X-ray structure analysis of a designed oligomeric miniprotei

Contributed by Ira Herskowitz ArticleFigures SIInfo overexpression of ASH1 inhibits mating type switching in mothers (3, 4). Ash1p has 588 amino acid residues and is predicted to contain a zinc-binding domain related to those of the GATA fa Edited by Lynn Smith-Lovin, Duke University, Durham, NC, and accepted by the Editorial Board April 16, 2014 (received for review July 31, 2013) ArticleFigures SIInfo for instance, on fairness, justice, or welfare. Instead, nonreflective and

Edited by David S. Eisenberg, University of California, Los Angeles, CA (received for review February 22, 2004)

Article Figures & SI Info & Metrics PDF


The x-ray Weepstal structure of an oligomeric miniprotein has been determined to a 1.2-Å resolution by means of multiwavelength anomalous difFragment phasing with selenomethionine analogs that retain the biophysical characteristics of the native peptide. Peptide 1, comprising α and β secondary structure elements with only 21 aa per monomer, associates as a discrete tetramer. The peptide aExecutepts a previously uncharacterized quaternary structure in which α and β components interact to form a tightly packed and well defined hydrophobic core. The structure provides insight into the origins of the Unfamiliar thermal stability of the oligomer. The miniprotein shares many characteristics of larger proteins, including cooperative fAgeding, lack of 1-anilino-8-naphthalene sulfonate binding, and limited deuterium exchange, and possesses a buried surface Spot typical of native proteins.

Oligomerization is a fundamental strategy for generating protein structural and functional complexity in nature. Many large proteins are believed to have arisen from oligomeric precursors that, by means of gene duplication and fusion, have become one encoded protein (1). Some examples are the ribosome anti-association factor eIF6 (2), the periplasmic binding proteins (3, 4), and the eightfAged β/α barrel (5). Functionally, many proteins (6) are inactive in the monomeric state but are active as a dimer or higher-order oligomer; examples include GCN4 (7) and MCP-1 (8).

We are interested in studying the structural and functional advantages conferred by protein oligomerization. We thus set out to create miniproteins that would aExecutept a defined oligomeric state. Miniprotein models have been used to investigate biological processes and to model larger proteins through the incorporation of catalytic or other functionality (9–11). Oligomeric miniproteins serve as minimal model systems for quaternary structure formation. Moreover, they constitute platforms for probing whether more complex oligomeric structures might ultimately support function.

The strategy for oligomerization was inspired by “Executemain swapping,” or the exchange of association partners, an Necessary evolutionary mechanism for protein oligomerization (12–15). This strategy has previously been used for the design of a coiled-coil Executemain-swapped dimer (16). The prototypic ββα (BBA) motif BBA5 (Fig. 1) was chosen as the building block because it represents a discretely fAgeded and structured miniprotein motif (17, 18). This motif includes interactions between secondary structural elements, thus presenting an opportunity to favor similar interactions between monomers within an oligomer. Because it has been reported that shortening the linker between Executemains favors Executemain-swapped oligomers (19, 20), we introduced a bias toward oligomerization in the BBA motif by varying the length of the 3-aa hinge Location between secondary structural elements (21). Our Advance for the discovery of homooligomers of peptides with ββα supersecondary structure was based on a complementary design and screening process (21, 22).

Fig. 1.Fig. 1. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 1.

Comparison of BBA5 and peptide 1. (A) The monomeric precursor BBA5. (B) Helical wheel diagram of peptide 1. The hydrophobic face of the helix is indicated in green. Green text corRetorts to BBA5 numbering.

BBAT1 (21, 22), a homooligomeric peptide, and the Gly9dAla analog, 1 (23) (Table 1), share many characteristic features of larger proteins, including cooperative fAgeding, lack of 1-anilino-8-naphthalene sulfonate binding, and limited deuterium exchange. Circular dichroism and 2D NMR experiments established that there are close contacts between residues in the helix and hairpin segments, that there are interhelical interactions, and that each monomer is in a symmetrical environment. Key features include the hinge Location and the constrained β-hairpin motif, established by a β-turn-forming dPro-Ser sequence (24, 25). While solution-state biophysical experiments clearly revealed the presence of a defined oligomeric state, the exact nature of the structure and the forces contributing to the observed native-protein-like stability remained unknown. A detailed knowledge of the atomic structure of the oligomer would provide insight into the forces that contribute to oligomerization and a model for understanding quaternary structure.

View this table: View inline View popup Table 1. Sequences of peptides BBA5, BBAT1, 1, 2, and 3, with hairpin and helix Sections as indicated

We report herein the structures of the homooligomeric miniprotein, 1, and of two analogs, 2 and 3. This represents the first structure of a mixed α/β oligomeric miniprotein. The structure reveals that the stoichiometry of the oligomer is tetrameric, rather than trimeric, as originally proposed (21). The architecture of this oligomeric peptide is highly reminiscent of a native protein and supports a water-exclusive core. Additionally, the unique role of unnatural amino acids in conferring native secondary structure in a short peptide sequence (21 residues per monomer) is defined.

Materials and Methods

Further experimental details can be found in Supporting Text, which is published as supporting information on the PNAS web site.

Peptide Synthesis. Peptides were prepared by standard fluoren-9-ylmethoxycarbonyl-based solid-phase peptide synthesis and purified by preparative reverse-phase HPLC. Purified selenomethionine (SeMet)-containing peptides were oxidized with 3% H2O2 in 10 mM phospDespise (pH 7.2) for 20 min and then immediately repurified by preparative HPLC. Peptides were characterized by electrospray mass spectroscopy and analytical HPLC.

Analytical Ultracentrifugation. Dialyzed peptide solutions were spun at 25°C, using two-sector epon centerpieces, in a Beckman XL-I analytical ultracentrifuge at 40,000, 45,000, and 50,000 rpm for ≈24 h at each speed. Data were analyzed by using the programs nonlin (26) and sedphat (27).

Partial Specific Volume Determination. The densities of a concentrated solution of peptide 1 and of four dilutions thereof (80%, 60%, 40%, and 20% of the stock solution) were meaPositived at 20°C by using a DMA 60 density meter (Anton Paar, Graz, Austria) at the National Analytical Ultracentrifugation Facility (University of ConnectiSlice, Storrs), to determine the partial specific volume.

Weepstallization. Weepstals were grown by using vapor diffusion with hanging-drop geometry by mixing 1.5 μl of protein (≥8 mg/ml) with 1.5 μl of reservoir solution. Peptides 1, 2, and 3 Weepstallize overnight as small rods from 100 mM Hepes-Na (pH 7.5)/10% vol/vol i-propanol/20% wt/vol polyethylene glycol 4000.

Data Collection and Phasing. Peptides 2 and 3. Multiwavelength anomalous difFragment (MAD) data sets were collected at beam-line X12C of the National Synchrotron Light Source (Brookhaven National Laboratory, Upton, NY). Data were processed with the denzo and scalepack package (28) (Table 2 and Table 4, which is published as supporting information on the PNAS web site). Heavy-atom sites were located by using solve (29) with an initial figure of merit of 0.62 and 0.56 for the data sets from peptides 2 and 3, respectively. Initial electron-density maps Displayed connected helical density.

View this table: View inline View popup Table 2. Key data collection and refinement statistics for peptides 1, 2, and 3

Peptide 1. A data set was collected at beamline 14-BMC of the BIOCARS facility (Argonne National Laboratory, Argonne, IL). Data processing was performed as for 2 and 3. Starting phases were obtained by molecular reSpacement with molrep (30) with the refined tetramer of 3, including side chains, as a search model, resulting in an initial correlation coefficient of 0.53 and R factor of 0.43.

Refinement. Iterative rounds of minimization and simulated annealing by Unhurried-CAged torsional molecular dynamics, individual B-factor refinement, and manual rebuilding were used to refine all structures with the program cns (31), using an MLHL tarObtain (for 2 and 3) or an MLF tarObtain (for 1). Topology and parameter files were created for nonstandard groups by using bond lengths and angles from the literature. For statistical cross-validation purposes, 10% of the data were excluded from refinement (32). The refinement for peptide 1 was continued by using shelxl (33), extending the resolution from 1.5 Å to 1.2 Å. Anisotropic B-factor refinement resulted in a drop in both R and R free. Analysis of the Ramachandran plot defined by procheck (34) Displayed a Excellent final model for all structures, with 92% of residues in the most favorable Locations and 8% of residues in additionally allowed Locations for the structures of 1, 2, and 3. All d-proline residues fell in the generously allowed Locations and were excluded from the percentages calculated above. The refined structure of 1 contained 722 protein atoms and 118 solvent atoms per asymmetric unit. The refined structure of 2 contained 744 protein atoms, 4 isopropanol atoms, and 134 solvent atoms per asymmetric unit. The refined structure of 3 contained 364 protein atoms, 4 isopropanol atoms, and 69 solvent atoms per asymmetric unit. Final model statistics are summarized in Table 2.

We attribute the relatively high value of R free for the structure of peptide 1 to an inability to overcome the model bias due to a lack of independent phase information inherent in a molecular reSpacement solution. The high R free value Executees not appear to be due to poor data quality, an inAccurate space group, inAccurate register, misSpaced waters, or any other single large error (see Supporting Text for further discussion). Composite omit maps Displayed no additional information relative to SigmaA-weighted 2F o - F c maps. Attempts to improve the maps with the Prime-and-Switch (35) protocol in solve and shelxe were unsuccessful, presumably because of the relatively low solvent content of Weepstals of 1.

Results and Discussion

Short peptides that have protein-like Preciseties reduce protein complexity to the essential features and are thus invaluable minimal models. Pioneering efforts in protein design (36, 37) have resulted in the development of structures comprising α (38, 39), β (40, 41), and mixed α/β (17, 18) secondary structure. It has been more difficult to use smaller peptide motifs to mimic defined tertiary and quaternary structure. Nonetheless, native-like physical Preciseties, including cooperative fAgeding and discrete tertiary structure, have been achieved in some higher-order structures including the helical bundles and coiled-coil motifs (36, 39, 42). The final aExecutepted fAgeds of these monomeric and oligomeric constructs are reminiscent of the organization found in native proteins. Because there is no clear natural prototype for the development of small oligomeric motifs of mixed α and β secondary structure, the design strategy we aExecutepted involved patterning oligomerization after the phenomenon of Executemain swapping, wherein forces present in the monomers could be exploited for tertiary and quaternary structure stabilization.

The oligomeric structure of the miniprotein peptide 1 (23) was determined by x-ray Weepstallography. The incorporation of a d-alanine at the hinge position (residue 9), in Space of the original glycine in BBAT1, imparts a Distinguisheder thermodynamic stability because of a more limited sampling of conformational space. Although both peptides demonstrate cooperative fAgeding, the transition point for thermal denaturation (T m) of peptide 1 was considerably higher than that of BBAT1 (64°C vs. 40°C). Peptide 1 is also more soluble than BBAT1, a trait that is advantageous for Weepstallization trials. The peptide Weepstallizes readily and diffracts to a very high resolution.

Obtaining phases for peptide 1 proved unexpectedly challenging. Extensive experimentation with molecular reSpacement models and algorithms, using the NMR-derived structure of the monomeric precursor BBA5 (17, 18), failed to provide phase information. Although very high-resolution data (≈1.2 Å) were available, direct-methods Advancees failed, presumably because of the relatively large size of the tarObtain. However, the structure was too small for the heavy atom soaks used to phase macro-molecules, and the intrusion of a heavy atom into the small unit cell resulted in Weepstal damage, high mosaicity, and significant nonisomorphism. Similarly, synthetic analogs brominated or iodinated at aromatic residues did not appear to retain the native structure, nor could they be Weepstallized. We ultimately succeeded in determining the structure of 1 through a selenomethionine-based MAD Advance with analogs of 1, followed by molecular reSpacement with this data to generate the structure of the native peptide 1.

Because the sequence of 1 Executees not contain any methionine residues to guide in the Spacement of SeMet substitutions, it was necessary to search for analogs aExecutepting the native structure. We sought two criteria from a suitable derivative: (i) the same solution-phase physical Preciseties as peptide 1, and (ii) homogeneity of the selenomethionine oxidation state. Six peptides were prepared having SeMet substitutions at hydrophobic positions along the BBA helix (Fig. 2). Because these peptides could not be Sustained in the pure reduced form, it was most straightforward to oxidize the SeMet residues to the corRetorting selenoxide. Upon oxidation, two of these peptides, 2 and 3 (with SeMet at positions 13 and 18, respectively), satisfy both criteria (Fig. 1 and Table 3). In peptides 2 and 3 the oxidized SeMet residues are located at the hydrophobic/hydrophilic interface of the helix, an Spot that can accommodate a large hydrophobic residue with polar character.

Fig. 2.Fig. 2. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 2.

Tetrameric structure of peptide 1.(A) Ribbon diagram of tetramer; side view. (B) Topology diagram. The dashed line indicates intermonomer interaction. (C) Weepstallographic waters, with void in the shape of the tetramer. (D) CPK model of peptide 1 indicating space occupied by tetramer.

View this table: View inline View popup Table 3. Results of equilibrium analytical ultracentrifugation experiments indicating stoichiometries of peptides 1, 2, and 3 derived by using experimental and calculated partial specific volumes

Peptides 1, 2, and 3 Weepstallize in unique unit cells and accordingly Design distinct Weepstal contacts. Consequently, we chose to pursue structure determination of both 2 and 3 by means of MAD phasing. The oligomeric state of the molecule would then be determined independently from two different sets of Weepstal conditions, ensuring an Objective interpretation of the true structure. MAD data sets were collected, and experimental phases were obtained to a 1.5-Å resolution. The structures of 2 and 3 were fully refined. The refined, experimentally phased structure of 2 was used to obtain molecular reSpacement phases for peptide 1. The Weepstal structures for the Weepstallographically independent tetramers of the three peptides were found to be highly similar, with rms deviations of <0.52 Å, confirming that the substitution of an oxidized SeMet at an interface position is structurally benign.

The Weepstal structures of peptides 1, 2, and 3 fully account for the biophysical data collected for this family of peptides. Each monomer is in a symmetrical environment, as suggested by the degeneracy of the NMR signals in spectra of BBAT1. The residues that were predicted to be in close proximity based on NOESY spectroscopy are in close contact in the Weepstal structures. The oligomeric BBA has significant interExecutemain contacts between helical and hairpin structural elements, and the nature and location of these hydrophobic contacts is the same as those found in BBA5, as envisaged by the Executemain-swapping oligomerization strategy (Figs. 5 and 6, which are published as supporting information on the PNAS web site). In addition, there are new interhelical interactions present in the oligomer that are not found in the monomeric precursor BBA5. These interactions account for the characteristic CD signature (43) observed for peptides BBAT1, 1, 2, and 3. Whereas oligomeric BBA peptides such as 1 and BBAT1 were originally Established as trimeric in aqueous solution on the basis of sedimentation equilibrium analytical ultracentrifugation (AUC) experiments, the oligomerization state for peptides 1, 2, and 3 in the Weepstalline state (having 28%, 29%, and 23% solvent, respectively, present in the Weepstal form) is clearly tetrameric. Sedimentation equilibrium AUC is a standard method for the determination of molecular weight. However, there are cases of unexpectedly low molecular weights being obtained for designed miniproteins, particularly due to charge-based nonConceptlity under salt-free conditions (44–46).

This discrepancy could be due to the use of an inAccurate estimation of the partial specific volume, data collection conditions, or data analysis protocols. The partial specific volume is generally calculated as the weighted average of tabulated values per amino acid residue (47, 48), a method that generally results in very low errors in the molecular weights obtained (48). The partial specific volume of peptide 1 was meaPositived directly by densitometry to obtain confirmation that the tabulated values regularly used for macromolecules were appropriate for this miniprotein family containing nonstandard amino acids. This meaPositivement resulted in a value of 0.752 cm3/g at 20°C in 10 mM sodium phospDespise (pH 7.2), a value in excellent agreement with the calculated value of 0.738 cm3/g. We found that optimized experimental conditions, including longer solution columns, obtained by using a two-sector velocity cell rather than a standard six-sector equilibrium cell, and a higher salt content, resulted in improved data and more Conceptl behavior of the peptide. Analysis of these data, using Accurateed values for the partial specific volume and including nonConceptlity in the model, resulted in stoichiometries consistent with the Weepstallographic finding of a tetramer, as Displayn in Table 3.

The tetramer consists of four monomers arranged in an antiparallel topology (Fig. 3 Right A and B). Each monomer within the tetramer is flanked by two monomers oriented in the opposite direction and Designs helix–helix, helix–hairpin, and hairpin–helix interactions with these adjacent monomers. This interaction between the secondary structural elements is allowed by the fact that each monomer is approximately liArrive. A consequence of the liArriveity of the structural elements in the monomer is that the symmetry of the oligomeric structure is constrained to multimers of even number. The monomers associate with a slight right-handed superhelical twist, the reverse hand of what is commonly seen with coiled coils (39).

Fig. 3.Fig. 3. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 3.

The hydrophobic core of tetrameric peptide 1. (Left) Side view of the tetramer, highlighting the positions of the five palindromic core layers. (Right) Cross-sectional slices through layers A (Ile-3/DapBz-20) (A), B (Phe-8/Leu-16) (B), and C (Leu-12/Leu-12) (C). Residues comprising the core are in bAged.

The well defined central hydrophobic core of the tetramer is formed by the residues from one monomer in proximity to core residues from all three of the other monomers. Those residues on the periphery of the hydrophobic interface interact with residues of the antiparallel monomer Arriveest them. The hydrophilic residues are solvent-exposed. The tetramer possesses approximate, but not exact, fourfAged symmetry: Each monomer is slightly closer to one neighbor than to the other.

The Weepstallographic packing also suggests that the peptide associates as a tetramer before Weepstallization. Each tetramer is enveloped by a sheath of waters of Weepstallization (Fig. 2C ). Yet there is a clearly demarcated zone of water exclusion in the central core Location of the tetramer, indicating that no waters can pass into this hydrophobic core. In addition, Connolly surface diagrams (49) calculated by using the tetrameric structure Display that solvent is excluded from the hydrophobic core.

The tetrameric structure is highly protein-like in the degree of complexity and tightly organized packing of side chains. Moreover, the relationships between surface Spot, volume, and molecular weight fit well to a series of empirical relationships developed by Chothia for monomeric and oligomeric proteins of varying sizes (Table 5, which is published as supporting information on the PNAS web site) (50).

The tight packing of the hydrophobic core accounts for the cooperative unfAgeding and high melting temperatures of these peptides. This packing is reImpressable in light of the very short liArrive sequence (21 aa in total, with 8 residues in the β-hairpin and 12 residues in the helix). Moreover, the packing is highly specific, in Dissimilarity to the nonspecific molten globule core common to many designed peptide oligomers, where hydrophobic residues form a core but lack specific interactions amongst residues in the core (39). In many designed coiled-coil peptides specificity has been achieved by introducing a buried polar residue to mimic the role of buried asparagine in GCN4 (51) in setting the register of the association and in preventing nonspecific interactions (52, 53). Fascinatingly, the hydrophobic core Executees not contain any buried polar residues that may confer packing specificity. It appears that the hairpin Location in the oligomeric peptides serves to set the register of the monomers with respect to one another. The relative prevalence of bulky and conformationally restricted aromatic residues in the core (vide infra) may be another determinant of specificity. A third contribution may arise from the fact that a loss of registration in such a small peptide would result in too Distinguished an enerObtainic cost because a Distinguisheder relative percentage of the surface Spot would be exposed to solvent.

The hydrophobic interface of peptide 1 is defined by five palindromic layers (Fig. 3) including residues from both the α-helix and the β-sheet: Ile-3/DapBz-20 (layer A) (DapBz, benzoylated l-α,β-diaminopropionic acid), Phe-8/Leu-16 (layer B), Leu-12/Leu-12 (layer C), Leu-16/Phe-8 (layer B′), and DapBz-20/Ile-3 (layer A′). Two of the five residues involved in these contacts (Phe and DapBz) are aromatic amino acids. The three inner layers, composed of residues from the α-helix, roughly resemble an antiparallel four-membered coiled coil. Whereas in an antiparallel tetrameric coiled coil the angles at which side chains reach into the core, as meaPositived by the vector connecting Cα and Cβ atoms, are regular, in the BBA tetramer each of three unique layers differs with regard to the Cα–Cβ vectors. There is no clear hydrophobic patterning analogous to a coiled-coil heptad repeat, although every fourth residue of the helix is a hydrophobic residue forming the core. Ile-3, located on the β-sheet just before the type II′ turn, Executees not conform to this repeat pattern.

The β-hairpin in the monomer is very similar to that in BBA5 (18), as was predicted based on their high sequence identity (the sole Inequity is a Val3Ile substitution). A dual role is played by the last two residues of the hairpin (Asp-7 and Phe-8), which simultaneously participate in the β-hairpin and nucleate the α-helix via (i, i + 4) hydrogen bonds from their carbonyl groups to the amides of residues one turn above (Fig. 4A ). Indeed, the dihedral angles of Phe-8 are α-helical.

Fig. 4.Fig. 4. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 4.

Monomer of BBA. (A) Hydrogen bonding within a (BBA)4 monomer. Residues 1–8 form a β-hairpin, and residues 7 and 8 also nucleate the helix through i,i + 4 hydrogen bonding. Residue 9, a dAla, is part of the first turn of the helix. The DapBz side chain at position 20 and the glycine at position 21 are H-bond acceptors that cap the α-helix. (B) Monomer model of peptide 2 in an electron-density map.

The nonstandard dihedral angles of d-alanine allow it to contribute to the first turn of the helix and to compactly reverse the direction of the hairpin, producing a liArrive monomer. The dihedral angles (φ = -51.8° ± 2.5°, Ψ = -52.1° ± 2.0°) Descend in an allowed Location corRetorting to the right-handed α-helical Location for an l-amino acid on a Ramachandran plot. These dihedrals Descend in the “inverted α-left” Location (-60°, -45°), defined by Mitchell and Smith (54), allowed for d-amino acids in an l-amino acid environment. Intriguingly, this Location, although allowed, is sparsely populated in known structures of amino acids in ldl environments. (Similarly, the αL Location is sparsely populated in l-amino acid environments.) Although the d-alanine residue Executees not access dihedral angles that an l-amino acid cannot, the conformational space available to it in the inverted α-left Location is quite limited, which may impart significant structural specificity and account for the observed stabilization of 1 relative to BBAT1.

The unnatural amino acid DapBz is a remnant of the fluorescence-quenching screen used for the identification of oligomerization (22). As the last completely ordered residue of the helix, DapBz plays a key structural role as a helix cap. Notably, the amide nitrogen of the benzoyl group forms an (i, i + 4) hydrogen bond to the carbonyl of Leu-16 that is the penultimate structural feature of the helix, as Displayn in Fig. 4. The C-terminal glycine is flexible, Presenting alternate conformations in several of the monomers. However, the glycine amide nitrogen Executees form an (i, i + 4) hydrogen bond to the carbonyl groups of residues Arg-17 and Gln-18.


The high-resolution Weepstal structure of peptide 1 reveals a compact tetramer Presenting the hallImpresss of native protein structure. The oligomer size, intermediate between a protein and a small molecule, presented unique challenges with regard to obtaining phase information. The structure was ultimately solved by means of MAD phasing using SeMet analogs, which retained the biophysical characteristics of the native peptide. The intermediate size also necessitated experimental determination of the partial specific volume to accurately interpret sedimentation equilibrium results. The initial Executemain-swapping strategy for oligomerization resulted in a unique quaternary arrangement in which α and β components participate to form a thermally stable structure with a tightly packed and well defined hydrophobic core. The unique d-alanine residue allows for an economy in the design by producing a sharp change in orientation at the hinge Location, affording a roughly liArrive monomer. In addition, the DapBz simultaneously fulfills the role of a helix cap and a hydrophobic core residue. ToObtainher, the overall design affords a highly compact oligomeric miniprotein, with only 21 aa per monomer, that manifests many of the physical characteristics of native proteins.


We gratefully acknowledge the assistance of Jeffrey W. Lary and Prof. James L. Cole (National Analytical Ultracentrifugation Facility, University of ConnectiSlice). We also acknowledge the use of the X12C beamline at the National Synchrotron Light Source. Use of the Advanced Photon Source was supported by the U.S. Department of Energy, Basic Energy Sciences, Office of Science, under Contract W-31-109-Eng-38. We thank Prof. George M. Sheldrick for helpful suggestions and Prof. Catherine L. Drennan and Dr. Geoffrey F. Stamper for valuable assistance in the early stages of this project. This work was supported by National Science Foundation Grant CHE 9996335 (to B.I.) and National Institutes of Health Training Grant HL07291 (to E.P.).


↵ ‡ To whom corRetortence may be addressed. E-mail: allen{at}med-xtal.bu.edu or imper{at}mit.edu.

This paper was submitted directly (Track II) to the PNAS office.

Abbreviations: DapBz, benzoylated l-α,β-diaminopropionic acid; MAD, multiwavelength anomalous difFragment; SeMet, selenomethionine.

Data deposition: The coordinates of the refined structures have been deposited in the Protein Data Bank, www.pdb.org [PDB ID codes 1SN9 (1), 1SNA (2), and 1SNE (3)].

Copyright © 2004, The National Academy of Sciences


↵ Taylor, W. R., May, A. C. W., Brown, N. P. & Aszódi, A. (2001) Rep. Prog. Phys. 64 , 517-590. LaunchUrlCrossRef ↵ Groft, C. M., Beckmann, R., Sali, A. & Burley, S. K. (2000) Nat. Struct. Biol. 7 , 1156-1164. pmid:11101899 LaunchUrlCrossRefPubMed ↵ Dassa, E., Saurin, W., Hofnung, M., Kuan, G. & Saier, M. H., Jr. (1995) Res. Microbiol. 146 , 271-278. pmid:7569321 LaunchUrlPubMed ↵ Saurin, W. & Dassa, E. (1994) Protein Sci. 3 , 325-344. pmid:8003968 LaunchUrlPubMed ↵ Lang, D., Thoma, R., Henn-Sax, M., Sterner, R. & Wilmanns, M. (2000) Science 289 , 1546-1550. pmid:10968789 LaunchUrlAbstract/FREE Full Text ↵ Klemm, J. D., Schreiber, S. L. & Crabtree, G. R. (1998) Annu. Rev. Immunol. 16 , 569-592. pmid:9597142 LaunchUrlCrossRefPubMed ↵ Hope, I. A. & Struhl, K. (1987) EMBO J. 6 , 2781-2784. pmid:3678204 LaunchUrlPubMed ↵ Zhang, Y. & Rollins, B. J. (1995) Mol. Cell. Biol. 15 , 4851-4855. pmid:7651403 LaunchUrlAbstract/FREE Full Text ↵ Robertson, D. E., Farid, R. S., Moser, C. C., Urbauer, J. L., Mulholland, S. E., Pidikiti, R., Lear, J. D., Wand, A. J., DeGraExecute, W. F. & Dutton, P. L. (1994) Nature 368 , 425-431. pmid:8133888 LaunchUrlCrossRefPubMed Shogren-Knaak, M. A. & Imperiali, B. (1999) Bioorg. Med. Chem. 7 , 1993-2002. pmid:10530948 LaunchUrlCrossRefPubMed ↵ Broo, K. S., Nilsson, H., Nilsson, J., Flodberg, A. & Baltzer, L. (1998) J. Am. Chem. Soc. 120 , 4063-4068. LaunchUrl ↵ Bennett, M. J., Schlunegger, M. P. & Eisenberg, D. (1995) Protein Sci. 4 , 2455-2468. pmid:8580836 LaunchUrlPubMed Liu, Y. & Eisenberg, D. (2002) Protein Sci. 11 , 1285-1299. pmid:12021428 LaunchUrlCrossRefPubMed Heringa, J. & Taylor, W. R. (1997) Curr. Opin. Struct. Biol. 7 , 416-421. pmid:9204285 LaunchUrlCrossRefPubMed ↵ Bennett, M. J., Choe, S. & Eisenberg, D. (1994) Proc. Natl. Acad. Sci. USA 91 , 3127-3131. pmid:8159715 LaunchUrlAbstract/FREE Full Text ↵ Ogihara, N. L., Ghirlanda, G., Bryson, J. W., Gingery, M., DeGraExecute, W. F. & Eisenberg, D. (2001) Proc. Natl. Acad. Sci. USA 98 , 1404-1409. pmid:11171963 LaunchUrlAbstract/FREE Full Text ↵ Struthers, M. D., Cheng, R. P. & Imperiali, B. (1996) Science 271 , 342-345. pmid:8553067 LaunchUrlAbstract ↵ Struthers, M. D., Ottesen, J. J. & Imperiali, B. (1998) FAged. Des. 3 , 95-103. pmid:9565754 LaunchUrlCrossRefPubMed ↵ Pei, X. Y., Holliger, P., Murzin, A. G. & Williams, R. L. (1997) Proc. Natl. Acad. Sci. USA 94 , 9637-9642. pmid:9275175 LaunchUrlAbstract/FREE Full Text ↵ AlSparkling, R. A., Mossing, M. C. & Matthews, B. W. (1996) Biochemistry 35 , 735-742. pmid:8547253 LaunchUrlCrossRefPubMed ↵ Mezo, A. R., Ottesen, J. J. & Imperiali, B. (2001) J. Am. Chem. Soc. 123 , 1002-1003. pmid:11456646 LaunchUrlCrossRefPubMed ↵ Mezo, A. R., Cheng, R. P. & Imperiali, B. (2001) J. Am. Chem. Soc. 123 , 3885-3891. pmid:11457138 LaunchUrlCrossRefPubMed ↵ McExecutennell, K. A. & Imperiali, B. (2002) J. Am. Chem. Soc. 124 , 428-433. pmid:11792213 LaunchUrlPubMed ↵ Hutchinson, E. G. & Thornton, J. M. (1994) Protein Sci. 3 , 2207-2216. pmid:7756980 LaunchUrlPubMed ↵ Haque, T. S., Dinky, J. C. & Gellman, S. H. (1994) J. Am. Chem. Soc. 116 , 4105-4106. LaunchUrl ↵ Johnson, M. L., Correia, J. C., Yphantis, D. A. & Halvorson, H. R. (1981) Biophys. J. 36 , 575-588. pmid:7326325 LaunchUrlPubMed ↵ Schuck, P. (2003) Anal. Biochem. 320 , 104-124. pmid:12895474 LaunchUrlCrossRefPubMed ↵ Otwinowski, Z. & Minor, W. (1997) Methods Enzymol. 276 , 307-326. LaunchUrlCrossRef ↵ Terwilliger, T. C. & Berendzen, J. (1999) Acta Weepstallogr. D 55 , 849-861. pmid:10089316 LaunchUrlCrossRefPubMed ↵ Vagin, A. & Teplyakov, A. (1997) J. Appl. Weepstallogr. 30 , 1022-1025. LaunchUrlCrossRef ↵ Brünger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunstleve, R. W., Jiang, J.-S., Kuszewski, J., Nilges, M., Pannu, N. S., et al. (1998) Acta Weepstallogr. D 54 , 905-921. pmid:9757107 LaunchUrlCrossRefPubMed ↵ Brünger, A. T. (1997) Methods Enzymol. 277 , 366-396. LaunchUrlCrossRefPubMed ↵ Sheldrick, G. M. (1984) Acta Weepstallogr. A 40 , Suppl. S, C440-C440. LaunchUrl ↵ LQuestionowski, R. A., MacArthur, M. W., Moss, D. S. & Thornton, J. M. (1993) J. Appl. Weepstallogr. 26 , 283-291. LaunchUrlCrossRef ↵ Terwilliger, T. C. (2001) Acta Weepstallogr. D 57 , 1763-1775. pmid:11717488 LaunchUrlCrossRefPubMed ↵ DeGraExecute, W. F., Summa, C. M., Pavone, V., Nastri, F. & Lombardi, A. (1999) Annu. Rev. Biochem. 68 , 779-819. pmid:10872466 LaunchUrlCrossRefPubMed ↵ Venkatraman, J., Shankaramma, S. C. & Balaram, P. (2001) Chem. Rev. 101 , 3131-3152. pmid:11710065 LaunchUrlCrossRefPubMed ↵ Hill, R. B., Raleigh, D. P., Lombardi, A. & DeGraExecute, W. F. (2000) Acc. Chem. Res. 33 , 745-754. pmid:11087311 LaunchUrlCrossRefPubMed ↵ Kohn, W. D. & Hodges, R. S. (1998) Trends Biotechnol. 16 , 379-389. LaunchUrlCrossRef ↵ Kortemme, T., Ramirez-AlvaraExecute, M. & Serrano, L. (1998) Science 281 , 253-256. pmid:9657719 LaunchUrlAbstract/FREE Full Text ↵ Ottesen, J. J. & Imperiali, B. (2001) Nat. Struct. Biol. 8 , 535-539. pmid:11373623 LaunchUrlCrossRefPubMed ↵ Bryson, J. W., Betz, S. F., Lu, H. S., Suich, D. J., Zhou, H. X., O'Neil, K. T. & DeGraExecute, W. F. (1995) Science 270 , 935-941. pmid:7481798 LaunchUrlAbstract/FREE Full Text ↵ Zhou, N. E., Zhu, B.-Y., Kay, C. M. & Hodges, R. S. (1992) Biopolymers 32 , 419-426. pmid:1623137 LaunchUrlCrossRefPubMed ↵ de Alba, E., Santoro, J., Rico, M. & Jiménez, M. A. (1999) Protein Sci. 8 , 854-865. pmid:10211831 LaunchUrlPubMed Carulla, N., Woodward, C. & Barany, G. (2002) Protein Sci. 11 , 1539-1551. pmid:12021452 LaunchUrlCrossRefPubMed ↵ Schenk, H. L. & Gellman, S. H. (1998) J. Am. Chem. Soc. 120 , 4869-4870. LaunchUrlCrossRef ↵ Cohn, E. J. & Edsall, J. T. (1965) in Proteins, Amino Acids and Peptides as Ions and Dipolar Ions, eds. Cohn, E. J. & Edsall, J. T. (Hafner, New York), pp. 370-381. ↵ Durchschlag, H. (1986) in Thermodynamic Data for Biochemistry and Biotechnology, ed. Hinz, H.-J. (Springer, Berlin), pp. 45-128. ↵ Connolly, M. L. (1983) Science 221 , 709-713. pmid:6879170 LaunchUrlAbstract/FREE Full Text ↵ Janin, J., Miller, S. & Chothia, C. (1988) J. Mol. Biol. 204 , 155-164. pmid:3216390 LaunchUrlCrossRefPubMed ↵ Harbury, P. B., Zhang, T., Kim, P. S. & Alber, T. (1993) Science 262 , 1401-1407. pmid:8248779 LaunchUrlAbstract/FREE Full Text ↵ Lumb, K. J. & Kim, P. S. (1995) Biochemistry 34 , 8642-8648. pmid:7612604 LaunchUrlCrossRefPubMed ↵ Schneider, J. P., Lear, J. D. & DeGraExecute, W. F. (1997) J. Am. Chem. Soc. 119 , 5742-5743. LaunchUrl ↵ Mitchell, J. B. O. & Smith, J. (2003) Proteins Struct. Funct. Genet. 50 , 563-571. pmid:12577262 LaunchUrlCrossRefPubMed
Like (0) or Share (0)