Evolvability as a function of purifying selection in TEM-1 ß-lactamase.

Evolvability­the capacity to generate beneficial heritable variation­is a central property of biological systems. However, its origins and modulation by environmental factors have not been examined systematically. Here, we analyze the fitness effects of all single mutations in TEM-1 ß-lactamase (4,997 variants) under selection for the wild-type function (ampicillin resistance) and for a new function (cefotaxime resistance). Tolerance to mutation in this enzyme is bimodal and dependent on the strength of purifying selection in vivo, a result that derives from a steep non-linear ampicillin-dependent relationship between biochemical activity and fitness. Interestingly, cefotaxime resistance emerges from mutations that are neutral at low levels of ampicillin but deleterious at high levels; thus the capacity to evolve new function also depends on the strength of selection. The key property controlling evolvability is an excess of enzymatic activity relative to the strength of selection, suggesting that fluctuating environments might select for high-activity enzymes.

Necl-4/Cadm4 recruits Par-3 to the Schwann cell adaxonal membrane.

Interactions between axons and Schwann cells are essential for the acquisition of Schwann cell radial and longitudinal polarity and myelin sheath assembly. In the internode, the largest of these longitudinal domains, axon-Schwann cell interactions are mediated by the Nectin-like (Necl) cell adhesion proteins, also known as SynCAMs or Cadms. In particular, Necl-1/Cadm3 expressed on the axon surface binds to Necl-4/Cadm4 expressed along the adaxonal membrane of myelinating Schwann cells. Necl-4 promotes myelination in vitro and is required for the timely onset of myelination and the fidelity of the organization of the myelin sheath and the internode in vivo. A key question is the identity of the downstream effectors of Necl-4 that mediate its effects. The cytoplasmic terminal region (CTR) of Necl-4 contains a PDZ-domain binding motif. Accordingly, we used the CTR of Necl-4 in an unbiased proteomic screen of PDZ-domain proteins. We identify Par-3, a multi-PDZ domain containing protein of the Par-aPKC polarity complex previously implicated in myelination, as an interacting protein. Necl-4 and Par-3 are colocalized along the inner Schwann cell membrane and coprecipitate from Schwann cell lysates. The CTR of Necl-4 binds to the first PDZ domain of Par-3 thereby recruiting Par-3 to sites of Necl-4/Necl-1 interaction. Knockdown of Necl-4 perturbs Par-3 localization to the inner membrane of Schwann cells in myelinating co-cultures. These findings implicate interactions of Necl-1/Necl-4 in the recruitment of Par-3 to the Schwann cell adaxonal membrane and the establishment of Schwann cell radial polarity.

Fidelity of tRNA 5'-maturation: a possible basis for the functional dependence of archaeal and eukaryal RNase P on multiple protein cofactors.

RNase P, which catalyzes tRNA 5'-maturation, typically comprises a catalytic RNase P RNA (RPR) and a varying number of RNase P proteins (RPPs): 1 in bacteria, at least 4 in archaea and 9 in eukarya. The four archaeal RPPs have eukaryotic homologs and function as heterodimers (POP5•RPP30 and RPP21•RPP29). By studying the archaeal Methanocaldococcus jannaschii RPR's cis cleavage of precursor tRNA(Gln) (pre-tRNA(Gln)), which lacks certain consensus structures/sequences needed for substrate recognition, we demonstrate that RPP21•RPP29 and POP5•RPP30 can rescue the RPR's mis-cleavage tendency independently by 4-fold and together by 25-fold, suggesting that they operate by distinct mechanisms. This synergistic and preferential shift toward correct cleavage results from the ability of archaeal RPPs to selectively increase the RPR's apparent rate of correct cleavage by 11,140-fold, compared to only 480-fold for mis-cleavage. Moreover, POP5•RPP30, like the bacterial RPP, helps normalize the RPR's rates of cleavage of non-consensus and consensus pre-tRNAs. We also show that archaeal and eukaryal RNase P, compared to their bacterial relatives, exhibit higher fidelity of 5'-maturation of pre-tRNA(Gln) and some of its mutant derivatives. Our results suggest that protein-rich RNase P variants might have evolved to support flexibility in substrate recognition while catalyzing efficient, high-fidelity 5'-processing.

Simultaneous enhancement of multiple functional properties using evolution-informed protein design.

A major challenge in protein design is to augment existing functional proteins with multiple property enhancements. Altering several properties likely necessitates numerous primary sequence changes, and novel methods are needed to accurately predict combinations of mutations that maintain or enhance function. Models of sequence co-variation (e.g., EVcouplings), which leverage extensive information about various protein properties and activities from homologous protein sequences, have proven effective for many applications including structure determination and mutation effect prediction. We apply EVcouplings to computationally design variants of the model protein TEM-1 ß-lactamase. Nearly all the 14 experimentally characterized designs were functional, including one with 84 mutations from the nearest natural homolog. The designs also had large increases in thermostability, increased activity on multiple substrates, and nearly identical structure to the wild type enzyme. This study highlights the efficacy of evolutionary models in guiding large sequence alterations to generate functional diversity for protein design applications.

Protein Structure from Experimental Evolution.

Natural evolution encodes rich information about the structure and function of biomolecules in the genetic record. Previously, statistical analysis of co-variation patterns in natural protein families has enabled the accurate computation of 3D structures. Here, we explored generating similar information by experimental evolution, starting from a single gene and performing multiple cycles of in vitro mutagenesis and functional selection in Escherichia coli. We evolved two antibiotic resistance proteins, ß-lactamase PSE1 and acetyltransferase AAC6, and obtained hundreds of thousands of diverse functional sequences. Using evolutionary coupling analysis, we inferred residue interaction constraints that were in agreement with contacts in known 3D structures, confirming genetic encoding of structural constraints in the selected sequences. Computational protein folding with interaction constraints then yielded 3D structures with the same fold as natural relatives. This work lays the foundation for a new experimental method (3Dseq) for protein structure determination, combining evolution experiments with inference of residue interactions from sequence information. A record of this paper's Transparent Peer Review process is included in the Supplemental Information.

Inferring protein 3D structure from deep mutation scans.

We describe an experimental method of three-dimensional (3D) structure determination that exploits the increasing ease of high-throughput mutational scans. Inspired by the success of using natural, evolutionary sequence covariation to compute protein and RNA folds, we explored whether 'laboratory', synthetic sequence variation might also yield 3D structures. We analyzed five large-scale mutational scans and discovered that the pairs of residues with the largest positive epistasis in the experiments are sufficient to determine the 3D fold. We show that the strongest epistatic pairings from genetic screens of three proteins, a ribozyme and a protein interaction reveal 3D contacts within and between macromolecules. Using these experimental epistatic pairs, we compute ab initio folds for a GB1 domain (within 1.8 Å of the crystal structure) and a WW domain (2.1 Å). We propose strategies that reduce the number of mutants needed for contact prediction, suggesting that genomics-based techniques can efficiently predict 3D structure.

A Protocol for Functional Assessment of Whole-Protein Saturation Mutagenesis Libraries Utilizing High-Throughput Sequencing.

Site-directed mutagenesis has long been used as a method to interrogate protein structure, function and evolution. Recent advances in massively-parallel sequencing technology have opened up the possibility of assessing the functional or fitness effects of large numbers of mutations simultaneously. Here, we present a protocol for experimentally determining the effects of all possible single amino acid mutations in a protein of interest utilizing high-throughput sequencing technology, using the 263 amino acid antibiotic resistance enzyme TEM-1 ß-lactamase as an example. In this approach, a whole-protein saturation mutagenesis library is constructed by site-directed mutagenic PCR, randomizing each position individually to all possible amino acids. The library is then transformed into bacteria, and selected for the ability to confer resistance to ß-lactam antibiotics. The fitness effect of each mutation is then determined by deep sequencing of the library before and after selection. Importantly, this protocol introduces methods which maximize sequencing read depth and permit the simultaneous selection of the entire mutation library, by mixing adjacent positions into groups of length accommodated by high-throughput sequencing read length and utilizing orthogonal primers to barcode each group. Representative results using this protocol are provided by assessing the fitness effects of all single amino acid mutations in TEM-1 at a clinically relevant dosage of ampicillin. The method should be easily extendable to other proteins for which a high-throughput selection assay is in place.

Quantifying protein-protein interactions in high throughput using protein domain microarrays.

Protein microarrays provide an efficient way to identify and quantify protein-protein interactions in high throughput. One drawback of this technique is that proteins show a broad range of physicochemical properties and are often difficult to produce recombinantly. To circumvent these problems, we have focused on families of protein interaction domains. Here we provide protocols for constructing microarrays of protein interaction domains in individual wells of 96-well microtiter plates, and for quantifying domain-peptide interactions in high throughput using fluorescently labeled synthetic peptides. As specific examples, we will describe the construction of microarrays of virtually every human Src homology 2 (SH2) and phosphotyrosine binding (PTB) domain, as well as microarrays of mouse PDZ domains, all produced recombinantly in Escherichia coli. For domains that mediate high-affinity interactions, such as SH2 and PTB domains, equilibrium dissociation constants (K(D)s) for their peptide ligands can be measured directly on arrays by obtaining saturation binding curves. For weaker binding domains, such as PDZ domains, arrays are best used to identify candidate interactions, which are then retested and quantified by fluorescence polarization. Overall, protein domain microarrays provide the ability to rapidly identify and quantify protein-ligand interactions with minimal sample consumption. Because entire domain families can be interrogated simultaneously, they provide a powerful way to assess binding selectivity on a proteome-wide scale and provide an unbiased perspective on the connectivity of protein-protein interaction networks.

Predicting PDZ domain-peptide interactions from primary sequences.

PDZ domains constitute one of the largest families of interaction domains and function by binding the C termini of their target proteins. Using Bayesian estimation, we constructed a three-dimensional extension of a position-specific scoring matrix that predicts to which peptides a PDZ domain will bind, given the primary sequences of the PDZ domain and the peptides. The model, which was trained using interaction data from 82 PDZ domains and 93 peptides encoded in the mouse genome, successfully predicts interactions involving other mouse PDZ domains, as well as PDZ domains from Drosophila melanogaster and, to a lesser extent, PDZ domains from Caenorhabditis elegans. The model also predicts the differential effects of point mutations in peptide ligands on their PDZ domain-binding affinities. Overall, we show that our approach captures, in a single model, the binding selectivity of the PDZ domain family.

PDZ domain binding selectivity is optimized across the mouse proteome.

PDZ domains have long been thought to cluster into discrete functional classes defined by their peptide-binding preferences. We used protein microarrays and quantitative fluorescence polarization to characterize the binding selectivity of 157 mouse PDZ domains with respect to 217 genome-encoded peptides. We then trained a multidomain selectivity model to predict PDZ domain-peptide interactions across the mouse proteome with an accuracy that exceeds many large-scale, experimental investigations of protein-protein interactions. Contrary to the current paradigm, PDZ domains do not fall into discrete classes; instead, they are evenly distributed throughout selectivity space, which suggests that they have been optimized across the proteome to minimize cross-reactivity. We predict that focusing on families of interaction domains, which facilitates the integration of experimentation and modeling, will play an increasingly important role in future investigations of protein function.

Uncovering quantitative protein interaction networks for mouse PDZ domains using protein microarrays.

One of the principal challenges in systems biology is to uncover the networks of protein-protein interactions that underlie most biological processes. To date, experimental efforts directed at this problem have largely produced only qualitative networks that are replete with false positives and false negatives. Here, we describe a domain-centered approach--compatible with genome-wide investigations--that enables us to measure the equilibrium dissociation constant (K(D)) of recombinant PDZ domains for fluorescently labeled peptides that represent physiologically relevant binding partners. Using a pilot set of 22 PDZ domains, 4 PDZ domain clusters, and 20 peptides, we define a gold standard dataset by determining the K(D) for all 520 PDZ-peptide combinations using fluorescence polarization. We then show that microarrays of PDZ domains identify interactions of moderate to high affinity (K(D) < or = 10 microM) in a high-throughput format with a false positive rate of 14% and a false negative rate of 14%. By combining the throughput of protein microarrays with the fidelity of fluorescence polarization, our domain/peptide-based strategy yields a quantitative network that faithfully recapitulates 85% of previously reported interactions and uncovers new biophysical interactions, many of which occur between proteins that are co-expressed. From a broader perspective, the selectivity data produced by this effort reveal a strong concordance between protein sequence and protein function, supporting a model in which interaction networks evolve through small steps that do not involve dramatic rewiring of the network.