A systematic family-wide investigation reveals that ~30% of mammalian PDZ domains engage in PDZ-PDZ interactions

PDZ domains are independently folded modules that typically mediate protein-protein interactions by binding to the C termini of their target proteins. However, in a few instances, PDZ domains have been reported to dimerize with other PDZ domains. To investigate this noncanonical-binding mode further, we used protein microarrays comprising virtually every mouse PDZ domain to systematically query all possible PDZ-PDZ pairs. We then used fluorescence polarization to retest and quantify interactions and coaffinity purification to test biophysically validated interactions in the context of their full-length proteins. Overall, we discovered 37 PDZ-PDZ interactions involving 46 PDZ domains (~30% of all PDZ domains tested), revealing that dimerization is a more frequently used binding mode than was previously appreciated. This suggests that many PDZ domains evolved to form multiprotein complexes by simultaneously interacting with more than one ligand.     

Domain Analysis and Motif Matcher (DAMM): A Program to Predict Selectivity Determinants in Monosiga brevicollis PDZ Domains Using Human PDZ Data

Choanoflagellates are single-celled eukaryotes with complex signaling pathways. They are considered the closest non-metazoan ancestors to mammals and other metazoans and form multicellular-like states called rosettes. The choanoflagellate Monosiga brevicollis contains over 150 PDZ domains, an important peptide-binding domain in all three domains of life (Archaea, Bacteria, and Eukarya). Therefore, an understanding of PDZ domain signaling pathways in choanoflagellates may provide insight into the origins of multicellularity. PDZ domains recognize the C-terminus of target proteins and regulate signaling and trafficking pathways, as well as cellular adhesion. Here, we developed a computational software suite, Domain Analysis and Motif Matcher (DAMM), that analyzes peptide-binding cleft sequence identity as compared with human PDZ domains and that can be used in combination with literature searches of known human PDZ-interacting sequences to predict target specificity in choanoflagellate PDZ domains. We used this program, protein biochemistry, fluorescence polarization, and structural analyses to characterize the specificity of A9UPE9_MONBE, a M. brevicollis PDZ domain-containing protein with no homology to any metazoan protein, finding that its PDZ domain is most similar to those of the DLG family. We then identified two endogenous sequences that bind A9UPE9 PDZ with <100 μM affinity, a value commonly considered the threshold for cellular PDZ–peptide interactions. Taken together, this approach can be used to predict cellular targets of previously uncharacterized PDZ domains in choanoflagellates and other organisms. Our data contribute to investigations into choanoflagellate signaling and how it informs metazoan evolution.     

A fluorescence polarization based screening assay for identification of small molecule inhibitors of the PICK1 PDZ domain

PDZ (PSD-95/Discs-large/ZO-1 homology) domains represent putative targets in several diseases including cancer, stroke, addiction and neuropathic pain. Here we describe the application of a simple and fast screening assay based on fluorescence polarization (FP) to identify inhibitors of the PDZ domain in PICK1 (protein interacting with C kinase 1). We screened 43,380 compounds for their ability to inhibit binding of an Oregon Green labeled C-terminal dopamine transporter peptide (OrG-DAT C13) to purified PICK1 in solution. The assay was highly reliable with excellent screening assay parameters (Z'≈0.7 and Z≈0.6). Out of ~200 compounds that reduced FP to less than 80% of the control wells, six compounds were further characterized. The apparent affinities of the compounds were determined in FP competition binding experiments and ranged from ~5.0 μM to ~193 μM. Binding to the PICK1 PDZ domain was confirmed for five of the compounds (CSC-03, CSC-04, CSC-43, FSC-231 and FSC-240) in a non-fluorescence based assay by their ability to inhibit pull-down of PICK1 by a C-terminal DAT GST fusion protein. CSC-03 displayed the highest apparent affinity (5.0 μM) in the FP assay, and was according to fluorescence resonance energy transfer (FRET) experiments capable of inhibiting the interaction between the C-terminus of the GluR2 subunit of the AMPA-type glutamate receptor and PICK1 in live cells. Additional experiments suggested that CSC-03 most likely is an irreversible inhibitor but with specificity for PICK1 since it did not bind three different PDZ domains of PSD-95. Summarized, our data suggest that FP based screening assays might be a widely applicable tool in the search for small molecule inhibitors of PDZ domain interactions.     

Simultaneous prediction of binding free energy and specificity for PDZ domain–peptide interactions

Interactions between protein domains and linear peptides underlie many biological processes. Among these interactions, the recognition of C-terminal peptides by PDZ domains is one of the most ubiquitous. In this work, we present a mathematical model for PDZ domain–peptide interactions capable of predicting both affinity and specificity of binding based on X-ray crystal structures and comparative modeling with Rosetta. We developed our mathematical model using a large phage display dataset describing binding specificity for a wild type PDZ domain and 91 single mutants, as well as binding affinity data for a wild type PDZ domain binding to 28 different peptides. Structural refinement was carried out through several Rosetta protocols, the most accurate of which included flexible peptide docking and several iterations of side chain repacking and backbone minimization. Our findings emphasize the importance of backbone flexibility and the energetic contributions of side chain-side chain hydrogen bonds in accurately predicting interactions. We also determined that predicting PDZ domain–peptide interactions became increasingly challenging as the length of the peptide increased in the N-terminal direction. In the training dataset, predicted binding energies correlated with those derived through calorimetry and specificity switches introduced through single mutations at interface positions were recapitulated. In independent tests, our best performing protocol was capable of predicting dissociation constants well within one order of magnitude of the experimental values and specificity profiles at the level of accuracy of previous studies. To our knowledge, this approach represents the first integrated protocol for predicting both affinity and specificity for PDZ domain–peptide interactions.     

Predicting protein-protein interactions using graph invariants and a neural network

The PDZ domain of proteins mediates a protein-protein interaction by recognizing the hydrophobic C-terminal tail of the target protein. One of the challenges put forth by the DREAM (Discussions on Reverse Engineering Assessment and Methods) 2009 Challenge consists of predicting a position weight matrix (PWM) that describes the specificity profile of five PDZ domains to their target peptides. We consider the primary structures of each of the five PDZ domains as a numerical sequence derived from graph-theoretic models of each of the individual amino acids in the protein sequence. Using available PDZ domain databases to obtain known targets, the graph-theoretic based numerical sequences are then used to train a neural network to recognize their targets. Given the challenge sequences, the target probabilities are computed and a corresponding position weight matrix is derived. In this work we present our method. The results of our method placed second in the DREAM 2009 challenge.     

Circumventing the problems caused by protein diversity in microarrays: implications for protein interaction networks

Protein microarrays provide a well-controlled, high-throughput way to uncover protein-protein interactions. One problem with this and other standardized assays, however, is that proteins vary considerably with respect to their physical properties. If a simple threshold-based approach is used to define protein-protein interactions, the resulting binary networks can be strongly biased. Here, we investigate the extent to which even closely related protein interaction domains vary when printed as microarrays. We find that, when a collection of well behaved, monomeric Src homology 2 (SH2) domains are printed at the same concentration, they vary by up to 50-fold with respect to the resulting surface density of active protein. When a threshold-based binding assay is performed on these domains using fluorescently labeled phosphopeptides, a misleading picture of the underlying biophysical interactions emerges. This problem can be circumvented, however, by obtaining saturation binding curves for each protein-peptide interaction. Importantly, the equilibrium dissociation constants obtained from these curves are independent of the surface density of active protein. We submit that an increased emphasis should be placed on obtaining quantitative information from protein microarrays and that this should serve as a more general goal in all efforts to define large-scale protein interaction networks.     

Novel peptide–protein assay for identification of antimicrobial peptides by fluorescence quenching

The specific interaction of peptides with proteins is often a key factor which determines biological activities. The determination of K(d) values of such interactions is commonly performed with fluorescence polarization. However, fluorescence polarization assays are prone to false-positive results due to the potential for non-specific interactions and only afford very low signal-to-background ratios. Here, we present as an alternative a fluorescence resonance energy transfer based quenching assay to measure peptide-protein interactions in solution. In a test setup where antimicrobial peptides were tested for their affinity towards the protein DnaK, the assay provided high specificity and good reproducibility and correlated with the results obtained by fluorescence polarization methods. Furthermore, we established a fast prescreening method which will allow a highly efficient screening of peptide libraries by reducing the amount of sample by 98% compared to conventional fluorescence polarization assays.     

A flexible docking procedure for the exploration of peptide binding selectivity to known structures and homology models of PDZ domains

PDZ domains are important scaffolding modules that typically bind to the C-termini of their interaction partners. Several structures of such complexes have been solved, revealing a conserved binding site in the PDZ domain and an extended conformation of the bound peptide. A compendium of information regarding PDZ complexes demonstrates that dissimilar C-terminal peptides bind to the same PDZ domain, and different PDZ domains can bind the same peptides. A detailed understanding of the PDZ-peptide recognition is needed to elucidate this complexity. To this end, we have designed a family of docking protocols for PDZ domains (termed PDZ-DocScheme) that is based on simulated annealing molecular dynamics and rotamer optimization, and is applicable to the docking of long peptides (20-40 rotatable bonds) to both known PDZ structures and to the more complicated problem of homology models of these domains. The resulting protocol reproduces the structures of PDZ complexes with peptides 4-8 amino acids long within 1-2 A from the experimental structure when the docking is performed to the original structure. If the structure of the target PDZ domain is an apo structure or a homology model, the docking protocol yields structures within 3 A in 9 out of 12 test cases. The automated docking procedure PDZ-DocScheme can serve in the generation of a structural context for validation of PDZ domain specificity from mutagenesis and ligand binding data.     

An improved method for the synthesis of cellulose membrane-bound peptides with free C termini is useful for PDZ domain binding studies

SPOT synthesis permits parallel synthesis and screening of thousands of cellulose membrane-bound peptides to study protein-protein interactions in a proteomic context. Recognition of C-terminal residues is one of the most common binding features of PDZ domains. Unfortunately, most solid support-bound peptide libraries lack a free C terminus due to C-terminal fixation on the solid support. To overcome this restriction, we developed a robust methodology based on our previous strategy for generating peptides with authentic C termini. To validate this improved method, we screened a human peptide library of 6223 C termini with the syntrophin PDZ domain. Furthermore, using the same library, new peptide ligands derived from membrane proteins and receptors were found for the ERBIN PDZ domain. Finally, we identified the protein kinase breakpoint cluster region, which is known as a negative regulator of cell proliferation and oncogenic transformation, as an ERBIN ligand.     

Thermodynamic basis for promiscuity and selectivity in protein-protein interactions: PDZ domains, a case study

Like other protein-protein interaction domains, PDZ domains are involved in many key cellular processes. These processes often require that specific multiprotein complexes be assembled, a task that PDZ domains accomplish by binding to specific peptide motifs in target proteins. However, a growing number of experimental studies show that PDZ domains (like other protein-protein interaction domains) can engage in a variety of interactions and bind distinct peptide motifs. Such promiscuity in ligand recognition raises intriguing questions about the molecular and thermodynamic mechanisms that can sustain it. To identify possible sources of promiscuity and selectivity underlying PDZ domain interactions, we performed molecular dynamics simulations of 20 to 25 ns on a set of 12 different PDZ domain complexes (for the proteins PSD-95, Syntenin, Erbin, GRIP, NHERF, Inad, Dishevelled, and Shank). The electrostatic, nonpolar, and configurational entropy binding contributions were evaluated using the MM/PBSA method combined with a quasi-harmonic analysis. The results revealed that PDZ domain interactions are characterized by overwhelmingly favorable nonpolar contributions and almost negligible electrostatic components, a mix that may readily sustain promiscuity. In addition, despite the structural similarity in fold and in recognition modes, the entropic and other dynamical aspects of binding were remarkably variable not only across PDZ domains but also for the same PDZ domain bound to distinct ligands. This variability suggests that entropic and dynamical components can play a role in determining selectivity either of PDZ domain interactions with peptide ligands or of PDZ domain complexes with downstream effectors.     

Discovery of a small-molecule inhibitor of Dvl–CXXC5 interaction by computational approaches

The Wnt/β-catenin signaling pathway plays a significant role in the control of osteoblastogenesis and bone formation. CXXC finger protein 5 (CXXC5) has been recently identified as a negative feedback regulator of osteoblast differentiation through a specific interaction with Dishevelled (Dvl) protein. It was reported that targeting the Dvl–CXXC5 interaction could be a novel anabolic therapeutic target for osteoporosis. In this study, complex structure of Dvl PDZ domain and CXXC5 peptide was simulated with molecular dynamics (MD). Based on the structural analysis of binding modes of MD-simulated Dvl PDZ domain with CXXC5 peptide and crystal Dvl PDZ domain with synthetic peptide–ligands, we generated two different pharmacophore models and applied pharmacophore-based virtual screening to discover potent inhibitors of the Dvl–CXXC5 interaction for the anabolic therapy of osteoporosis. Analysis of 16 compounds selected by means of a virtual screening protocol yielded four compounds that effectively disrupted the Dvl–CXXC5 interaction in the fluorescence polarization assay. Potential compounds were validated by fluorescence spectroscopy and nuclear magnetic resonance. We successfully identified a highly potent inhibitor, BMD4722, which directly binds to the Dvl PDZ domain and disrupts the Dvl–CXXC5 interaction. Overall, CXXC5–Dvl PDZ domain complex based pharmacophore combined with various traditional and simple computational methods is a promising approach for the development of modulators targeting the Dvl–CXXC5 interaction, and the potent inhibitor BMD4722 could serve as a starting point to discover or design more potent and specific the Dvl–CXXC5 interaction disruptors.     

A selective irreversible inhibitor targeting a PDZ protein interaction domain

Irreversible inhibitors of proteases have proven themselves useful tools for determining which proteases are active under given conditions in tissues or cells and for studying the functional role that a protease plays in physiological processes. The application of such techniques to the study of the activity and function of protein-protein interactions has been hindered by the lack of guiding principles for the mechanistic design of irreversible inhibitors targeting the "active site" of a protein interaction. We report herein the first example of a mechanism-based irreversible inhibitor of a protein interaction that has been specifically targeted to one member of the PDZ family of protein interaction domains: the second PDZ domain of the membrane-associated guanylate kinase MAGI3. This inhibitor was designed using rationally directed computational evaluation to take advantage of a conserved histidine in the PDZ domain by introducing an ionizable group that will be held in close proximity to that nucleophile during binding. The novel compound exhibits all of the characteristics of an irreversible inhibitor of the interaction of the tumor suppressor PTEN with MAGI3 in in vitro models. In cells, the inhibitor can be shown to release PTEN from sequestration by MAGI3 and consequently upregulate the PKB signaling pathway.     

Bridged peptide macrocycles as ligands for PDZ domain proteins

[reaction: see text] Conformationally constrained side chain-bridged cyclic peptides were prepared using bis-carboxylic acid ring spacers. These macrocyles were designed to inhibit protein-protein interactions mediated by the third PDZ domain (PDZ3) of a mammalian neuronal protein, PSD-95. Isothermal titration calorimetry (ITC) experiments measured dissociation constants in the low micromolar range. For each compound, the change in entropy (TdeltaS) of binding either is comparable in magnitude to the enthalpy change (deltaH) or is the predominant driving force for association.     

Detecting protein-protein interactions with a green fluorescent protein fragment reassembly trap: scope and mechanism

Identification of protein binding partners is one of the key challenges of proteomics. We recently introduced a screen for detecting protein-protein interactions based on reassembly of dissected fragments of green fluorescent protein fused to interacting peptides. Here, we present a set of comaintained Escherichia coli plasmids for the facile subcloning of fusions to the green fluorescent protein fragments. Using a library of antiparallel leucine zippers, we have shown that the screen can detect very weak interactions (K(D) approximately 1 mM). In vitro kinetics show that the reassembly reaction is essentially irreversible, suggesting that the screen may be useful for detecting transient interactions. Finally, we used the screen to discriminate cognate from noncognate protein-ligand interactions for tetratricopeptide repeat domains. These experiments demonstrate the general utility of the screen for larger proteins and elucidate mechanistic details to guide the further use of this screen in proteomic analysis. Additionally, this work gives insight into the positional inequivalence of stabilizing interactions in antiparallel coiled coils.     

Linderstrom-Lang-Schellman's model for protein stabilization revisited

The fact that cleavage of single peptide linkages in proteins often leads to extensive conformational alteration, including regions far removed from the cleavage site is not fully understood. We propose, based on the work of Linderstrom-Lang and Schellman, that disruption primarily occurs within protein structural domains that are stabilized by cooperative interactions and that cleavage of single peptide linkages of the domain perturbs the entire cooperative interaction. For this model we review experimental observations: on fragment complexation (ribonuclease A, staphylococcal nuclease and cytochrome c), destabilized N-terminal large fragments (ribonuclease A and nuclease), cooperative folding and stabilization of proteins (ribonuclease A, nuclease and cytochrome c), the close relationship of the three-dimensional structure between fragment complexes and the original protein (ribonuclease A and nuclease), ligand induced stabilization (nuclease), 3D domain swapping, circular permutation (dihydrofolate reductase), evolutionary conservation (cytochrome c fold). Based on analysis of these observations, we conclude that the cooperative interactions of domains are important for the mechanism of 3D domain swapping as well as for stabilization and thereby, determination of the ground state of native proteins. Furthermore, analysis of the observations reveals that domains generally contain a hydrophobic core. Further, based on studies of cytochrome c and the Tsao, Evans and Wennerstrom model of electrostatic interactions between two hydrophobic monolayers, we propose the model that the hydrophobic core of a domain is polarizable and responds to the surface charges through its polarizability to stabilize the domain, explaining in part the nature of the cooperative interactions.     

Energy exchange network of inter‐residue interactions within a thermally fluctuating protein molecule: A computational study

Protein function is regulated not only by the structure but also by physical dynamics and thermal fluctuations. We have developed the computer program, CURrent calculation for proteins (CURP), for the flow analysis of physical quantities within thermally fluctuating protein media. The CURP program was used to calculate the energy flow within the third PDZ domain of the neuronal protein PSD‐95, and the results were used to illustrate the energy exchange network of inter‐residue interactions based on atomistic molecular dynamics simulations. The removal of the α3 helix is known to decrease ligand affinity by 21‐fold without changing the overall protein structure; nevertheless, we demonstrated that the helix constitutes an essential part of the network graph. © 2015 Wiley Periodicals, Inc.     

Targeting specific PDZ domains of PSD-95; structural basis for enhanced affinity and enzymatic stability of a cyclic peptide

A cyclic peptide, Tyr-Lys-c[-Lys-Thr-Glu(betaAla)-]-Val, incorporating a beta-Ala lactam side chain linker and designed to target the PDZ domains of the postsynaptic density protein 95 (PSD-95), has been synthesized and structurally characterized by NMR while free and bound to the PDZ1 domain of PSD-95. While bound, the lactam linker of the peptide makes a number of unique contacts outside the canonical PDZ binding motif, providing a novel target for PDZ-domain specificity as well as producing a 10-fold enhancement in binding affinity. Additionally, the cyclization greatly enhances the enzymatic stability, increasing the duration that the peptide inhibits the association between PSD-95 and glutamate receptors, effectively inhibiting the clustering of kainate receptors for over 14 hr after application. Highly specific regulation of kainate receptor action may provide a novel route for treatment of drug addiction and epilepsy.     

Molecular self-assembly of solid-supported protein crystals

Highly ordered protein arrays have been proposed as a means for templating the organization of nanomaterials. Toward this end, we investigate the ability of the protein streptavidin to self-assemble into various configurations on solid-supported phospholipids. We identify two genetic variants of streptavidin (comprising amino acids 14-136 and 13-139) and examine their molecular organization at the liquid-solid interface. Our results demonstrate that the structural differences between these two protein variants affect both crystalline lattice and domain morphology. In general, these results for the liquid-solid interface are similar and consistent with those at the air-water interface with a few notable differences. Analogous to crystallization at the air-water interface, both forms of streptavidin yield H-like domains with lattice parameters that have C222 symmetry at pH 7. At pH 4, the native, truncated form of streptavidin yields needle-like domains consisting of molecules arranged in P1 symmetry. Unlike crystalline domains grown at the air-water interface, however, the lattice parameters of this P1 crystal are unique and have not yet been reported. The presence of a solid substrate does not appear to dramatically alter streptavidin's two-dimensional crystallization behavior, suggesting that local intermolecular interactions between proteins are more significant than interactions between the interface and protein. Our results also demonstrate that screening the electrostatic repulsion between protein molecules by modulating ionic strength will increase growth rate while decreasing crystalline domain size and macroscopic defects. Finally, we show that these domains are indeed functional by attaching biotinylated gold nanoparticles to the crystals. The ability to modulate molecular configuration, crystalline defects, and domain size on a functional array supports the potential application of this system toward materials assembly.     

Peptide Targeting of PDZ-Dependent Interactions as Pharmacological Intervention in Immune-Related Diseases

PDZ (postsynaptic density (PSD95), discs large (Dlg), and zonula occludens (ZO-1)-dependent interactions are widely distributed within different cell types and regulate a variety of cellular processes. To date, some of these interactions have been identified as targets of small molecules or peptides, mainly related to central nervous system disorders and cancer. Recently, the knowledge of PDZ proteins and their interactions has been extended to various cell types of the immune system, suggesting that their targeting by viral pathogens may constitute an immune evasion mechanism that favors viral replication and dissemination. Thus, the pharmacological modulation of these interactions, either with small molecules or peptides, could help in the control of some immune-related diseases. Deeper structural and functional knowledge of this kind of protein–protein interactions, especially in immune cells, will uncover novel pharmacological targets for a diversity of clinical conditions.     

Characterization of PDZ domain–peptide interactions using an integrated protocol of QM/MM,PB/SA,and CFEA analyses

Protein–protein interactions, particularly weak and transient ones, are often mediated by peptide recognition domains. Characterizing the interaction interface of domain–peptide complexes and analyzing binding specificity for modular domains are critical for deciphering protein–protein interaction networks. In this article, we report the successful use of an integrated computational protocol to dissect the energetic profile and structural basis of peptide binding to third PDZ domain (PDZ3) from the PSD-95 protein. This protocol employs rigorous quantum mechanics/molecular mechanics (QM/MM), semi-empirical Poisson–Boltzmann/surface area (PB/SA), and empirical conformational free energy analysis (CFEA) to quantitatively describe and decompose systematic energy changes arising from, respectively, noncovalent interaction, desolvation effect, and conformational entropy loss associated with the formation of 30 affinity-known PDZ3–peptide complexes. We show that the QM/MM-, PB/SA-, and CFEA-derived energy components can work together fairly well in reproducing experimentally measured affinity after a linearly weighting treatment, albeit they are not compatible with each other directly. We also demonstrate that: (1) noncovalent interaction and desolvation effect donate, respectively, stability and specificity to complex architecture, while entropy loss contributes modestly to binding; (2) P₀ and P₋₂ of peptide ligand are the most important positions for determining both the stability and specificity of the PDZ3–peptide complex, P₋₁ and P₋₃ can confer substantial stability (but not specificity) for the complex, and N-terminal P₋₄ and P₋₅ have only a very limited effect on binding.