Skip to main content
NCBI home page
Protein Science : A Publication of the Protein Society logoLink to Protein Science : A Publication of the Protein Society
. 2015 Mar 25;24(5):841–849. doi: 10.1002/pro.2657

Aromatic cluster mutations produce focal modulations of β-sheet structure

Matthew Biancalana 1, Koki Makabe 1,2, Shude Yan 1, Shohei Koide 1,*
PMCID: PMC4420532  PMID: 25645104

Abstract

Site-directed mutagenesis is a powerful tool for altering the structure and function of proteins in a focused manner. Here, we examined how a model β-sheet protein could be tuned by mutation of numerous surface-exposed residues to aromatic amino acids. We designed these aromatic side chain “clusters” at highly solvent-exposed positions in the flat, single-layer β-sheet of Borrelia outer surface protein A (OspA). This unusual β-sheet scaffold allows us to interrogate the effects of these mutations in the context of well-defined structure but in the absence of the strong scaffolding effects of globular protein architecture. We anticipated that the introduction of a cluster of aromatic amino acid residues on the β-sheet surface would result in large conformational changes and/or stabilization and thereby provide new means of controlling the properties of β-sheets. Surprisingly, X-ray crystal structures revealed that the introduction of aromatic clusters produced only subtle conformational changes in the OspA β-sheet. Additionally, despite burying a large degree of hydrophobic surface area, the aromatic cluster mutants were slightly less stable than the wild-type scaffold. These results thereby demonstrate that the introduction of aromatic cluster mutations can serve as a means for subtly modulating β-sheet conformation in protein design.

Keywords: protein design, conformational stability, beta-sheet formation, secondary structure

Introduction

The design and analysis of protein architectures provide a critical test of our understanding of the fundamental principles of protein architecture and stability. Altering the surface of an existing protein scaffold through systematic mutation is a powerful but conceptually simple design strategy, in which surface-exposed residues are mutated to modulate a protein's functional or biophysical properties.1 For instance, the “super-charging” protein design strategy is a striking examples of surface engineering, in which a protein is imparted with an unusually high net charge by mutating a large number of surface-exposed residues to amino acids sharing the same charge (e.g., Glu and Asp).2 This method was used to engineer GFP variants that were resistant to aggregation over a wide range of conditions,2 as well as highly-soluble and stable variants of enteropeptidase.3 An additional method termed “surface entropy reduction” replaces high-entropy side chains such as Lys and Glu with comparatively lower-entropy Ser residues in order to facilitate protein crystallization.4,5 Together, these protein design strategies demonstrate how selective mutation of a subset of surface-exposed residues can be used to modulate the behavior of the folded polypeptide.

We questioned whether the surface-engineering strategy could be applied more broadly to the replacement of surface-exposed residues with aromatic residues, thereby exploiting the additive or cooperative effects of multiple bulky side chains sharing similar properties. Unlike the unbranched amino acids used in the above-mentioned examples, aromatic amino acids, with their bulky side chains, are likely to produce substantial interactions among adjacent residues, which may in turn modulate the backbone conformation. As β-sheets are excellent platforms for creating stable scaffolds, we envisioned that introducing an aromatic cluster might be a viable approach to fine-tune scaffold conformation. Here, we describe a series of mutants of a flat β-sheet protein (discussed below) in which aromatic residues are introduced at a large number of positions that are adjacent in space in the native structure. These mutations create a sterically crowded patch (“cluster”) of aromatic side chains on the protein surface.

Aromatic residues have critical structural roles in many β-rich motifs and globular proteins.6–8 Understanding the contributions of the chemical properties of aromatic amino acids, including their high β-sheet propensity, hydrophobicity, and π-stacking interactions, has been a major focus of peptide models for β-sheet formation.9–11 Particular attention has been paid to the higher-order effects of multiple interacting aromatic residues occurring across β-strands [Fig. 1(A)],12,13 which have been found to possess uniquely favorable self-association energy.12,14,15 Unfortunately, many models used to explore the role of dense aromatic clusters in β-sheet systems are formed from small peptides or peptide mimics with highly malleable backbone conformations.16 It remains to be seen whether we can apply the conclusions drawn from studies on these systems as predictive tools for larger or more complex β-sheets, because stably folded protein systems do not share the great conformational flexibility of peptides.

Figure 1.

Figure 1

Open in a new tab

Design and stability measurements of OspA aromatic cluster mutants. A: Positions for making cluster mutants in β-sheets. A small β-sheet is schematically shown as a connection of α-carbons, with side chains depicted as colored spheres. A 3×3 cluster of residues whose side chains are solvent-exposed on the same face of the β-sheet are shown. B: Positions in the OspA scaffold used in this work, shown as spheres and colored as in A. C: Schematic of the OspA SLB, with positions used for mutagenesis colored as in A. E104 and K107, colored in orange, are mutated to Ser in the “surface mutant” (sm1) constructs used for crystallization. D: Amino acid composition of the aromatic cluster mutants studied in this work. The 3×3 array corresponds to the positions shown in blue, green, and red in the OspA SLB schematic. E: OspA unfolding reaction and experimental detection techniques. The three-state unfolding mechanism of OspA is depicted as native (N), intermediate (I), and unfolded (U) states. The unfolding of the OspA C-terminal domain (native-intermediate transition) is monitored by a change in fluorescence of an intrinsic Trp residue. The unfolding of the C-terminal domain (native-intermediate transition) and unfolding of the SLB remainder of the protein (intermediate-unfolded transition) are both monitored by CD spectroscopy. F: CD (top) and Trp fluorescence (bottom) denaturation data for Y3, F3, and wild-type OspA. G: CD (top) and Trp fluorescence (bottom) denaturation data for Y3x3 sm1, Y3x3W sm1, and OspA sm1 (see Methods).

We have developed a design strategy in which in which a large number of amino acids that are adjacent in the native structure of a model protein are mutated to aromatic residues.13 We employed a large β-sheet scaffold derived from Borrelia outer surface protein A (OspA) [Fig. 1(B)],17 which contains a central flat, single-layer β-sheet (SLB) comprised of two β-hairpins (Strands 7–8 and 9–10) flanked by N- and C-terminal globular domains [Fig. 1(B,C)]. The OspA scaffold is monomeric, water-soluble, crystallization-efficient and highly amenable to site-directed mutagenesis.18 Importantly, the SLB region is not involved in the formation of a hydrophobic core that is the predominant determinant of the structure and stability of common β-sheet proteins. Thus, the OspA system provides a unique opportunity to investigate how the introduction of aromatic residues at numerous adjacent positions can be used to modify the structure of a β-sheet.

Results

Aromatic cluster mutations in the OspA SLB

We designed a series of aromatic clusters within the OspA scaffold that exploit the close spatial proximity of both cross- and intra-strand residues, in hopes of amplifying the effects of local contacts between mutated side chains. We designed and characterized four variants of OspA, termed “Y3”, “F3”, “Y3x3”, and “Y3x3W” [Fig. 1(D)]. The mutated positions are located centrally in the SLB to minimize effects from the adjacent N- and C-terminal domains of OspA. Y3, whose structure has been described previously,13 was designed by introducing three cross-strand Tyr. F3 was designed by introducing three Phe residues in these same positions. Y3x3 was designed by forming a 3x3 cluster of aromatic residues using three adjacent rows of three cross-strand residues on the same face of the β-sheet. Finally, Y3x3W was designed by replacing the central residue of the Y3x3 array with Trp. All proteins were expressed recombinantly in E. coli and purified as monomers, as judged by size-exclusion chromatography.

Effects of aromatic cluster mutations on stability

We characterized the stability of the aromatic cluster mutants using urea denaturation, as previously applied to numerous OspA variants. Wild-type OspA displays a three-state unfolding profile consisting of native, intermediate, and unfolded states [Fig. 1(E)].19 The native-intermediate transition reflects the unfolding of the C-terminal domain, and the intermediate-unfolded transition reflects cooperative unfolding of the SLB and N-terminal domain. In wild-type OspA, the native-intermediate transition is monitored by an increase in Trp fluorescence from the sole endogenous Trp residue located in the OspA C-terminal domain. The overall unfolding reaction is monitored by circular dichroism (CD) spectroscopy. The CD denaturation profile displays a single transition resulting from the sum of two unresolved transitions corresponding to: (1) the unfolding of the C-terminal domain of OspA, followed by (2) subsequent unfolding of the remainder of the polypeptide. By integrating fluorescence and CD data, we can dissect the two transitions and their energetic parameters. Although the OspA variants studied here have many additional aromatic residues, we were able to apply the same method to analyzing most of the mutants. Stability measurements for Y3 and F3 were performed using proteins containing the wild-type N- and C-terminal globular domain sequences of OspA. We obtained better-quality data for the Y3x3 and Y3x3W mutants using the “sm1” proteins employed for crystallization (see Methods), owing to a larger difference in CD signal between the folded and unfolded forms of the proteins. All mutants are compared relative to their respective OspA scaffold, i.e. Y3 and F3 are compared with the wild-type scaffold, whereas Y3x3 sm1 and Y3x3W sm1 are compared with the sm1 scaffold.

Y3 and F3 were slightly less stable than the corresponding wild-type scaffold, as judged by a subtle shift in their fluorescence and CD denaturation profiles [Fig. 1(F)]. We likewise found that the Y3x3 mutations destabilized the scaffold, though these were to an appreciably greater degree [Fig. 1(G)]. Intriguingly, Y3x3W alone displayed stability changes that are partitioned differently between the two unfolding transitions: as judged by the CD denaturation profile, the native state is destabilized, while the intermediate state is stabilized. Unfortunately but not surprisingly, the additional surface-exposed Trp residue in Y3x3W produced a complex pattern of fluorescence data [Fig. 1(G) lower panel], preventing us from applying our standard method for interpreting folding transitions in structural terms.

Structures of the introduced aromatic clusters

To determine how the aromatic residues are accommodated and to clarify the molecular basis for their stability changes, we determined the crystal structures of Y3, F3, Y3x3, and Y3x3W at 1.3 Å, 1.4 Å, 1.7 Å, and 1.6 Å resolutions, respectively [Fig. 2(A), Table1]. The proteins crystallized in the same space group with very similar crystal dimensions, though they have two distinct packing modes. In F3 and Y3x3, packing interactions position the C-terminal domain of OspA in proximity to the SLB of an adjacent molecule [Fig. 2(B)]. These crystal contacts are mediated by only a fraction of the SLB residues, and as judged by their composition are likely opportunistic rather than causative (e.g., the vast majority of contacts with the introduced aromatics are formed by polar Ser/Thr/Asn or positively charged Lys side chains, rather than prototypical protein-protein interactions formed between oppositely-charged or hydrophobic residues). The aromatic clusters in the wild type, Y3, and Y3x3W structures form minimal contacts with adjacent symmetry-related molecules [Fig. 2(B)], obviating crystal packing as a major influence in the structures of these SLB regions. Thus, despite differences in crystal packing, the OspA mutants can be directly compared to dissect structural variations arising from their distinct sets of aromatic mutations. Moreover, these different packing patterns eliminate the possibility that the observed structures are uniformly constrained by similar lattice contacts, further strengthening the significance of any trends or tendencies shared across multiple structures.

Figure 2.

Figure 2

Open in a new tab

Crystal structures of the aromatic cluster mutants of OspA. A: Conformations of the introduced aromatic clusters. The side chains comprising the mutated face of the OspA SLB (top) are shown as sticks. The side chains of the introduced aromatic residues are colored in green and the side chains of neighboring endogenous residues are colored in lavender. The SLB is in the plane of the page, with the N- and C-terminal globular domains above and below, respectively. The globular domains are omitted for clarity in the top row. The structures of the entire OspA proteins are shown (lower row) with aromatic clusters shown as space-filling spheres. The OspA backbone is shown as grey ribbons. The rotamers of the introduced aromatic mutations are shown below. The mutated positions are colored yellow, orange, or red depending on whether each is the first, second, or third backbone-dependent rotamer preference at this position. B: Crystal packing of the aromatic cluster mutants. Left, the packing mode of Y3x3 (and also F3) that involves interaction between the C-terminal globular domain of one OspA molecule with the SLB of another. Right, the packing mode of Y3x3W (as well as wild type and Y3) showing effectively no contacts involving the mutated face of the SLB. C: The Cβ positions of the Tyr side chains of Y3x3 (green) are displaced away from the side chains of their cross-strand neighbors, relative to wild type (gray). This comparison was generated by aligning Strands 7 and 8 of the Y3x3 and wild type structures. The cross-strand side-chain pairs are shown as sticks, while the backbones are shown as lines.

Table 1.

Statistics for Crystal Structures of OspA Aromatic Cluster Mutants

Protein OspA F3 OspA Y3x3 OspA Y3x3W
PDB code 2PI3 2OY5 3AUM
Data collection statistics
Space group P21 P21 P21
Cell parameters a = 36.133 a = 37.004 a = 35.241
b = 55.617 b = 57.602 b = 53.036
c = 66.124 c = 67.746 c = 65.093
β = 98.97 β = 95.43 β = 96.97
Beamline APS 23-ID APS 17-ID PF-AR NW12
Wavelength 0.9793 1.0000 1.0000
Resolution (Å) a) 50–1.4 (1.45–1.40) 50–1.7 (1.81–1.75) 20–1.6 (1.63–1.60)
Completeness (%) 99.9 (99.7) 99.8 (98.1) 98.0 (96.4)
I/σ (I) 13.82 (3.59) 24.26 (2.37) 39.37 (8.19)
Rmergeb 0.077 (0.454) 0.067 (0.623) 0.039 (0.178)
Average redundancy 3.7 (3.5) 5.4 (5.2) 3.8 (3.8)
Refinement statistics
Resolution range (Å) 20.00–1.40 20.00–1.80 20–1.60
Reflections used (free) 48229 (2593) 25029 (1340) 29294 (1566)
R-factorc 0.16881 0.18526 0.17765
Rfreed 0.20681 0.22365 0.19860
RMS deviations
Bonds (Å) 0.009 0.017 0.010
Angles (°) 1.313 1.365 1.387
No. protein residues 246 250 247
No. waters 435 293 314
Average B factor (Å2) 14.426 27.300 12.794
Ramachandran plot statistics
Most favored (%) 90.6 91.2 91.1
Additionally allowed (%) 8.5 8.8 8.0
Generally allowed (%) 0.4 0.0 0.9
Open in a new tab
a

Highest resolution shell is shown in parenthesis.

b

R-merge = ΣhklΣi| I(hkl)i – <I(hkl)> | /ΣhklΣi<I(hkl)i> over i observations of a reflection hkl.

c

R-factor = Σ||F(obs)| – |F(calc)| |/Σ|F(obs)|.

d

Rfree is R with 5% of reflections sequestered before refinement.

The protein backbone in all mutants was well defined, except for residues 117–119 of Y3x3W, corresponding to a turn. The conformations of all the introduced aromatic amino acids were unambiguous. Y3, F3, Y3x3, and Y3x3W buried an additional 130 Å2, 152 Å2, 424 Å2, and 497 Å2, respectively, as compared with the wild-type SLB. As expected from the spatial proximity of the mutations, all of the aromatic cluster mutants showed some degrees of interaction between adjacent aromatic residues (thus contributing, at least in part, to their increased hydrophobic surface burial over wild type). However, F3 was the only mutant to display a canonical “parallel-displaced” orientation between the cluster mutation side chains (F113 and F123), a geometry that is characteristic of π-stacking between the aromatic rings [Fig. 2(A)].7 In contrast, the side chains of Y3 formed no mutual-interactions, and therefore buried additional surface area primarily through interactions with either the protein backbone or with other endogenous residues of the β-sheet scaffold. In contrast, the Y3x3 and Y3x3W mutations form dense patches of tightly clustered aromatic side chains. The Tyr residues in Y3x3 form a stunning rosette of aromatic rings, arranged around the central Tyr residue [Fig. 2(A)]. Intriguingly, the single Trp mutation in Y3x3W directed a shift in the overall packing of the 3x3 array, in which the aromatic side chains lay across the β-sheet in the same direction. The aromatic residues within each strand of Y3x3 and Y3x3W assume generally similar conformations and thereby pack against one another in a repetitive geometry across each strand.

The side-chain conformers of amino acids can have substantially different energy levels and thus contribute to protein stability.8,20,21 For example, an abundance of disfavored side-chain conformers would destabilize a protein and thereby reflect structural compromises that are required in order to maintain the overall structure.22,23 Using the backbone-dependent rotamer library of Dunbrack,24 we found that the vast majority of χ1 conformers of the mutated residues were in either the first or second most-preferred backbone-dependent conformation out of the three possible conformations [trans, gauche-, and gauche+; Fig. 2(A)]. Interestingly, gauche- and gauche+ conformers are preferred in certain positions of the aromatic clusters, despite the fact that these are considered to be less favorable conformers when viewed independent of the backbone conformation.25 Similarly, different conformers are observed at analogous positions in the aromatic cluster mutants (i.e., the trans conformer of Y113 in Y3, as compared with the gauche- conformer of F113 in F3), and are the preferred backbone-dependent rotamers in their respective contexts. These observations therefore suggest that the backbones of these structures have responded to remove unfavorable backbone/rotamer combinations.

In β-sheets containing cross-strand aromatic pairs, the gauche+ conformer is energetically disfavored and infrequently observed.25 However, when present, this conformer is often in the context of a hydrogen-bonded cross-strand pairing with a glycine, as other side chains (i.e., those with a β-carbon) clash with the aromatic ring.25 It is therefore particularly surprising that the gauche+ Tyr conformers in Y3×3 are present cross-strand from the β-branched residues L99 and V101 [Fig. 2(C)]. When compared with the wild-type residues at these positions, the gauche+ Tyr residues lean away from their cross-strand partners with concomitant local backbone distortion [Fig. 2(C)]. These results indicate that the high density of bulky aromatic side chains restrict the ability to assume particular conformers. Moreover, our observations indicate that the energy differences between favorable and unfavorable conformers of Tyr and Trp are relatively small and can be overcome by subtle adjustments to backbone conformation that compensate for steric restrictions imposed by bulky aromatic side chains [see below, Fig. 3(A)].

Figure 3.

Figure 3

Open in a new tab

Changes in the β-sheet backbone conformation by aromatic cluster mutations. A: Values of the three parameters Twist, Bend, Bend' describing the three-dimensional rotations between the two β-hairpin units (Strands 7 and 8, and Strands 9 and 10, respectively) in the SLB. Values of zero for the three parameters define a perfectly flat, rectangular β-sheet. The resulting conformational differences in the β-sheet were visualized by aligning the hairpins formed by Strands 9 and 10, with the turn regions (which are not included in the alignment) removed for clarity. B: Backbone adjustments to relieve over-packing revealed by hybrid analysis. Each strand from the aromatic cluster mutants was aligned separately with its counterpart in the wild-type structure using all backbone atoms. The adjusted coordinates of the aromatic side chains were then grafted onto the wild-type backbone to form the hybrids. The red dots indicate steric clashes identified using the Probe and King programs.

Effects of aromatic cluster mutations on the β-sheet backbone conformation

Despite the considerable number of mutations introduced into OspA to generate the Y3, F3, Y3x3, and Y3x3W proteins, they are nevertheless remarkably similar in structure to the starting wild-type scaffold (with Cα RMSDs of 0.14 Å, 1.07 Å, 1.72 Å, and 1.36 Å for the entire polypeptide, respectively). We examined structural variations among these proteins by analyzing the relative orientations of the two adjacent β-hairpin units (Strands 7–8 and 9–10) comprising the SLB into which the mutations were introduced. This analysis deconvolutes conformational changes of the β-sheet backbone into rotations about three orthogonal axes termed Twist, Bend, and Bend', respectively, as we established previously [Fig. 3(A)].13,17 Our analysis revealed that the F3, Y3x3, and Y3x3W mutations bent the β-sheet away from the face on which the aromatic side chains were introduced. F3 and Y3x3 were also less twisted than the wild-type scaffold, consistent with the prediction that aromatic residues are energetically primed for the formation of flat β-strands.26 Both of these effects were the largest in Y3x3. Unexpectedly, Y3x3W was essentially as twisted as wild type, which we attribute to the peculiarly “lopsided” direction of aromatic residue packing in this mutant. In contrast, Y3 showed marginal differences relative to wild type in all three rotational parameters, as expected from its small Cα RMSD. Thus, the β-sheets in Y3, F3, Y3x3, and Y3x3W were only nominally perturbed by the aromatic mutations. These results stand in contrast to the substantial twist facilitating the interactions between the multiple aromatic mutations used to generate the “TrpZip” motif in a small and highly distorted β-hairpin27 (see Discussion). Taken together, our observations demonstrate a high level of tolerance and inherent resiliency of the underlying β-sheet to the introduction of a dense cluster of aromatic side chains.

To elucidate the structural basis for conformational changes seen in the mutants, we examined whether the aromatic side-chain conformers were compatible with the more twisted wild-type backbone conformation [Fig. 3(B)]. The side-chain conformers found in the aromatic mutants were grafted onto the wild-type backbone to form “hybrid” models. These models showed steric clashes of in several regions within and around the aromatic clusters. The parallel-displaced π-stacked residues F113 and F123 also show a minor steric clash in the hybrid model, suggesting that their positioning is optimized within the specific context of the F3 backbone. The Y3x3 and Y3x3W hybrids exhibit the most steric clashes, especially along the β-strands and at the perimeter of the SLB. These results indicate that the bulky, interacting aromatic side chains placed in contiguous positions indeed impose some steric restrictions on the β-sheet backbone conformation. These restrictions are likely responsible for the subtle differences in backbone conformation between the mutants, as we have previously observed.13 However, it is also clear that many clashes appear to be with residues outside the 3x3 array, indicating that interactions of the cluster with its environment are also an important factor. Taken together, these conformational effects are small, given the large number of adjacent, non-conservative mutations that were introduced into the OspA scaffold.

Discussion

The design and analysis of synthetic proteins provides a critical test of our understanding of the fundamental principles of protein architecture and stability. Our results show that the aromatic cluster design strategy can be used to successfully generate stable proteins containing a large number of adjacent and tightly packed amino acid residues. This work provides an extension of surface engineering that can be used to produce subtle, defined conformational perturbations to the underlying scaffold.

The Y3x3 and Y3x3W mutants are unique among natural and synthetic proteins for their accommodation of a large number of proximal aromatic mutations into a β-sheet while remaining stable, folded, soluble, and monomeric. The surprisingly large structural and energetic differences between Y3x3 and Y3x3W demonstrate the complex nature of side-chain conformations and energetics of surface-exposed residues of a β-sheet. The greater stability of Y3x3W over Y3x3 is consistent with the notion that increased hydrophobic surface burial should produce more stable proteins. These results mimic an analysis of Trp and Tyr pairs that showed Trp-Tyr interactions are far more stabilizing than Tyr-Tyr interactions.15 Unfortunately, the dramatic reorganization of the aromatic side chains between Y3x3 and Y3x3W makes it difficult to concretely identify the basis for their distinct stability profiles. Nevertheless, this reorganization (caused by a single Tyr to Trp mutation) suggests there is a fairly small energetic barrier between different side-chain packing structures.

Unlike studies that have examined the effects of aromatic ladders in the context of the hydrophobic cores of globular proteins,28 our system based on a large, solvent exposed β-sheet helps dissect the energetic contributions of aromatic mutations by removing many higher-order tertiary packing interactions. For instance, based on a hydrophobic core model containing cross-strand aromatics,28 the F3 mutations would have been predicted to be strongly stabilizing, rather than nominally destabilizing. Likewise, the results from our present study would grossly underestimate the effect of F3 mutations placed in a hydrophobic core. Studies on the role of aromatic residues in globular proteins are likely to be heavily influenced by both their local and tertiary environment, thus underscoring the importance of examining mutations in a relevant structural context.29

The flat β-sheet of OspA readily accommodated the aromatic clusters with only modest structural rearrangements. The burial of large amounts of hydrophobic surface by assuming a large twist is therefore likely to be possible only within small β-rich peptides that have large degrees of backbone freedom.30,31 In contrast, the close proximity of aromatic mutations in the OspA mutants generally prevented edge-to-face or parallel-displaced interactions of the aromatic rings, which are strongly stabilizing in smaller and more flexible β-peptides.15 The aromatic clusters instead introduced subtle and widespread strain into the OspA scaffold, as indicated by the numerous high-energy aromatic ring conformers and in the relative decrease in scaffold stability. The disparate consequences of aromatic clusters between smaller scaffolds and OspA systems demonstrate the respective shortcomings of using β-motifs as models for larger β-sheets, and vice versa.

The ease with which the clustered aromatic mutations were engineered in OspA suggests that this simple strategy may be transferable to other β-rich systems as well. The design principles described here can be readily adapted to study patches composed of a variety of amino acids, which will further clarify how other repetitive amino acid motifs might be adapted as structural and functional tools in globular proteins. Also, because these proteins represents highly unusual amino acid compositions on a β-sheet surface, our structures should serve as interesting test cases for evaluating protein structure prediction algorithms.

Methods

Protein production

Mutagenesis, expression, and purification of the OspA mutants were performed as described previously.19 The mutations used in this work (relative to the wild-type scaffold) are as follows: F3, K113F, E123F, I136F; Y3x3, S111Y, K113Y, T115Y, S121Y, E123Y, K125Y, E134Y, I136Y, T138Y; and Y3x3W, S111Y, K113Y, T115Y, S121Y, E123W, K125Y, E134Y, I136Y, T138Y. Crystal structures were obtained using OspA mutants constructed from the “sm1” scaffold, which contains mutations in the N- and C-terminal globular domains of OspA that increase crystallization efficiency without perturbing global structure or the SLB region.5 An N-terminal His-tag was removed prior to crystallization trials by thrombin protease cleavage.

Unfolding measurements

Urea-induced unfolding of the OspA mutants was monitored using circular dichroism elipticity at 235 nm and Trp fluorescence emission at 290 nm on an AVIV 202 CD Spectrometer (Lakewood, NJ) at 30°C, as previously established.19 We did not use the fitting procedure described previously to assign numerical stabilities due to the difficulty in establishing an ambiguous model for the unfolding reaction for each mutant.32 Comparisons derived from these analyses assume that the OspA mutants share relatively similar unfolded state structures, which is unlikely in the case of the aromatic cluster mutants.

Crystallization and structure determination

Crystals were obtained using the hanging-drop vapor diffusion method at 20 °C. F3 was crystallized by mixing 1 μL of 10.0 mg/mL protein in 10 mM Tris-HCl pH 8.0 with 1 μL of a well solution containing 28% PEG400, 0.1M Tris-HCl pH 7.0. Y3x3 was crystallized by mixing 1 μL of 19.0 mg/mL protein in 10 mM Tris-HCl pH 8.0 with 1 μL of a well solution containing 18% PEG4000, 0.1M HEPES pH 7.0, 8.5% isopropanol, 15% glycerol. Y3x3W was crystallized by mixing 1 μL of 24.4 mg/mL protein in 10 mM Tris-HCl pH 8.0 with 1 μL of a well solution containing 17% PEG4000, 0.1M Tris-HCl pH 8.0, 8.5% isopropanol, 15% glycerol. X-ray diffraction data were collected using the Advanced Photon Source at the Argonne National Laboratory and the Photon Factory at the High Energy Accelerator Research Organization (KEK). Molecular replacement was performed using the wild-type OspA sm1 scaffold (PDB ID: 2G8C) as a search model. The structure was split into an N-terminal half (residues 23–131) and C-terminal half (residues 132–273) that were used simultaneously as search models. Model building and refinement were performed in the same manner as described.5,17

Data deposition

The crystal structures of F3, Y3x3, and Y3x3W were deposited in the PDB with accession codes 2PI3, 2OY5, and 3AUM, respectively.

Structure analysis

The parameters Twist, Bend, Bend' were determined as described.17 The alignment and RMSD calculations were performed using CCP4.33 Steric clashes in the hybrid models were visualized using Probe and King (http://kinemage.biochem.duke.edu). All other structure images were generated with PyMOL (http://www.pymol.org). The backbone dihedral angles (φ and ψ) and side-chain conformations (χ1) of the mutants were calculated using MolProbity (http://molprobity.biochem.duke.edu). The most recent iteration (2010) of the backbone-dependent rotamer library was obtained from the Dunbrack lab (http://dunbrack.fccc.edu).24

Acknowledgments

The authors thank V. Terechko for assistance in X-ray diffraction data collection. They also thank the staff of synchrotron beamline NW12 at the Photon Factory, KEK.

References

  1. Gilbreth RN, Koide S. Structural insights for engineering binding proteins based on non-antibody scaffolds. Curr Opin Struct Biol. 2012;22:413–420. doi: 10.1016/j.sbi.2012.06.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Lawrence MS, Phillips KJ, Liu DR. Supercharging proteins can impart unusual resilience. J Am Chem Soc. 2007;129:10110–10112. doi: 10.1021/ja071641y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Simeonov P, Berger-Hoffmann R, Hoffmann R, Strater N, Zuchner T. Surface supercharged human enteropeptidase light chain shows improved solubility and refolding yield. Protein Eng Des Sel. 2011;24:261–268. doi: 10.1093/protein/gzq104. [DOI] [PubMed] [Google Scholar]
  4. Derewenda ZS, Vekilov PG. Entropy and surface engineering in protein crystallization. Acta crystallographica Section D. Biological crystallography. 2006;62:116–124. doi: 10.1107/S0907444905035237. [DOI] [PubMed] [Google Scholar]
  5. Makabe K, Tereshko V, Gawlak G, Yan S, Koide S. Atomic-resolution crystal structure of Borrelia burgdorferi outer surface protein A via surface engineering. Protein Sci. 2006;15:1907–1914. doi: 10.1110/ps.062246706. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Burley SK, Petsko GA. Aromatic-aromatic interaction: a mechanism of protein structure stabilization. Science. 1985;229:23–28. doi: 10.1126/science.3892686. [DOI] [PubMed] [Google Scholar]
  7. McGaughey GB, Gagne M, Rappe AK. pi-Stacking interactions. Alive and well in proteins. J Biol Chem. 1998;273:15458–15463. doi: 10.1074/jbc.273.25.15458. [DOI] [PubMed] [Google Scholar]
  8. Thomas A, Meurisse R, Charloteaux B, Brasseur R. Aromatic side-chain interactions in proteins. I. Main structural features. Proteins. 2002;48:628–634. doi: 10.1002/prot.10190. [DOI] [PubMed] [Google Scholar]
  9. Minor DL., Jr Kim PS. Measurement of the beta-sheet-forming propensities of amino acids. Nature. 1994;367:660–663. doi: 10.1038/367660a0. [DOI] [PubMed] [Google Scholar]
  10. Nesloney CL, Kelly JW. Progress towards understanding beta-sheet structure. Bioorg Med Chem. 1996;4:739–766. doi: 10.1016/0968-0896(96)00051-x. [DOI] [PubMed] [Google Scholar]
  11. Santiveri CM, Rico M, Jimenez MA, Pastor MT, Perez-Paya E. Insights into the determinants of beta-sheet stability: 1H and 13C NMR conformational investigation of three-stranded antiparallel beta-sheet-forming peptides. J Pept Res. 2003;61:177–188. doi: 10.1034/j.1399-3011.2003.00045.x. [DOI] [PubMed] [Google Scholar]
  12. Smith CK, Regan L. Guidelines for protein design: the energetics of beta sheet side chain interactions. Science. 1995;270:980–982. doi: 10.1126/science.270.5238.980. [DOI] [PubMed] [Google Scholar]
  13. Biancalana M, Makabe K, Koide A, Koide S. Aromatic cross-strand ladders control the structure and stability of beta-rich peptide self-assembly mimics. J Mol Biol. 2008;383:205–213. doi: 10.1016/j.jmb.2008.08.031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Tatko CD, Waters ML. Selective aromatic interactions in beta-hairpin peptides. J Am Chem Soc. 2002;124:9372–9373. doi: 10.1021/ja0262481. [DOI] [PubMed] [Google Scholar]
  15. Wu L, McElheny D, Takekiyo T, Keiderling TA. Geometry and efficacy of cross-strand Trp/Trp, Trp/Tyr, and Tyr/Tyr aromatic interaction in a beta-hairpin peptide. Biochemistry. 2010;49:4705–4714. doi: 10.1021/bi100491s. [DOI] [PubMed] [Google Scholar]
  16. Hughes RM, Waters ML. Model systems for beta-hairpins and beta-sheets. Curr Opin Struct Biol. 2006;16:514–524. doi: 10.1016/j.sbi.2006.06.008. [DOI] [PubMed] [Google Scholar]
  17. Makabe K, McElheny D, Tereshko V, Hilyard A, Gawlak G, Yan S, Koide A, Koide S. Atomic structures of peptide self-assembly mimics. Proc Natl Acad Sci U S A. 2006;103:17753–17758. doi: 10.1073/pnas.0606690103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Makabe K, Biancalana M, Yan S, Tereshko V, Gawlak G, Miller-Auer H, Meredith SC, Koide S. High-resolution structure of a self-assembly-competent form of a hydrophobic peptide captured in a soluble beta-sheet scaffold. J Mol Biol. 2008;378:459–467. doi: 10.1016/j.jmb.2008.02.051. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Yan S, Gawlak G, Smith J, Silver L, Koide A, Koide S. Conformational heterogeneity of an equilibrium folding intermediate quantified and mapped by scanning mutagenesis. J Mol Biol. 2004;338:811–825. doi: 10.1016/j.jmb.2004.02.063. [DOI] [PubMed] [Google Scholar]
  20. Sengupta A, Mahalakshmi R, Shamala N, Balaram P. Aromatic interactions in tryptophan-containing peptides: crystal structures of model tryptophan peptides and phenylalanine analogs. J Pept Res. 2005;65:113–129. doi: 10.1111/j.1399-3011.2004.00191.x. [DOI] [PubMed] [Google Scholar]
  21. Roy RS, Gopi HN, Raghothama S, Gilardi RD, Karle IL, Balaram P. Peptide hairpins with strand segments containing alpha- and beta-amino acid residues: cross-strand aromatic interactions of facing Phe residues. Biopolymers. 2005;80:787–799. doi: 10.1002/bip.20294. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Butterfoss GL, Hermans J. Boltzmann-type distribution of side-chain conformation in proteins. Protein Sci. 2003;12:2719–2731. doi: 10.1110/ps.03273303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Tuffery P, Etchebest C, Hazout S. Prediction of protein side chain conformations: a study on the influence of backbone accuracy on conformation stability in the rotamer space. Protein engineering. 1997;10:361–372. doi: 10.1093/protein/10.4.361. [DOI] [PubMed] [Google Scholar]
  24. Shapovalov MV, Dunbrack RL., Jr A smoothed backbone-dependent rotamer library for proteins derived from adaptive kernel density estimates and regressions. Structure. 2011;19:844–858. doi: 10.1016/j.str.2011.03.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Merkel JS, Regan L. Aromatic rescue of glycine in beta sheets. Fold Des. 1998;3:449–455. doi: 10.1016/s1359-0278(98)00062-5. [DOI] [PubMed] [Google Scholar]
  26. Tsai HH, Gunasekaran K, Nussinov R. Sequence and structure analysis of parallel beta helices: implication for constructing amyloid structural models. Structure. 2006;14:1059–1072. doi: 10.1016/j.str.2006.03.015. [DOI] [PubMed] [Google Scholar]
  27. Cochran AG, Skelton NJ, Starovasnik MA. Tryptophan zippers: stable, monomeric beta -hairpins. Proc Natl Acad Sci U S A. 2001;98:5578–5583. doi: 10.1073/pnas.091100898. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Budyak IL, Zhuravleva A, Gierasch LM. The role of aromatic-aromatic interactions in strand-strand stabilization of beta-sheets. J Mol Biol. 2013;425:3522–3535. doi: 10.1016/j.jmb.2013.06.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Minor DL, Jr, Kim PS. Context is a major determinant of beta-sheet propensity. Nature. 1994;371:264–267. doi: 10.1038/371264a0. [DOI] [PubMed] [Google Scholar]
  30. Santiveri CM, Santoro J, Rico M, Jimenez MA. Factors involved in the stability of isolated beta-sheets: turn sequence, beta-sheet twisting, and hydrophobic surface burial. Protein Sci. 2004;13:1134–1147. doi: 10.1110/ps.03520704. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Aravinda S, Shamala N, Rajkishore R, Gopi HN, Balaram P. A crystalline beta-hairpin peptide nucleated by a type I' Aib-D-Ala beta-turn: evidence for cross-strand aromatic interactions. Angew Chem Int Ed Engl. 2002;41:3863–3865. doi: 10.1002/1521-3773(20021018)41:20<3863::AID-ANIE3863>3.0.CO;2-A. [DOI] [PubMed] [Google Scholar]
  32. Yan S, Gawlak G, Makabe K, Tereshko V, Koide A, Koide S. Hydrophobic surface burial is the major stability determinant of a flat, single-layer beta-sheet. J Mol Biol. 2007;368:230–243. doi: 10.1016/j.jmb.2007.02.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Collaborative-Computational-Project. The CCP4 suite: programs for protein crystallography. Acta Crystallography D. 1994;50:760–763. doi: 10.1107/S0907444994003112. [DOI] [PubMed] [Google Scholar]

Articles from Protein Science : A Publication of the Protein Society are provided here courtesy of The Protein Society

RESOURCES