Sequence-function Relationships Research Articles

Leveraging implicit knowledge in neural networks for functional dissection and engineering of proteins

Proteins are nature’s most versatile molecular machines. Deep neural networks trained on large protein datasets have recently been used to tackle the unmet complexity of protein sequence–function relationships. The implicit knowledge contained in these networks represents a powerful, but thus far inaccessible, resource for understanding protein biology. Here, we show that occlusion-based sensitivity analysis can leverage the knowledge present in deep-neural-network-based protein sequence classifiers to identify functionally relevant parts of proteins. We first validated our approach by successfully predicting positions that mediate small molecule binding or catalytic activity across different protein classes. Next, we inferred the impact of point mutations on the activity of ERK and HRas, signalling factors frequently deregulated in cancer. Finally, we used our approach to identify engineering hotspots in CRISPR–Cas9 and anti-CRISPR protein AcrIIA4. Our work demonstrates how implicit knowledge in neural networks can be harnessed for protein functional dissection and protein engineering. Deep neural networks are a powerful tool for predicting protein function, but identifying the specific parts of a protein sequence that are relevant to its functions remains a challenge. An occlusion-based sensitivity technique helps interpret these deep neural networks, and can guide protein engineering by locating functionally relevant protein positions.

Activity and Thermostability of GH5 Endoglucanase Chimeras from Mesophilic and Thermophilic Parents.

Cellulases from glycoside hydrolase family 5 (GH5) are key endoglucanase enzymes in the degradation of diverse polysaccharide substrates and are used in industrial enzyme cocktails to break down biomass. The GH5 family shares a canonical (βα)8-barrel structure, where each (βα) module is essential for the enzyme's stability and activity. Despite their shared topology, the thermostability of GH5 endoglucanase enzymes can vary significantly, and highly thermostable variants are often sought for industrial applications. Based on the previously characterized thermophilic GH5 endoglucanase Egl5A from Talaromyces emersonii (TeEgl5A), which has an optimal temperature of 90°C, we created 10 hybrid enzymes with elements of the mesophilic endoglucanase Cel5 from Stegonsporium opalus (SoCel5) to determine which elements are responsible for enhanced thermostability. Five of the expressed hybrid enzymes exhibit enzyme activity. Two of these hybrids exhibited pronounced increases in the temperature optimum (10 and 20°C), the temperature at which the protein lost 50% of its activity (T50) (15 and 19°C), and the melting temperature (Tm ) (16.5 and 22.9°C) and extended half-lives (t1/2) (∼240- and 650-fold at 55°C) relative to the values for the mesophilic parent enzyme and demonstrated improved catalytic efficiency on selected substrates. The successful hybridization strategies were validated experimentally in another GH5 endoglucanase, Cel5 from Aspergillus niger (AnCel5), which demonstrated a similar increase in thermostability. Based on molecular dynamics (MD) simulations of both the SoCel5 and TeEgl5A parent enzymes and their hybrids, we hypothesize that improved hydrophobic packing of the interface between α2 and α3 is the primary mechanism by which the hybrid enzymes increase their thermostability relative to that of the mesophilic parent SoCel5.IMPORTANCE Thermal stability is an essential property of enzymes in many industrial biotechnological applications, as high temperatures improve bioreactor throughput. Many protein engineering approaches, such as rational design and directed evolution, have been employed to improve the thermal properties of mesophilic enzymes. Structure-based recombination has also been used to fuse TIM barrel fragments, and even fragments from unrelated folds, to generate new structures. However, little research has been done on GH5 endoglucanases. In this study, two GH5 endoglucanases exhibiting TIM barrel structure, SoCel5 and TeEgl5A, with different thermal properties, were hybridized to study the roles of different (βα) motifs. This work illustrates the role that structure-guided recombination can play in helping to identify sequence function relationships within GH5 enzymes by supplementing natural diversity with synthetic diversity.

Applications of high-throughput sequencing to analyze and engineer ribozymes.

A large number of catalytic RNAs, or ribozymes, have been identified in the genomes of various organisms and viruses. Ribozymes are involved in biological processes such as regulation of gene expression and viral replication, but biological roles of many ribozymes still remain unknown. Ribozymes have also inspired researchers to engineer synthetic ribozymes that function as sensors or gene switches. To gain deeper understanding of the sequence-function relationship of ribozymes and to efficiently engineer synthetic ribozymes, a large number of ribozyme variants need to be examined which was limited to hundreds of sequences by Sanger sequencing. The advent of high-throughput sequencing technologies, however, has allowed us to sequence millions of ribozyme sequences at low cost. This review focuses on the recent applications of high-throughput sequencing to both characterize and engineer ribozymes, to highlight how the large-scale sequence data can advance ribozyme research and engineering.

Improved mutant function prediction via PACT: Protein Analysis and Classifier Toolkit

Deep mutational scanning experiments have enabled the measurement of the sequence-function relationship for thousands of mutations in a single experiment. The Protein Analysis and Classifier Toolkit (PACT) is a Python software package that marries the fitness metric of a given mutation within these experiments to sequence and structural features enabling downstream analyses. PACT enables the easy development of user sharable protocols for custom deep mutational scanning experiments as all code is modular and reusable between protocols. Protocols for mutational libraries with single or multiple mutations are included. To exemplify its utility, PACT assessed two deep mutational scanning datasets that measured the tradeoff of enzyme activity and enzyme stability. PACT efficiently evaluated classifiers that predict protein mutant function tested on deep mutational scanning screens. We found that the classifiers with the lowest false positive and highest true positive rate assesses sequence homology, contact number and if mutation involves proline. PACT and the processed datasets are distributed freely under the terms of the GPL-3 license. The source code is available at GitHub (https://github.com/JKlesmith/PACT). Supplementary data are available at Bioinformatics online.

DeepT3: deep convolutional neural networks accurately identify Gram-negative bacterial type III secreted effectors using the N-terminal sequence.

Various bacterial pathogens can deliver their secreted substrates also called effectors through Type III secretion systems (T3SSs) into host cells and cause diseases. Since T3SS secreted effectors (T3SEs) play important roles in pathogen-host interactions, identifying them is crucial to our understanding of the pathogenic mechanisms of T3SSs. However, the effectors display high level of sequence diversity, therefore making the identification a difficult process. There is a need to develop a novel and effective method to screen and select putative novel effectors from bacterial genomes that can be validated by a smaller number of key experiments. We develop a deep convolution neural network to directly classify any protein sequence into T3SEs or non-T3SEs, which is useful for both effector prediction and the study of sequence-function relationship. Different from traditional machine learning-based methods, our method automatically extracts T3SE-related features from a protein N-terminal sequence of 100 residues and maps it to the T3SEs space. We train and test our method on the datasets curated from 16 species, yielding an average classification accuracy of 83.7% in the 5-fold cross-validation and an accuracy of 92.6% for the test set. Moreover, when comparing with known state-of-the-art prediction methods, the accuracy of our method is 6.31-20.73% higher than previous methods on a common independent dataset. Besides, we visualize the convolutional kernels and successfully identify the key features of T3SEs, which contain important signal information for secretion. Finally, some effectors reported in the literature are used to further demonstrate the application of DeepT3. DeepT3 is freely available at: https://github.com/lje00006/DeepT3. Supplementary data are available at Bioinformatics online.

High activity chimeric snake gamma-type phospholipase A2 inhibitor created by DNA shuffling

The gamma-type inhibitor of snake venom phospholipase A2 (PLIγ) is expressed extensively in livers of both venomous and non-venomous snakes. It is not clear why PLIγs from different snake species possess diverse activities. To obtain high activity PLIγs and interpret the sequence-function relationships, we used DNA shuffling to hybridize the PLIγs of Sinonatrix annularis (saPLIγ) and Elaphe carinata (ecPLIγ). Chimera PLIγs (cPLIγ) of ∼550 bp were obtained by a series of gene manipulations including DNase I digestion, primer-free PCR, and PCR amplification with PLIγs primer pair. After successful insertion of cPLIγs into pCANTAB5e phage vector, the transformed TG1 strain of Esherichia coli was achieved. The cPLIγ phage library was produced and panned in a five-pace snake venom-coated immune tube. Three high affinity cPLIγ isoforms survived two rounds of panning. Prokaryote expression by the pET28c vector was employed for production of the three cPLIγs and the two parental PLIγs. These all showed anti-hemorrhage activity with cPLIγ 2 demonstrating superior inhibition to the parent PLIγs. Sequence alignment showed that the three kinds of cPLIγ were produced by gene splicing of S. annularis and E. carinata at different sites. Primary sequence changes brought regional changes in secondary and tertiary structure, which may explain the differences in PLIγ activity.

Codeinone reductase isoforms with differential stability, efficiency and product selectivity in opium poppy.

Codeinone reductase (COR) catalyzes the reversible NADPH-dependent reduction of codeinone to codeine as the penultimate step of morphine biosynthesis in opium poppy (Papaver somniferum). It also irreversibly reduces neopinone, which forms by spontaneous isomerization in aqueous solution from codeinone, to neopine. In a parallel pathway involving 3-O-demethylated analogs, COR converts morphinone to morphine, and neomorphinone to neomorphine. Similar to neopine, the formation of neomorphine by COR is irreversible. Neopine is a minor substrate for codeine O-demethylase (CODM), yielding morphine. In the plant, neopine levels are low and neomorphine has not been detected. Silencing of CODM leads to accumulation of upstream metabolites, such as codeine and thebaine, but does not result in a shift towards higher relative concentrations of neopine, suggesting a mechanism in the plant for limiting neopine production. In yeast (Saccharomyces cerevisiae) engineered to produce opiate alkaloids, the catalytic properties of COR lead to accumulation of neopine and neomorphine as major products. An isoform (COR-B) was isolated from opium poppy chemotype Bea's Choice that showed higher catalytic activity than previously characterized CORs, and it yielded mostly neopine invitro and in engineered yeast. Five catalytically distinct COR isoforms (COR1.1-1.4 and COR-B) were used to determine sequence-function relationships that influence product selectivity. Biochemical characterization and site-directed mutagenesis of native COR isoforms identified four residues (V25, K41, F129 and W279) that affected protein stability, reaction velocity, and product selectivity and output. Improvement of COR performance coupled with an ability to guide pathway flux is necessary to facilitate commercial production of opiate alkaloids in engineered microorganisms.

Fungal-type carbohydrate binding modules from the coccolithophore Emiliania huxleyi show binding affinity to cellulose and chitin.

Six fungal-type cellulose binding domains were found in the genome of the coccolithophore Emiliania huxleyi and cloned and expressed in Escherichia coli. Sequence comparison indicate high similarity to fungal cellulose binding domains, raising the question of why these domains exist in coccolithophores. The proteins were tested for binding with cellulose and chitin as ligands, which resulted in the identification of two functional carbohydrate binding modules: EHUX2 and EHUX4. Compared to benchmark fungal cellulose binding domain Cel7A-CBM1 from Trichoderma reesei, these proteins showed slightly lower binding to birch and bacterial cellulose, but were more efficient chitin binders. Finally, a set of cellulose binding domains was created based on the shuffling of one well-functioning and one non-functional domain. These were characterized in order to get more information of the binding domain’s sequence–function relationship, indicating characteristic differences between the molecular basis of cellulose versus chitin recognition. As previous reports have showed the presence of cellulose in coccoliths and here we find functional cellulose binding modules, a possible connection is discussed.

Affinity maturation of an TpoR targeting antibody in full-length IgG form for enhanced agonist activity.

It has been observed that converting scFv formatted antibodies to full-length IgG often associates with loss of affinity. We aim to address this issue in this paper by establishing an integrated affinity maturation method applying yeast display technology platform. To demonstrate that, we employed a human thrombopoietin receptor targeting antibody named 3D9 which was identified previously from a combinational antibody library in scFv-Fc fusion protein form. We have observed that significant potency loss happened when 3D9 was transformed to full-length IgG form. In this study, we tested whether the potency of the full-length IgG can be improved by affinity maturation of 3D9 using a modified Fab yeast display platform. An efficient CDR3 targeted mutagenesis strategy was designed for Fab library with pre-designed CDR diversity. Next generation sequencing was also used for evaluation of the enrichment process and investigation of sequence-function relationship of the antibody. A variant with improved affinity and higher potency was identified. The study demonstrates that the strategy we used here are efficient for optimizing affinity and activity of full-length IgGs.

SACCHARIS: an automated pipeline to streamline discovery of carbohydrate active enzyme activities within polyspecific families and de novo sequence datasets

BackgroundDeposition of new genetic sequences in online databases is expanding at an unprecedented rate. As a result, sequence identification continues to outpace functional characterization of carbohydrate active enzymes (CAZymes). In this paradigm, the discovery of enzymes with novel functions is often hindered by high volumes of uncharacterized sequences particularly when the enzyme sequence belongs to a family that exhibits diverse functional specificities (i.e., polyspecificity). Therefore, to direct sequence-based discovery and characterization of new enzyme activities we have developed an automated in silico pipeline entitled: Sequence Analysis and Clustering of CarboHydrate Active enzymes for Rapid Informed prediction of Specificity (SACCHARIS). This pipeline streamlines the selection of uncharacterized sequences for discovery of new CAZyme or CBM specificity from families currently maintained on the CAZy website or within user-defined datasets.ResultsSACCHARIS was used to generate a phylogenetic tree of a GH43, a CAZyme family with defined subfamily designations. This analysis confirmed that large datasets can be organized into sequence clusters of manageable sizes that possess related functions. Seeding this tree with a GH43 sequence from Bacteroides dorei DSM 17855 (BdGH43b, revealed it partitioned as a single sequence within the tree. This pattern was consistent with it possessing a unique enzyme activity for GH43 as BdGH43b is the first described α-glucanase described for this family. The capacity of SACCHARIS to extract and cluster characterized carbohydrate binding module sequences was demonstrated using family 6 CBMs (i.e., CBM6s). This CBM family displays a polyspecific ligand binding profile and contains many structurally determined members. Using SACCHARIS to identify a cluster of divergent sequences, a CBM6 sequence from a unique clade was demonstrated to bind yeast mannan, which represents the first description of an α-mannan binding CBM. Additionally, we have performed a CAZome analysis of an in-house sequenced bacterial genome and a comparative analysis of B. thetaiotaomicron VPI-5482 and B. thetaiotaomicron 7330, to demonstrate that SACCHARIS can generate “CAZome fingerprints”, which differentiate between the saccharolytic potential of two related strains in silico.ConclusionsEstablishing sequence-function and sequence-structure relationships in polyspecific CAZyme families are promising approaches for streamlining enzyme discovery. SACCHARIS facilitates this process by embedding CAZyme and CBM family trees generated from biochemically to structurally characterized sequences, with protein sequences that have unknown functions. In addition, these trees can be integrated with user-defined datasets (e.g., genomics, metagenomics, and transcriptomics) to inform experimental characterization of new CAZymes or CBMs not currently curated, and for researchers to compare differential sequence patterns between entire CAZomes. In this light, SACCHARIS provides an in silico tool that can be tailored for enzyme bioprospecting in datasets of increasing complexity and for diverse applications in glycobiotechnology.

Control of Cell-Selective Activity of Membrane-Active Polyleucine-Based Peptides using Database-Guided High-throughput Screening

We have developed a new 14-residue antimicrobial peptide template sequence containing four residues: L, D, K, and A. This template, called LDKA, which was designed using atomic detail structural information derived from unbiased long-timescale equilibrium molecular dynamics folding, partitioning, and pore assembly simulations, shows low micro-molar antibacterial activity against several gram-positive and gram-negative bacteria. Furthermore, LDKA was shown to form pores in microbial and eukaryotic membrane models using fluorescent dye leakage assays. A library of 2,916 peptides was designed using this template sequence and high-throughput screened using an orthogonal vesicle leakage assay. Dyes of different sizes were embedded into vesicles with varying lipid composition to simultaneously screen for both pore size and affinity to charged and neutral membranes. From this screen, 9 different LDKA variants were selected, sequenced, synthesized, and further characterized biophysically. Remarkably, even though the selected sequences displayed only minor mutational changes, each of these peptides has unique functional properties, forming either small or large pores and being selective for either neutral or anionic charged bilayers. Further screening of these peptides for bacterial and haemolytic activity revealed good correlation between charged membrane model leakage and antibacterial activity, as well as neutral membrane leakage and haemolytic activity. Antimicrobial activity is different for each peptide and depends on the bacterial species. Analysis of the sequences reveals no clear motifs or explanation for these strong preferences, but suggests that hydrophobic moment and sequence arrangement are more important for membrane selectivity than net charge. The results reveal that simple sequence-function relationships for peptides that can form a large repertoire of functional structures may remain elusive and that it may be the peptide sequences propensity to aggregate and assemble in a given environment that holds the key to functional prediction.

Multiplexed gene synthesis in emulsions for exploring protein functional landscapes.

Improving our ability to construct and functionally characterize DNA sequences would broadly accelerate progress in biology. Here, we introduce DropSynth, a scalable, low-cost method to build thousands of defined gene-length constructs in a pooled (multiplexed) manner. DropSynth uses a library of barcoded beads that pull down the oligonucleotides necessary for a gene's assembly, which are then processed and assembled in water-in-oil emulsions. We used DropSynth to successfully build more than 7000 synthetic genes that encode phylogenetically diverse homologs of two essential genes in Escherichia coli We tested the ability of phosphopantetheine adenylyltransferase homologs to complement a knockout E. coli strain in multiplex, revealing core functional motifs and reasons underlying homolog incompatibility. DropSynth coupled with multiplexed functional assays allows us to rationally explore sequence-function relationships at an unprecedented scale.

Characterizing Protein-Protein Interactions Using Deep Sequencing Coupled to Yeast Surface Display.

In this chapter, we discuss a method to determine the affinity and specificity of nearly all single-point mutants for a full-length protein binder. This method combines deep sequencing, comprehensive mutagenesis, yeast surface display, and fluorescence-activated cell sorting. This approach has been used to study sequence-function relationships for protein-protein interactions. The data can be used to determine the fine conformational epitope on the protein binder.

Coevolutionary Landscape of Kinase Family Proteins: Sequence Probabilities and Functional Motifs

The protein kinase catalytic domain is one of the most abundant domains across all branches of life. Although kinases share a common core function of phosphoryl-transfer, they also have wide functional diversity and play varied roles in cell signaling networks, and for this reason are implicated in a number of human diseases. This functional diversity is primarily achieved through sequence variation, and uncovering the sequence-function relationships for the kinase family is a major challenge. In this study we use a statistical inference technique inspired by statistical physics, which builds a coevolutionary “Potts” Hamiltonian model of sequence variation in a protein family. We show how this model has sufficient power to predict the probability of specific subsequences in the highly diverged kinase family, which we verify by comparing the model’s predictions with experimental observations in the Uniprot database. We show that the pairwise (residue-residue) interaction terms of the statistical model are necessary and sufficient to capture higher-than-pairwise mutation patterns of natural kinase sequences. We observe that previously identified functional sets of residues have much stronger correlated interaction scores than are typical.

Insights into substrate binding of ferulic acid esterases by arabinose and methyl hydroxycinnamate esters and molecular docking

Ferulic acid esterases (FAE, EC 3.1.1.73) cleave the arabinose hydroxycinnamate ester in plant hemicellulose and other related substrates. FAE are commonly categorised as type A-D based on catalytic activities towards model, short alkyl chain esters of hydroxycinnamates. However, this system correlates poorly with sequence and structural features of the enzymes. In this study, we investigated the basis of the type A categorisation of an FAE from Aspergillus niger, AnFaeA, by comparing its activity toward methyl and arabinose hydroxycinnamate esters. kcat/Km ratios revealed that AnFaeA hydrolysed arabinose ferulate 1600-fold, and arabinose caffeate 6.5 times more efficiently than their methyl ester counterparts. Furthermore, small docking studies showed that while all substrates adopted a catalytic orientation with requisite proximity to the catalytic serine, methyl caffeate and methyl p-coumarate preferentially formed alternative non-catalytic conformations that were energetically favoured. Arabinose ferulate was unable to adopt the alternative conformation while arabinose caffeate preferred the catalytic orientation. This study demonstrates that use of short alkyl chain hydroxycinnnamate esters can result in activity misclassification. The findings of this study provide a basis for developing a robust classification system for FAE and form the basis of sequence-function relationships for this class.

Pichia pastoris Alcohol Oxidase 1 (AOX1) Core Promoter Engineering by High Resolution Systematic Mutagenesis.

Unravelling the core promoter sequence-function relationship is fundamental for engineering transcription initiation and thereby a feasible "tuning knob" for fine-tuning expression in synthetic biology and metabolic engineering applications. Here a systematic replacement studies of the core promoter and 5' untranslated region (5'UTR) of the exceptionally strong and tightly methanol regulated Komagataella phaffii (syn. Pichia pastoris) alcohol oxidase 1 (AOX1) promoter at unprecedented resolution is performed. Adjacent triplets of the 200 bp long core promoter are mutated at a time by changing the wild-type sequence into cytosine or adenine triplets, resulting in 130 variants that are cloned upstream of an eGFP reporter gene providing a library for expression fine-tuning. Mutations in the TATA box motif, regions downstream of the transcription start site or next to the start codon in the 5'UTR had a significant effect on the eGFP fluorescence. Surprisingly, mutations in most other regions are tolerated, indicating that yeast core promoters can show a high tolerance toward small mutations, supporting regulatory models of degenerate motifs, or redundant design. The authors exploited these neutral core promoter positions, not affecting expression, to introduce extrinsic sequence elements such as cloning sites (allowing targeted core promoter/5'UTR modifications) and bacterial promoters (applicable in multi host vectors).

Directed Evolution of Proteins Based on Mutational Scanning.

Directed evolution has emerged as one of the most effective protein engineering methods in basic research as well as in applications in synthetic organic chemistry and biotechnology. The successful engineering of protein activity, allostery, binding affinity, expression, folding, fluorescence, solubility, substrate scope, selectivity (enantio-, stereo-, and regioselectivity), and/or stability (temperature, organic solvents, pH) is just limited by the throughput of the genetic selection, display, or screening system that is available for a given protein. Sometimes it is possible to analyze millions of protein variants from combinatorial libraries per day. In other cases, however, only a few hundred variants can be screened in a single day, and thus the creation of smaller yet smarter libraries is needed. Different strategies have been developed to create these libraries. One approach is to perform mutational scanning or to construct "mutability landscapes" in order to understand sequence-function relationships that can guide the actual directed evolution process. Herein we provide a protocol for economically constructing scanning mutagenesis libraries using a cytochrome P450 enzyme in a high-throughput manner. The goal is to engineer activity, regioselectivity, and stereoselectivity in the oxidative hydroxylation of a steroid, a challenging reaction in synthetic organic chemistry. Libraries based on mutability landscapes can be used to engineer any fitness trait of interest. The protocol is also useful for constructing gene libraries for deep mutational scanning experiments.

Protein Science by DNA Sequencing: How Advances in Molecular Biology Are Accelerating Biochemistry.

A fundamental goal of protein biochemistry is to determine the sequence-function relationship, but the vastness of sequence space makes comprehensive evaluation of this landscape difficult. However, advances in DNA synthesis and sequencing now allow researchers to assess the functional impact of every single mutation in many proteins, but challenges remain in library construction and the development of general assays applicable to a diverse range of protein functions. This Perspective briefly outlines the technical innovations in DNA manipulation that allow massively parallel protein biochemistry and then summarizes the methods currently available for library construction and the functional assays of protein variants. Areas in need of future innovation are highlighted with a particular focus on assay development and the use of computational analysis with machine learning to effectively traverse the sequence-function landscape. Finally, applications in the fundamentals of protein biochemistry, disease prediction, and protein engineering are presented.

Systematic optimization of L-tryptophan riboswitches for efficient monitoring of the metabolite in Escherichia coli.

Riboswitches form a class of genetically encoded sensor-regulators and are considered as promising tools for monitoring various metabolites. Functional parameters of a riboswitch, like dynamic or operational range, should be optimized before the riboswitch is implemented in a specific application for monitoring the target molecule efficiently. However, optimization of a riboswitch was not straightforward and required detailed studies owing to its complex sequence-function relationship. Here, we present three approaches for tuning and optimization of functional parameters of a riboswitch using an artificial L-tryptophan riboswitch as an example. First, the constitutive expression level was adjusted to control the dynamic range of an L-tryptophan riboswitch. The dynamic range increased as the constitutive expression level increased. Then, the function of a riboswitch-encoded protein was utilized to connect the regulatory response of the riboswitch to another outcome for amplifying the dynamic range. Riboswitch-mediated control of the host cell growth enabled the amplification of the riboswitch response. Finally, L-tryptophan aptamers with different dissociation constants were employed to alter the operational range of the riboswitch. The dose-response curve was shifted towards higher L-tryptophan concentrations when an aptamer with higher dissociation constant was employed. All strategies were effective in modifying the distinct functional parameters of the L-tryptophan riboswitch, and they could be easily applied to optimization of other riboswitches owing to their simplicity.

Sort-Seq Approach to Engineering a Formaldehyde-Inducible Promoter for Dynamically Regulated Escherichia coli Growth on Methanol.

Tight and tunable control of gene expression is a highly desirable goal in synthetic biology for constructing predictable gene circuits and achieving preferred phenotypes. Elucidating the sequence–function relationship of promoters is crucial for manipulating gene expression at the transcriptional level, particularly for inducible systems dependent on transcriptional regulators. Sort-seq methods employing fluorescence-activated cell sorting (FACS) and high-throughput sequencing allow for the quantitative analysis of sequence–function relationships in a robust and rapid way. Here we utilized a massively parallel sort-seq approach to analyze the formaldehyde-inducible Escherichia coli promoter (Pfrm) with single-nucleotide resolution. A library of mutated formaldehyde-inducible promoters was cloned upstream of gfp on a plasmid. The library was partitioned into bins via FACS on the basis of green fluorescent protein (GFP) expression level, and mutated promoters falling into each expression bin were identified with high-throughput sequencing. The resulting analysis identified two 19 base pair repressor binding sites, one upstream of the −35 RNA polymerase (RNAP) binding site and one overlapping with the −10 site, and assessed the relative importance of each position and base therein. Key mutations were identified for tuning expression levels and were used to engineer formaldehyde-inducible promoters with predictable activities. Engineered variants demonstrated up to 14-fold lower basal expression, 13-fold higher induced expression, and a 3.6-fold stronger response as indicated by relative dynamic range. Finally, an engineered formaldehyde-inducible promoter was employed to drive the expression of heterologous methanol assimilation genes and achieved increased biomass levels on methanol, a non-native substrate of E. coli.