaced by alanines. In the second peptide, denoted elbow, the cysteines at position 73 and 91 are included and the nativelike disulfide bond is formed. Both peptides are monomeric and unstructured in aqueous solution and bind calcium weakly with apparent K(d)'s on the order of 10(-2) M. In 50% trifluoroethanol (v/v), the peptides are 45% helical as judged by CD. NMR studies performed on elbow and elbow-A in TFE indicate that the helical structure is confined to the C-helix. In this solvent system elbow binds calcium one-to-one with a K(d) of 50 microM. Removing the disulfide bond reduces, but does not eliminate calcium binding (K(d) = 170 microM in 50% TFE). These results suggest that formation of the C-helix promotes calcium binding and may be a key determinant of calcium binding in the transition state.