Guiding discovery of protein sequence-structure-function modeling.

MOTIVATION: Protein engineering techniques are key in designing novel catalysts for a wide range of reactions. Although approaches vary in their exploration of the sequence-structure-function paradigm, they are often hampered by the labor-intensive steps of protein expression and screening. In this work, we describe the development and testing of a high-throughput in silico sequence-structure-function pipeline using AlphaFold2 and fast Fourier transform docking that is benchmarked with enantioselectivity and reactivity predictions for an ancestral sequence library of fungal flavin-dependent monooxygenases. RESULTS: The predicted enantioselectivities and reactivities correlate well with previously described screens of an experimentally available subset of these proteins and capture known changes in enantioselectivity across the phylogenetic tree representing ancestorial proteins from this family. With this pipeline established as our functional screen, we apply ensemble decision tree models and explainable AI techniques to build sequence-function models and extract critical residues within the binding site and the second-sphere residues around this site. We demonstrate that the top-identified key residues in the control of enantioselectivity and reactivity correspond to experimentally verified residues. The in silico sequence-to-function pipeline serves as an accelerated framework to inform protein engineering efforts from vast informative sequence landscapes contained in protein families, ancestral resurrects, and directed evolution campaigns. AVAILABILITY: Jupyter notebooks detailing the sequence-structure-function pipeline are available at https://github.com/BrooksResearchGroup-UM/seq_struct_func.

Enhanced Sampling of Buried Charges in Free Energy Calculations Using Replica Exchange with Charge Tempering.

Buried ionizable groups in proteins often play important structural and functional roles. However, it is generally challenging to study the detailed molecular mechanisms solely based on experimental measurements. Free energy calculations using atomistic simulations, on the other hand, complement experimental studies and can provide high temporal and spatial resolution information that can lead to mechanistic insights. Nevertheless, it is also well recognized that sufficient sampling of such atomistic simulations can be challenging, considering that structural changes related to the buried charges may be very slow. In the present study, we describe a simple but effective enhanced sampling technique called replica exchange with charge tempering (REChgT) with a novel free energy method, multisite λ dynamics (MSλD), to study two systems containing buried charges, pKa prediction of a small molecule, orotate, in complex with the dihydroorotate dehydrogenase, and relative stability of a Glu-Lys pair buried in the hydrophobic core of two variants of Staphylococcal nuclease. Compared to the original MSλD simulations, the usage of REChgT dramatically increases sampling in both conformational and alchemical spaces, which directly translates into a significant reduction of wall time to converge the free energy calculations. This study highlights the importance of sufficient sampling toward developing improved free energy methods.

Covalent docking in CDOCKER.

Targeted covalent inhibitors (TCIs) are considered to be an important component in the toolbox of drug discovery and about 30% of currently marketed drugs are TCIs. Although these drugs raise concerns about toxicity, their high potencies and prolonged effects result in less-frequent drug dosing and wide therapeutic margins for patients. This leads to increased interests in developing new computational methods to identify novel covalent inhibitors. The implementation of successful in silico docking algorithms have the potential to provide significant savings of time and money in the discovery of lead compounds. In this paper, we describe the implementation and testing of a covalent docking methodology in Rigid CDOCKER and the optimization of the corresponding physics-based scoring function with an additional customizable covalent bond grid potential which represents the free energy change of bond formation between the ligand and the receptor. We optimize the covalent bond grid potential for different common covalent bond formation reaction in TCIs. The average runtime for docking one covalent compound is 15 minutes which is comparable or faster than other well-established covalent docking methods. We demonstrate comparable top rank accuracy compared with other covalent docking algorithms using the pose prediction benchmark dataset for covalent docking algorithms developed by the Keseru group. Finally, we construct a retrospective virtual screening benchmark dataset containing 8 different receptor targets with different covalent bond formation reactions. To our knowledge, this is the largest dataset for benchmarking covalent docking methods. We show that our new covalent docking algorithm has the ability to identify lead compounds among a large chemical space. The largest AUC value is 0.909 for the target receptor CATK and the warhead chemistry of the covalent inhibitors is addition to the aldehyde functionality.

Electrostatic Forces Control the Negative Allosteric Regulation in a Disordered Protein Switch.

The transcriptional adaptor zinc-binding 1 (TAZ1) domain of the transcriptional coactivator CBP/P300 and two disordered peptides, HIF-1α and CITED2, form a delicate protein switch that regulates cellular hypoxic response. In hypoxia, HIF-1α binds TAZ1 to control the transcription of adaptive genes critical for the recovery from hypoxic stress. CITED2 acts as the negative feedback regulator to rapidly displace HIF-1α and efficiently attenuate the hypoxic response. Though CITED2 and HIF-1α have the same dissociation constant (Kd = 10 nM) in their binary complexes with TAZ1, CITED2 is much more competitive than HIF-1α upon binding the same target TAZ1 in ternary ( Berlow et al. Nature 2017 , 543 , 447 - 451 ). Here we demonstrate that a simple coarse-grained model can recapitulate this negative allosteric effect and provide detailed physical insights into the displacement mechanism. We find that long-range electrostatic forces are essential for the efficient displacement of HIF-1α by CITED2. The strong electrostatic interactions between CITED2 and TAZ1, along with the unique binding mode, make CITED2 much more competitive than HIF-1α in binding TAZ1.

Deciphering protein evolution and fitness landscapes with latent space models.

Protein sequences contain rich information about protein evolution, fitness landscapes, and stability. Here we investigate how latent space models trained using variational auto-encoders can infer these properties from sequences. Using both simulated and real sequences, we show that the low dimensional latent space representation of sequences, calculated using the encoder model, captures both evolutionary and ancestral relationships between sequences. Together with experimental fitness data and Gaussian process regression, the latent space representation also enables learning the protein fitness landscape in a continuous low dimensional space. Moreover, the model is also useful in predicting protein mutational stability landscapes and quantifying the importance of stability in shaping protein evolution. Overall, we illustrate that the latent space models learned using variational auto-encoders provide a mechanism for exploration of the rich data contained in protein sequences regarding evolution, fitness and stability and hence are well-suited to help guide protein engineering efforts.

Molecular Mechanisms of Interactions between Monolayered Transition Metal Dichalcogenides and Biological Molecules.

Single layered two-dimensional (2D) materials such as transition metal dichalcogenides (TMDs) show great potential in many microelectronic or nanoelectronic applications. For example, because of extremely high sensitivity, TMD-based biosensors have become promising candidates for next-generation label-free detection. However, very few studies have been conducted on understanding the fundamental interactions between TMDs and other molecules including biological molecules, making the rational design of TMD-based sensors (including biosensors) difficult. This study focuses on the investigations of the fundamental interactions between proteins and two widely researched single-layered TMDs, MoS2, and WS2 using a combined study with linear vibrational spectroscopy attenuated total reflectance FTIR and nonlinear vibrational spectroscopy sum frequency generation vibrational spectroscopy, supplemented by molecular dynamics simulations. It was concluded that a large surface hydrophobic region in a relatively flat location on the protein surface is required for the protein to adsorb onto a monolayered MoS2 or WS2 surface with preferred orientation. No disulfide bond formation between cysteine groups on the protein and MoS2 or WS2 was found. The conclusions are general and can be used as guiding principles to engineer proteins to attach to TMDs. The approach adopted here is also applicable to study interactions between other 2D materials and biomolecules.

Molecular Interactions between Graphene and Biological Molecules.

Applications of graphene have extended into areas of nanobio-technology such as nanobio-medicine, nanobio-sensing, as well as nanoelectronics with biomolecules. These applications involve interactions between proteins, peptides, DNA, RNA etc. and graphene, therefore understanding such molecular interactions is essential. For example, many applications based on using graphene and peptides require peptides to interact with (e.g., noncovalently bind to) graphene at one end, while simultaneously exposing the other end to the surrounding medium (e.g., to detect analytes in solution). To control and characterize peptide behavior on a graphene surface in solution is difficult. Here we successfully probed the molecular interactions between two peptides (cecropin P1 and MSI-78(C1)) and graphene in situ and in real-time using sum frequency generation (SFG) vibrational spectroscopy and molecular dynamics (MD) simulation. We demonstrated that the distribution of various planar (including aromatic (Phe, Trp, Tyr, and His)/amide (Asn and Gln)/Guanidine (Arg)) side-chains and charged hydrophilic (such as Lys) side-chains in a peptide sequence determines the orientation of the peptide adsorbed on a graphene surface. It was found that peptide interactions with graphene depend on the competition between both planar and hydrophilic residues in the peptide. Our results indicated that part of cecropin P1 stands up on graphene due to an unbalanced distribution of planar and hydrophilic residues, whereas MSI-78(C1) lies down on graphene due to an even distribution of Phe residues and hydrophilic residues. With such knowledge, we could rationally design peptides with desired residues to manipulate peptide-graphene interactions, which allows peptides to adopt optimized structure and exhibit excellent activity for nanobio-technological applications. This research again demonstrates the power to combine SFG vibrational spectroscopy and MD simulation in studying interfacial biological molecules.

Functionally important conformations of the Met20 loop in dihydrofolate reductase are populated by rapid thermal fluctuations.

Conformational changes in enzymes are well recognized to play an important role in the organization of the reactive groups for efficient catalysis. This study reveals atomic and energetic details of the conformational change process that precedes the catalytic reaction of the enzyme dihydrofolate reductase. The computed free energy profile provides insights into the ligand binding mechanism and a quantitative estimate of barrier heights separating different conformational states along the pathway. Studies show that the ternary complex comprised of NADPH cofactor and substrate dihydrofolate undergoes transitions between a closed state and an occluded state via an intermediate "open" conformation. During these transitions the largest conformational change occurs in the Met20 loop of DHFR and is accompanied by the motion of the cofactor into and out of the binding pocket. When the cofactor is out of the binding pocket, the enzyme frequently samples open and occluded conformations with a small (approximately 5 k(B)T) free energy barrier between the two states. However, when the cofactor is in the binding pocket, the closed conformation is thermodynamically most favored. The determination of a profile characterizing the position-dependent diffusion of the Met20 loop allowed us to apply reaction rate theory and deduce the kinetics of loop motions based on the computed free energy landscape.

Exploring assembly energetics of the 30S ribosomal subunit using an implicit solvent approach.

To explore the relationship between the assembly of the 30S ribosomal subunit and interactions among the constituent components, 16S RNA and proteins, relative binding free energies of the T. thermophilus 30S proteins to the 16S RNA were studied based on an implicit solvent model of electrostatic, nonpolar, and entropic contributions. The late binding proteins in our assembly map were found not to bind to the naked 16S RNA. The 5' domain early kinetic class proteins, on average, carry the highest positive charge, get buried the most upon binding to 16S RNA, and show the most favorable binding. Some proteins (S10/S14, S6/S18, S13/S19) have more stabilizing interactions while binding as dimers. Our computed assembly map resembles that of E. coli; however, the central domain path is more similar to that of A. aeolicus, a hyperthermophilic bacteria.