Pooling for improved screening of combinatorial libraries for directed evolution

Following diversity generation in combinatorial protein engineering, a significant amount of effort is expended in screening the library for improved variants. Pooling, or combining multiple cells into the same assay well when screening, is a means to increase throughput and screen a larger portion of the library with less time and effort. We have developed and validated a Monte Carlo simulation model of pooling and used it to screen a library of beta-galactosidase mutants randomized in the active site to increase their activity toward fucosides. Here, we show that our model can successfully predict the number of highly improved mutants obtained via pooling and that pooling does increase the number of good mutants obtained. In unpooled conditions, we found a total of three mutants with higher activity toward p-nitrophenyl-beta-D-fucoside than that of the wild-type beta-galactosidase, whereas when pooling 10 cells per well we found a total of approximately 10 improved mutants. In addition, the number of "supermutants", those with the highest activity increase, was also higher when pooling was used. Pooling is a useful tool for increasing the efficiency of screening combinatorial protein engineering libraries.     

Unbiased libraries in protein directed evolution

Directed evolution is a powerful approach to study the molecular basis of protein evolution and to engineer proteins for a wide range of applications in synthetic organic chemistry and biotechnology. There are many methods based on random or focused mutagenesis to engineer successfully any protein trait. Focused approaches such as site-directed and saturation mutagenesis have become methods of choice for improving protein activity, selectivity, stability and many other traits because the screening step can be practically handled (bottleneck in directed evolution). Although novel mutagenesis methods based on CRISPR or solid-phase gene synthesis can eliminate bias when creating protein libraries, traditional PCR approaches, although imperfect, remain widely used due to their ease and low cost. One of the most common approaches in focused mutagenesis relies on NNK mutagenesis, however, the primer-based 22c-trick and small-intelligent methods have emerged as key tools for constructing less biased and unbiased libraries when all 20 canonical amino acids are needed for various reasons. In this minireview, we assess studies employing such methods for library creation and their areas of application. We also discuss the advantages and disadvantages of both methods and provide a perspective for creating smarter libraries.     

Computational design strategies for combinatorial libraries

Medicinal chemistry principles are being increasingly applied to the design of smaller, high purity, information-rich libraries. Recent computational advances in statistical methodology, the design of libraries to reduce ADMET problems, targeting protein families and revisiting natural products as sources of inspiration for scaffolds and reagents are all areas of progressive research.     

Optical processing of bacterial libraries for directed evolution

Selection of phenotypically distinct bacterial colonies on a Petri dish is typically performed by one of two methods: chemical or mechanical. Chemical methods (e.g., antibiotic selection) rely on inherent growth advantages of the unique phenotypes desired and thus have limited applicability. Mechanical methods are generally slow and require relatively large colonies (typically hundreds of colonies per plate). Here the use of imaged light to select bacterial colonies is explored, employing either photodynamic therapy agents or a ferrochelatase mutation in combination with porphyrin precursors to sensitize the bacteria to light and a computer-controlled light projection system to illuminate some bacterial colonies while leaving others in the dark. A CCD camera was used to distinguish between bacteria expressing green fluorescent protein (GFP) from nonfluorescent colonies. The fluorescence image from the camera was then used to create a virtual masking image for photoselection. Using a simple commercial projector it was possible to confer a 56-fold selective advantage to colonies expressing GFP. This represents a potentially powerful tool in directed evolution experiments using large libraries.     

Design of combinatorial protein libraries of optimal size

In this article we introduce a computational procedure, OPTCOMB (Optimal Pattern of Tiling for COMBinatorial library design), for designing protein hybrid libraries that optimally balance library size with quality. The proposed procedure is directly applicable to oligonucleotide ligation-based protocols such as GeneReassembly, DHR, SISDC, and many more. Given a set of parental sequences and the size ranges of the parental sequence fragments, OPTCOMB determines the optimal junction points (i.e., crossover positions) and the fragment contributing parental sequences at each one of the junction points. By rationally selecting the junction points and the contributing parental sequences, the number of clashes (i.e., unfavorable interactions) in the library is systematically minimized with the aim of improving the overall library quality. Using OPTCOMB, hybrid libraries containing fragments from three different dihydrofolate reductase sequences (Escherichia coli, Bacillus subtilis, and Lactobacillus casei) are computationally designed. Notably, we find that there exists an optimal library size when both the number of clashes between the fragments composing the library and the average number of clashes per hybrid in the library are minimized. Results reveal that the best library designs typically involve complex tiling patterns of parental segments of unequal size hard to infer without relying on computational means.     

Designing gene libraries from protein profiles for combinatorial protein experiments

Protein combinatorial libraries provide new ways to probe the determinants of folding and to discover novel proteins. Such libraries are often constructed by expressing an ensemble of partially random gene sequences. Given the intractably large number of possible sequences, some limitation on diversity must be imposed. A non-uniform distribution of nucleotides can be used to reduce the number of possible sequences and encode peptide sequences having a predetermined set of amino acid probabilities at each residue position, i.e., the amino acid sequence profile. Such profiles can be determined by inspection, multiple sequence alignment or physically-based computational methods. Here we present a computational method that takes as input a desired sequence profile and calculates the individual nucleotide probabilities among partially random genes. The calculated gene library can be readily used in the context of standard DNA synthesis to generate a protein library with essentially the desired profile. The fidelity between the desired profile and the calculated one coded by these partially random genes is quantitatively evaluated using the linear correlation coefficient and a relative entropy, each of which provides a measure of profile agreement at each position of the sequence. On average, this method of identifying such codon frequencies performs as well or better than other methods with regard to fidelity to the original profile. Importantly, the method presented here provides much better yields of complete sequences that do not contain stop codons, a feature that is particularly important when all or large fractions of a gene are subject to combinatorial mutation.     

A convenient and robust method for construction of combinatorial and random mutant libraries

Here we describe a convenient and robust ligase-independent method for construction of combinatorial and random mutant libraries. The homologous genes flanked by plasmid-derived DNA sequences are fragmented, and the random fragments are reassembled in a self-priming polymerase reaction to obtain chimeric genes. The product is then mixed with linearized vector and two pairs of flanking primers, followed by assembly of the chimeric genes and linearized vector by PCR to introduce recombinant plasmids of a combinatorial library. Commonly, it is difficult to find proper restriction sites during the construction of recombinant plasmids after DNA shuffling with multiple homologous genes. However, this disadvantage can be overcome by using the ligase-independent method because the steps of DNA digestion and ligation can be avoided during library construction. Similarly, DNA sequences with random mutations introduced by error-prone PCR can be used to construct recombinant plasmids of a random mutant library with this method. Additionally, this method can meet the needs of large and comprehensive DNA library construction.     

Random multi-recombinant PCR for the construction of combinatorial protein libraries

Development of a new methodology to create protein libraries, which enable the exploration of global protein space, is an exciting challenge. In this study we have developed random multi-recombinant PCR (RM-PCR), which permits the shuffling of several DNA fragments without homologous sequences. In order to evaluate this methodology, we applied it to create two different combinatorial DNA libraries. For the construction of a ‘random shuffling library’, RM-PCR was used to shuffle six DNA fragments each encoding 25 amino acids; this affords many different fragment sequences whose every position has an equal probability to encode any of the six blocks. For the construction of the ‘alternative splicing library’, RM-PCR was used to perform different alternative splicings at the DNA level, which also yields different block sequences. DNA sequencing of the RM-PCR products in both libraries revealed that most of the sequences were quite different, and had a long open reading frame without a frame shift or stop codon. Furthermore, no distinct bias among blocks was observed. Here we describe how to use RM-PCR for the construction of combinatorial DNA libraries, which encode protein libraries that would be suitable for selection experiments in the global protein space.     

Phage display combinatorial libraries of short peptides: ligand selection for protein purification

A library of heptapeptides displayed on the surface of filamentous phage M13 was evaluated as a potential source of affinity ligands for the purification of Rhizomucor miehei lipase. Two independent selection (biopanning) protocols were employed: the enzyme was either physically adsorbed on polystyrene or chemically immobilized on small magnetic beads. From screening with the polystyrene-adsorbed lipase it was found that there was a rapid enrichment of the library with “doublet” clones i.e. the phage species which carried two consecutive sequences of heptapeptides, whilst no such clones were observed from the screening using lipase attached to magnetic beads. The binding of the best clones to the enzyme was unambiguously confirmed by ELISA. However the synthetic heptapeptide of identical sequence to the best “monomeric” clone did not act as a satisfactory affinity ligand after immobilization on Sepharose. This indicated that the interaction with lipase was due to both the heptapeptide and the presence of a part of the phage coat protein. This conclusion was further verified by immobilizing the whole phage on the surface of magnetic beads and using the resulting conjugate as an affinity adsorbent. The scope of application of this methodology and the possibility of preparing phage-based affinity materials are briefly discussed.     

Genetic optimization of combinatorial libraries

Most agrochemical and pharmaceutical companies have set up high-throughput screening programs which require large numbers of compounds to screen. Combinatorial libraries provide an attractive way to deliver these compounds. A single combinatorial library with four variable positions can yield more than 10(12) potential compounds, if one assumes that about 1000 reagents are available for each position. This is far more than any high-throughput screening facility can afford to screen. We have proposed a method for iterative compound selection from large databases, which identifies the most active compounds by examining only a small fraction of the database. In this article, we describe the extension of this method to the problem of selecting compounds from large combinatorial libraries. Copyright 1998 John Wiley & Sons, Inc.     

Molecular evolution: dynamic combinatorial libraries, autocatalytic networks and the quest for molecular function

The principles of Darwinian evolution have been explored in molecular systems such as autocatalytic networks and dynamic combinatorial libraries. Molecular evolution in such systems manifests itself as ligand or receptor amplification by selection. Research efforts exploring these concepts may provide a mechanism for the identification of novel catalysts, molecular receptors and bioactive molecules.     

Improved PCR method for the creation of saturation mutagenesis libraries in directed evolution: application to difficult-to-amplify templates

Saturation mutagenesis constitutes a powerful method in the directed evolution of enzymes. Traditional protocols of whole plasmid amplification such as Stratagene’s QuikChange™ sometimes fail when the templates are difficult to amplify. In order to overcome such restrictions, we have devised a simple two-primer, two-stage polymerase chain reaction (PCR) method which constitutes an improvement over existing protocols. In the first stage of the PCR, both the mutagenic primer and the antiprimer that are not complementary anneal to the template. In the second stage, the amplified sequence is used as a megaprimer. Sites composed of one or more residues can be randomized in a single PCR reaction, irrespective of their location in the gene sequence.The method has been applied to several enzymes successfully, including P450-BM3 from Bacillus megaterium, the lipases from Pseudomonas aeruginosa and Candida antarctica and the epoxide hydrolase from Aspergillus niger. Here, we show that megaprimer size as well as the direction and design of the antiprimer are determining factors in the amplification of the plasmid. Comparison of the results with the performances of previous protocols reveals the efficiency of the improved method. Joaquin Sanchis, Layla Fernández, and J. Daniel Carballeira contributed equally.     

DNA display II. Genetic manipulation of combinatorial chemistry libraries for small-molecule evolution

Biological in vitro selection techniques, such as RNA aptamer methods and mRNA display, have proven to be powerful approaches for engineering molecules with novel functions. These techniques are based on iterative amplification of biopolymer libraries, interposed by selection for a desired functional property. Rare, promising compounds are enriched over multiple generations of a constantly replicating molecular population, and subsequently identified. The restriction of such methods to DNA, RNA, and polypeptides precludes their use for small-molecule discovery. To overcome this limitation, we have directed the synthesis of combinatorial chemistry libraries with DNA "genes," making possible iterative amplification of a nonbiological molecular species. By differential hybridization during the course of a traditional split-and-pool combinatorial synthesis, the DNA sequence of each gene is read out and translated into a unique small-molecule structure. This "chemical translation" provides practical access to synthetic compound populations 1 million-fold more complex than state-of-the-art combinatorial libraries. We carried out an in vitro selection experiment (iterated chemical translation, selection, and amplification) on a library of 10(6) nonnatural peptides. The library converged over three generations to a high-affinity protein ligand. The ability to genetically encode diverse classes of synthetic transformations enables the in vitro selection and potential evolution of an essentially limitless collection of compound families, opening new avenues to drug discovery, catalyst design, and the development of a materials science "biology."     

Different strategies to recover the activity of monomeric triosephosphate isomerase by directed evolution

A monomeric version of triosephosphate isomerase from Trypanosoma brucei, MonoTIM, has very low activity, and the same is true for all of the additional monomeric variants so far constructed. Here, we subjected MonoTIM to directed evolution schemes to achieve an activity improvement. The construction of a suitable strain for genetic selection provided an effective way to obtain active catalysts from a diverse population of protein variants. We used this tool to identify active mutants from two different strategies of mutagenesis: random mutagenesis of the whole gene and randomization of loop 2. Both strategies converged in the isolation of mutations Ala43 to Pro and Thr44 to either Ala or Ser, when randomizing the entire gene or to Arg in the case of randomization of loop 2. The kinetic characterization of the two more active mutants showed an increase of 11-fold in k(cat) and a reduction of 4-fold in K(m) for both of them, demonstrating the sensitivity of the selection method. A small difference in growth rate is observed when both mutant genes are compared, which seems to be attributable to a difference in solubility of the expressed proteins.     

Spatially addressed combinatorial protein libraries for recombinant antibody discovery and optimization

Antibody discovery typically uses hybridoma- or display-based selection approaches, which lack the advantages of directly screening spatially addressed compound libraries as in small-molecule discovery. Here we apply the latter strategy to antibody discovery, using a library of ～10,000 human germline antibody Fabs created by de novo DNA synthesis and automated protein expression and purification. In multiplexed screening assays, we obtained specific hits against seven of nine antigens. Using sequence-activity relationships and iterative mutagenesis, we optimized the binding affinities of two hits to the low nanomolar range. The matured Fabs showed full and partial antagonism activities in cell-based assays. Thus, protein drug leads can be discovered using surprisingly small libraries of proteins with known sequences, questioning the requirement for billions of members in an antibody discovery library. This methodology also provides sequence, expression and specificity information at the first step of the discovery process, and could enable novel antibody discovery in functional screens.     

Optimizing non-natural protein function with directed evolution

Developing technologies such as unnatural amino acid mutagenesis, non-natural cofactor engineering, and computational design are generating proteins with novel functions; these proteins, however, often do not reach performance targets and would benefit from further optimization. Evolutionary methods can complement these approaches: recent work combining unnatural amino acid mutagenesis and phage selection has created useful proteins of novel composition. Weak initial activity in a computationally designed enzyme has been improved by iterative rounds of mutagenesis and screening. A marriage of ingenuity and evolution will expand the scope of protein function well beyond Mother Nature's designs.     

Chemical and biochemical strategies for the randomization of protein encoding DNA sequences: library construction methods for directed evolution

Directed molecular evolution and combinatorial methodologies are playing an increasingly important role in the field of protein engineering. The general approach of generating a library of partially randomized genes, expressing the gene library to generate the proteins the library encodes and then screening the proteins for improved or modified characteristics has successfully been applied in the areas of protein–ligand binding, improving protein stability and modifying enzyme selectivity. A wide range of techniques are now available for generating gene libraries with different characteristics. This review will discuss these different methodologies, their accessibility and applicability to non-expert laboratories and the characteristics of the libraries they produce. The aim is to provide an up to date resource to allow groups interested in using directed evolution to identify the most appropriate methods for their purposes and to guide those moving on from initial experiments to more ambitious targets in the selection of library construction techniques. References are provided to original methodology papers and other recent examples from the primary literature that provide details of experimental methods.     

Identification of affinity ligands for protein purification from synthetic chemical combinatorial libraries

A method to screen combinatorial libraries for the development of selective ligands for protein affinity chromatographic purification is described. The method is based on the application of parallel combinatorial libraries, and it has several potential advantages. The screening procedure is simple and straightforward, and it does not require the chemical derivatization of the target proteins or even that the target protein be pure. The experiment can also be designed to select binders that are less likely to cause protein denaturation. Feasibility of this approach is demonstrated with a model study of the chromatographic purification of bovine albumin serum (BSA) and Avidin.     

Artificial ribonucleases: from combinatorial libraries to efficient catalysts of RNA cleavage

Combinatorial libraries of small organic compounds capable of cleaving RNA were synthesized. The compounds contain benzene ring substituted with two residues of bis quaternary salt of diazabicyclo[2.2.2]octane (DABCO) bearing hydrophobic fragments of different length and structure, attached to DABCO at the bridge position. These compounds, lacking traditional functionalities involved in transesterification reaction, exhibit pronounced RNA cleavage activity. To identify the most active artificial ribonucleases, sublibraries and truncated libraries, containing compounds lacking one of substituents were synthesized. Analysis of ribonuclease activity of truncated libraries resulted in identification of the most active compounds, which are characterized by the presence of at least one long oligomethylene substituent.