Christine E. Tinberg

Computational design of ligand-binding proteins with high affinity and selectivity.

The ability to design proteins with high affinity and selectivity for any given small molecule is a rigorous test of our understanding of the physiochemical principles that govern molecular recognition. Attempts to rationally design ligand-binding proteins have met with little success, however, and the computational design of protein–small-molecule interfaces remains an unsolved problem. Current approaches for designing ligand-binding proteins for medical and biotechnological uses rely on raising antibodies against a target antigen in immunized animals and/or performing laboratory-directed evolution of proteins with an existing low affinity for the desired ligand, neither of which allows complete control over the interactions involved in binding. Here we describe a general computational method for designing pre-organized and shape complementary small-molecule-binding sites, and use it to generate protein binders to the steroid digoxigenin (DIG). Of seventeen experimentally characterized designs, two bind DIG; the model of the higher affinity binder has the most energetically favourable and pre-organized interface in the design set. A comprehensive binding-fitness landscape of this design, generated by library selections and deep sequencing, was used to optimize its binding affinity to a picomolar level, and X-ray co-crystal structures of two variants show atomic-level agreement with the corresponding computational models. The optimized binder is selective for DIG over the related steroids digitoxigenin, progesterone and β-oestradiol, and this steroid binding preference can be reprogrammed by manipulation of explicitly designed hydrogen-bonding interactions. The computational design method presented here should enable the development of a new generation of biosensors, therapeutics and diagnostics.

Bioluminescent sensor proteins for point-of-care therapeutic drug monitoring

For many drugs, finding the balance between efficacy and toxicity requires monitoring their concentrations in the patients blood. Quantifying drug levels at the bedside or at home would have advantages in terms of therapeutic outcome and convenience, but current techniques require the setting of a diagnostic laboratory. We have developed semisynthetic bioluminescent sensors that permit precise measurements of drug concentrations in patient samples by spotting minimal volumes on paper and recording the signal using a simple point-and-shoot camera. Our sensors have a modular design consisting of a protein-based and a synthetic part and can be engineered to selectively recognize a wide range of drugs, including immunosuppressants, antiepileptics, anticancer agents and antiarrhythmics. This low-cost point-of-care method could make therapies safer, increase the convenience of doctors and patients and make therapeutic drug monitoring available in regions with poor infrastructure.

A general strategy to construct small molecule biosensors in eukaryotes.

Biosensors for small molecules can be used in applications that range from metabolic engineering to orthogonal control of transcription. Here, we produce biosensors based on a ligand-binding domain (LBD) by using a method that, in principle, can be applied to any target molecule. The LBD is fused to either a fluorescent protein or a transcriptional activator and is destabilized by mutation such that the fusion accumulates only in cells containing the target ligand. We illustrate the power of this method by developing biosensors for digoxin and progesterone. Addition of ligand to yeast, mammalian, or plant cells expressing a biosensor activates transcription with a dynamic range of up to ~100-fold. We use the biosensors to improve the biotransformation of pregnenolone to progesterone in yeast and to regulate CRISPR activity in mammalian cells. This work provides a general methodology to develop biosensors for a broad range of molecules in eukaryotes. DOI: http://dx.doi.org/10.7554/eLife.10606.001

Improving the Catalytic Performance of an Artificial Metalloenzyme by Computational Design.

Artifical metalloenzymes combine the reactivity of small molecule catalysts with the selectivity of enzymes, and new methods are required to tune the catalytic properties of these systems for an application of interest. Structure-based computational design could help to identify amino acid mutations leading to improved catalytic activity and enantioselectivity. Here we describe the application of Rosetta Design for the genetic optimization of an artificial transfer hydrogenase (ATHase hereafter), [(η(5)-Cp*)Ir(pico)Cl] ⊂ WT hCA II (Cp* = Me5C5(-)), for the asymmetric reduction of a cyclic imine, the precursor of salsolsidine. Based on a crystal structure of the ATHase, computational design afforded four hCAII variants with protein backbone-stabilizing and hydrophobic cofactor-embedding mutations. In dansylamide-competition assays, these designs showed 46-64-fold improved affinity for the iridium pianostool complex [(η(5)-Cp*)Ir(pico)Cl]. Gratifyingly, the new designs yielded a significant improvement in both activity and enantioselectivity (from 70% ee (WT hCA II) to up to 92% ee and a 4-fold increase in total turnover number) for the production of (S)-salsolidine. Introducing additional hydrophobicity in the Cp*-moiety of the Ir-catalyst provided by adding a propyl substituent on the Cp* moiety yields the most (S)-selective (96% ee) ATHase reported to date. X-ray structural data indicate that the high enantioselectivity results from embedding the piano stool moiety within the protein, consistent with the computational model.

CSAR Benchmark Exercise 2013: Evaluation of Results from a Combined Computational Protein Design, Docking, and Scoring/Ranking Challenge

Community Structure-Activity Resource (CSAR) conducted a benchmark exercise to evaluate the current computational methods for protein design, ligand docking, and scoring/ranking. The exercise consisted of three phases. The first phase required the participants to identify and rank order which designed sequences were able to bind the small molecule digoxigenin. The second phase challenged the community to select a near-native pose of digoxigenin from a set of decoy poses for two of the designed proteins. The third phase investigated the ability of current methods to rank/score the binding affinity of 10 related steroids to one of the designed proteins (pKd = 4.1 to 6.7). We found that 11 of 13 groups were able to correctly select the sequence that bound digoxigenin, with most groups providing the correct three-dimensional structure for the backbone of the protein as well as all atoms of the active-site residues. Eleven of the 14 groups were able to select the appropriate pose from a set of plausible decoy poses. The ability to predict absolute binding affinities is still a difficult task, as 8 of 14 groups were able to correlate scores to affinity (Pearson-r > 0.7) of the designed protein for congeneric steroids and only 5 of 14 groups were able to correlate the ranks of the 10 related ligands (Spearman-ρ > 0.7).

Engineering of Kuma030: A Gliadin Peptidase That Rapidly Degrades Immunogenic Gliadin Peptides in Gastric Conditions

Celiac disease is characterized by intestinal inflammation triggered by gliadin, a component of dietary gluten. Oral administration of proteases that can rapidly degrade gliadin in the gastric compartment has been proposed as a treatment for celiac disease; however, no protease has been shown to specifically reduce the immunogenic gliadin content, in gastric conditions, to below the threshold shown to be toxic for celiac patients. Here, we used the Rosetta Molecular Modeling Suite to redesign the active site of the acid-active gliadin endopeptidase KumaMax. The resulting protease, Kuma030, specifically recognizes tripeptide sequences that are found throughout the immunogenic regions of gliadin, as well as in homologous proteins in barley and rye. Indeed, treatment of gliadin with Kuma030 eliminates the ability of gliadin to stimulate a T cell response. Kuma030 is capable of degrading >99% of the immunogenic gliadin fraction in laboratory-simulated gastric digestions within physiologically relevant time frames, to a level below the toxic threshold for celiac patients, suggesting great potential for this enzyme as an oral therapeutic for celiac disease.

Improved Free-Energy Landscape Quantification Illustrated with a Computationally Designed Protein-Ligand Interaction

Quantifying the energy landscape underlying protein-ligand interactions leads to an enhanced understanding of molecular recognition. A powerful yet accessible single-molecule technique is atomic force microscopy (AFM)-based force spectroscopy, which generally yields the zero-force dissociation rate constant (koff ) and the distance to the transition state (Δx≠ ). Here, we introduce an enhanced AFM assay and apply it to probe the computationally designed protein DIG10.3 binding to its target ligand, digoxigenin. Enhanced data quality enabled an analysis that yielded the height of the transition state (ΔG≠ =6.3±0.2 kcal mol-1 ) and the shape of the energy barrier at the transition state (linear-cubic) in addition to the traditional parameters [koff (=4±0.1×10-4  s-1 ) and Δx≠ (=8.3±0.1 Å)]. We expect this automated and relatively rapid assay to provide a more complete energy landscape description of protein-ligand interactions and, more broadly, the diverse systems studied by AFM-based force spectroscopy.

Computational Design of Ligand Binding Proteins

The ability to design novel small-molecule binding sites in proteins is a stringent test of our understanding of the principles of molecular recognition, and would have many practical applications, in synthetic biology and medicine. Here, we describe a computational method in the context of the macromolecular modeling suite Rosetta to designing proteins with sites featuring predetermined interactions to ligands of choice. The required inputs for the method are a model of the small molecule and the desired interactions (e.g., hydrogen bonding, electrostatics, steric packing), and a set of crystallographic structures of proteins containing existing or predicted binding pockets. Constellations of backbones surrounding the putative pocket are searched for compatibility with the desired binding site conception using RosettaMatch and surrounding amino acid side chain identities are optimized using RosettaDesign. Validation of the design is performed using metrics that evaluate the interface energy of the predicted binding pose, the preformation of key binding site features in the apo-state, and the local compatibility of the designed sequence changes with the wild type backbone structure, and top-ranking candidate designs are generated for experimental validation. This approach can allow for the creation of novel binding sites and for the rational tuning of specificity for congeneric ligands by altering the programmed interactions by design, thus offering a general computational protocol for construction and modulation of protein-small molecule interfaces.

Improving Binding Affinity and Selectivity of Computationally Designed Ligand-Binding Proteins Using Experiments

The ability to de novo design proteins that can bind small molecules has wide implications for synthetic biology and medicine. Combining computational protein design with the high-throughput screening of mutagenic libraries of computationally designed proteins is emerging as a general approach for creating binding proteins with programmable binding modes, affinities, and selectivities. The computational step enables the creation of a binding site in a protein that otherwise does not (measurably) bind the intended ligand, and targeted mutagenic screening allows for validation and refinement of the computational model as well as provides orders-of-magnitude increases in the binding affinity. Deep sequencing of mutagenic libraries can provide insights into the mutagenic binding landscape and enable further affinity improvements. Moreover, in such a combined computational-experimental approach where the binding mode is preprogrammed and iteratively refined, selectivity can be achieved (and modulated) by the placement of specified amino acid side chain groups around the ligand in defined orientations. Here, we describe the experimental aspects of a combined computational-experimental approach for designing-using the software suite Rosetta-proteins that bind a small molecule of choice and engineering, using fluorescence-activated cell sorting and high-throughput yeast surface display, high affinity and ligand selectivity. We illustrated the utility of this approach by performing the design of a selective digoxigenin (DIG)-binding protein that, after affinity maturation, binds DIG with picomolar affinity and high selectivity over structurally related steroids.