Faculty Bibliography

Most natural proteins performing sophisticated tasks contain multiple domains where an active site is located at the domain interface. Comparative structural analyses suggest that major leaps in protein function occur through gene recombination events that connect two or more protein domains to generate a new active site, frequently occurring at the newly created domain interface. However, such functional leaps by combination of unrelated domains have not been directly demonstrated. Here we show that highly specific and complex protein functions can be generated by joining a low-affinity peptide-binding domain with a functionally inert second domain and subsequently optimizing the domain interface. These directed evolution processes dramatically enhanced both affinity and specificity to a level unattainable with a single domain, corresponding to >500-fold and >2,000-fold increases of affinity and specificity, respectively. An x-ray crystal structure revealed that the resulting "affinity clamp" had clamshell architecture as designed, with large additional binding surface contributed by the second domain. The affinity clamps having a single-nanomolar dissociation constant outperformed a monoclonal antibody in immunochemical applications. This work establishes evolutionary paths from isolated domains with primitive function to multidomain proteins with sophisticated function and introduces a new protein-engineering concept that allows for the generation of highly functional affinity reagents to a predefined target. The prevalence and variety of natural interaction domains suggest that numerous new functions can be designed by using directed domain interface evolution.

beta-Rich self-assembly is a major structural class of polypeptides, but still little is known about its atomic structures and biophysical properties. Major impediments for structural and biophysical studies of peptide self-assemblies include their insolubility and heterogeneous composition. We have developed a model system, termed peptide self-assembly mimic (PSAM), based on the single-layer beta-sheet of Borrelia outer surface protein A. PSAM allows for the capture of a defined number of self-assembly-like peptide repeats within a water-soluble protein, making structural and energetic studies possible. In this work, we extend our PSAM approach to a highly hydrophobic peptide sequence. We show that a penta-Ile peptide (Ile(5)), which is insoluble and forms beta-rich self-assemblies in aqueous solution, can be captured within the PSAM scaffold in a form capable of self-assembly. The 1.1-A crystal structure revealed that the Ile(5) stretch forms a highly regular beta-strand within this flat beta-sheet. Self-assembly models built with multiple copies of the crystal structure of the Ile(5) peptide segment showed no steric conflict, indicating that this conformation represents an assembly-competent form. The PSAM retained high conformational stability, suggesting that the flat beta-strand of the Ile(5) stretch primed for self-assembly is a low-energy conformation of the Ile(5) stretch and rationalizing its high propensity for self-assembly. The ability of the PSAM to "solubilize" an otherwise insoluble peptide stretch suggests the potential of the PSAM approach to the characterization of self-assembling peptides.

Antibodies that bind protein antigens are indispensable in biochemical research and modern medicine. However, knowledge of RNA-binding antibodies and their application in the ever-growing RNA field is lacking. Here we have developed a robust approach using a synthetic phage-display library to select specific antigen-binding fragments (Fabs) targeting a large functional RNA. We have solved the crystal structure of the first Fab-RNA complex at 1.95 A. Capability in phasing and crystal contact formation suggests that the Fab provides a potentially valuable crystal chaperone for RNA. The crystal structure reveals that the Fab achieves specific RNA binding on a shallow surface with complementarity-determining region (CDR) sequence diversity, length variability, and main-chain conformational plasticity. The Fab-RNA interface also differs significantly from Fab-protein interfaces in amino acid composition and light-chain participation. These findings yield valuable insights for engineering of Fabs as RNA-binding modules and facilitate further development of Fabs as possible therapeutic drugs and biochemical tools to explore RNA biology.

We present detailed computational studies based on electrostatic calculations to evaluate the origins of pKa values and the pH dependence of stability for the 10th type III domain of human fibronectin (FNfn10). One of our goals is to validate the calculation protocols by comparison to experimental data (Koide, A.; Jordan, M. R.; Horner, S.; Batori, V.; Koide, S. Biochemistry 2001, 40, 10326-10333). Another goal is to evaluate the sensitivity of the calculated ionization free energies and apparent pKa values on local structural fluctuations, which do not alter the structural convergence to a particular architecture, by using a complete ensemble of solution NMR structures and the NMR average minimized structure of FNfn10 (Main, A. L.; Harvey, T. S.; Baron, M.; Boyd, J.; Campbell, I. D. Cell 1992, 71, 671-678). Our calculations demonstrate that, at high ionic strength, FNfn10 is more stable at low pH compared to neutral pH, in overall agreement with experimental data. This behavior is attributed to contributions from unfavorable Coulombic interactions in a surface patch for the pairs Asp7-Glu9 and Asp7-Asp23. The unfavorable interactions are decreased at low pH, where the acidic residues become neutral, and are further decreased at high ionic strength because of increased screening by salt ions. Elimination of the unfavorable interactions in the theoretical mutants Asp7Asn (D7N) and Asp7Lys (D7K) produce higher calculated stabilities at neutral pH and any ionic strength compared to the wild-type, in agreement with the experimental data. We also discuss subtleties in the calculated apparent pKa values and ionization free energies, which are not in agreement with the experimental data. This work demonstrates that comparative electrostatic calculations can provide rapid predictions of pH-dependent properties of proteins and can be significant aids in guiding the design of proteins with tailored properties.

We investigated how the register between adjacent beta-strands is specified using a series of mutants of the single-layer beta-sheet (SLB) in Borrelia OspA. The single-layer architecture of this system eliminates structural restraints imposed by a hydrophobic core, enabling us to address this question. A critical turn (turn 9/10) in the SLB was replaced with a segment with an intentional structural mismatch. Its crystal structure revealed a one-residue insertion into the central beta-strand (strand 9) of the SLB. This insertion triggered a surprisingly large-scale structural rearrangement: (i) the central strand (strand 9) was shifted by one residue, causing the strand to flip with respect to the adjacent beta-strands and thus completely disrupting the native side-chain contacts; (ii) the three-residue turn located on the opposite end of the beta-strand (turn 8/9) was pushed into its preceding beta-strand (strand 8); (iii) the register between strands 8 and 9 was shifted by three residues. Replacing the original sequence for turn 8/9 with a stronger turn motif restored the original strand register but still with a flipped beta-strand 9. The stability differences of these distinct structures were surprisingly small, consistent with an energy landscape where multiple low-energy states with different beta-sheet configurations exist. The observed conformations can be rationalized in terms of maximizing the number of backbone H-bonds. These results suggest that adjacent beta-strands "stick" through the use of factors that are not highly sequence specific and that beta-strands could slide back and forth relatively easily in the absence of external elements such as turns and tertiary packing.

A major architectural class in engineered binding proteins ("antibody mimics") involves the presentation of recognition loops off a single-domain scaffold. This class of binding proteins, both natural and synthetic, has a strong tendency to bind a preformed cleft using a convex binding interface (paratope). To explore their capacity to produce high-affinity interfaces with diverse shape and topography, we examined the interface energetics and explored the affinity limit achievable with a flat paratope. We chose a minimalist paratope limited to two loops found in a natural camelid heavy-chain antibody (VHH) that binds to ribonuclease A. Ala scanning of the VHH revealed only three "hot spot" side chains and additional four residues important for supporting backbone-mediated interactions. The small number of critical residues suggested that this is not an optimized paratope. Using selection from synthetic combinatorial libraries, we enhanced its affinity by >100-fold, resulting in variants with Kd as low as 180 pM with no detectable loss of binding specificity. High-resolution crystal structures revealed that the mutations induced only subtle structural changes but extended the network of interactions. This resulted in an expanded hot spot region including four additional residues located at the periphery of the paratope with a concomitant loss of the so-called "O-ring" arrangement of energetically inert residues. These results suggest that this class of simple, single-domain scaffolds is capable of generating high-performance binding interfaces with diverse shape. More generally, they suggest that highly functional interfaces can be designed without closely mimicking natural interfaces.

We have previously established a minimalist approach to antibody engineering by using a phage-displayed framework to support complementarity determining region (CDR) diversity restricted to a binary code of tyrosine and serine. Here, we systematically augmented the original binary library with additional levels of diversity and examined the effects. The diversity of the simplest library, in which only heavy chain CDR positions were randomized by the binary code, was expanded in a stepwise manner by adding diversity to the light chain, by diversifying non-paratope residues that may influence CDR conformations, and by adding additional chemical diversity to CDR-H3. The additional diversity incrementally improved the affinities of antibodies raised against human vascular endoethelial growth factor and the structure of an antibody-antigen complex showed that tyrosine side-chains are sufficient to mediate most of the interactions with antigen, but a glycine residue in CDR-H3 was critical for providing a conformation suitable for high-affinity binding. Using new high-throughput procedures and the most complex library, we produced multiple high-affinity antibodies with dissociation constants in the single-digit nanomolar range against a wide variety of protein antigens. Thus, this fully synthetic, minimalist library has essentially recapitulated the capacity of the natural immune system to generate high-affinity antibodies. Libraries of this type should be highly useful for proteomic applications, as they minimize inherent complexities of natural antibodies that have hindered the establishment of high-throughput procedures. Furthermore, analysis of a large number of antibodies derived from these well-defined and simplistic libraries allowed us to uncover statistically significant trends in CDR sequences, which provide valuable insights into antibody library design and into factors governing protein-protein interactions.

Phage display is the longest-standing platform among molecular display technologies. Recent developments have extended its utility to proteins that were previously recalcitrant to phage display. The technique has played a dominant role in forming the field of synthetic binding protein engineering, where novel interfaces have been generated from libraries built using antibody fragment frameworks and also alternative scaffolds. Combinatorial methods have also been developed for the rapid analysis of binding energetics across protein interfaces. The ability to rapidly select and analyze binding interfaces, and compatibility with high-throughput methods under diverse conditions, makes it likely that the combination of phage display and synthetic combinatorial libraries will prove to be the method of choice for synthetic binding protein engineering for broad applications.

Formation of a flat beta-sheet is a fundamental event in beta-sheet-mediated protein self-assembly. To investigate the contributions of various factors to the stability of flat beta-sheets, we performed extensive alanine-scanning mutagenesis experiments on the single-layer beta-sheet segment of Borrelia outer surface protein A (OspA). This beta-sheet segment consists of beta-strands with highly regular geometries that can serve as a building block for self-assembly. Our Ala-scanning approach is distinct from the conventional host-guest method, in that it introduces only conservative, truncation mutations that should minimize structural perturbation. Our results showed very weak correlation with experimental beta-sheet propensity scales, statistical beta-sheet propensity scales, or cross-strand pairwise correlations. In contrast, our data showed strong positive correlation with the change in buried non-polar surface area. Polar interactions including prominent Glu-Lys cross-strand pairs contribute marginally to the beta-sheet stability. These results were corroborated by results from additional non-Ala mutations. Taken together, these results demonstrate the dominant contribution of non-polar surface burial to flat beta-sheet stability even at solvent-exposed positions. The OspA single-layer beta-sheet achieves efficient hydrophobic surface burial without forming a hydrophobic core by a strategic placement of a variety of side-chains. These findings further suggest the importance of hydrophobic interactions within a beta-sheet layer in peptide self-assembly.

High degrees of sequence and conformation complexity found in natural protein interaction interfaces are generally considered essential for achieving tight and specific interactions. However, it has been demonstrated that specific antibodies can be built by using an interface with a binary code consisting of only Tyr and Ser. This surprising result might be attributed to yet undefined properties of the antibody scaffold that uniquely enhance its capacity for target binding. In this work we tested the generality of the binary-code interface by engineering binding proteins based on a single-domain scaffold. We show that Tyr/Ser binary-code interfaces consisting of only 15-20 positions within a fibronectin type III domain (FN3; 95 residues) are capable of producing specific binding proteins (termed "monobodies") with a low-nanomolar K(d). A 2.35-A x-ray crystal structure of a monobody in complex with its target, maltose-binding protein, and mutation analysis revealed dominant contributions of Tyr residues to binding as well as striking molecular mimicry of a maltose-binding protein substrate, beta-cyclodextrin, by the Tyr/Ser binary interface. This work suggests that an interaction interface with low chemical diversity but with significant conformational diversity is generally sufficient for tight and specific molecular recognition, providing fundamental insights into factors governing protein-protein interactions.