High-throughput Selection Research Articles

Growth-Based, High-Throughput Selection for NADH Preference in an Oxygen-Dependent Biocatalyst.

Cyclohexanone monooxygenases (CHMO) consume molecular oxygen and NADPH to catalyze the valuable oxidation of cyclic ketones. However, CHMO usage is restricted by poor stability and stringent specificity for NADPH. Efforts to engineer CHMO have been limited by the sensitivity of the enzyme to perturbations in conformational dynamics and long-range interactions that cannot be predicted. We demonstrate an aerobic, high-throughput growth selection platform in Escherichia coli for oxygenase evolution based on NADH redox balance. We applied this NADH-dependent selection to alter the cofactor specificity of CHMO to accept NADH, a less expensive cofactor than NADPH. We first identified the variant CHMO DTNP (S208D-K326T-K349N-L143P) with a ∼1200-fold relative cofactor specificity switch from NADPH to NADH compared to the wild type through semirational design. Molecular modeling suggests CHMO DTNP activity is driven by cooperative fine-tuning of cofactor contacts. Additional evolution of CHMO DTNP through random mutagenesis yielded the variant CHMO DTNPY with a ∼2900-fold relative specificity switch compared to the wild type afforded by an additional distal mutation, H163Y. These results highlight the difficulty in engineering functionally innovative variants from static models and rational designs, and the need for high throughput selection methods. Our introduced tools for oxygenase engineering accelerate the advancements of characteristics essential for industrial feasibility.

High-Throughput Selection and Characterisation of Aptamers on Optical Next-Generation Sequencers.

Aptamers feature a number of advantages, compared to antibodies. However, their application has been limited so far, mainly because of the complex selection process. ‘High-throughput sequencing fluorescent ligand interaction profiling’ (HiTS–FLIP) significantly increases the selection efficiency and is consequently a very powerful and versatile technology for the selection of high-performance aptamers. It is the first experiment to allow the direct and quantitative measurement of the affinity and specificity of millions of aptamers simultaneously by harnessing the potential of optical next-generation sequencing platforms to perform fluorescence-based binding assays on the clusters displayed on the flow cells and determining their sequence and position in regular high-throughput sequencing. Many variants of the experiment have been developed that allow automation and in situ conversion of DNA clusters into base-modified DNA, RNA, peptides, and even proteins. In addition, the information from mutational assays, performed with HiTS–FLIP, provides deep insights into the relationship between the sequence, structure, and function of aptamers. This enables a detailed understanding of the sequence-specific rules that determine affinity, and thus, supports the evolution of aptamers. Current variants of the HiTS–FLIP experiment and its application in the field of aptamer selection, characterisation, and optimisation are presented in this review.

High-throughput 96-well bioelectrochemical platform for screening of electroactive microbial consortia

The development of bioelectrochemical systems reinforces the necessity for the identification and engineering of electroactive bacteria with improved performance and novel biochemical properties. In this study, using a newly designed 96-well-plate array of microbial fuel cells (MFCs), we compared the electroactive capabilities of microbial communities derived from four mine drainages. The maximum power density of individual wells after two weeks of inoculation was 102 mW/m3, whereas the maximum current density was 1.6 A/m3. Transferring communities from individual wells into larger MFCs comprising low (20 mg/L) and high (200 mg/L) concentrations of Cu yielded maximum power densities of 445 and 58 mW/m3, respectively, with up to a 3.7 fold decrease in Cu2+ ions within 24 h. Electrochemical data analysis revealed that microbial consortia can be distinguished based on their electrochemical profiles. Our results showed that a 96-well MFC array is a suitable platform for high-throughput screening, selection, and subsequent source of enriched electroactive consortia. The quantitative and comparative analysis followed by principal component analysis indicated that the initial environmental conditions, as well as physical and chemical parameters of the lakes were crucial to develop an efficient electroactive community. Further applications of the proposed platform include genetic engineering, phenotype screening, and mutagenesis studies of both microbial communities and single cultures. This is the first time that a high-throughput MFC platform is used to evaluate the performance of multiple electroactive consortia towards Cu removal.

Exploring phenotypic variation and associations in root nodulation, morphological, and growth character traits among 50 guar genotypes

Guar [Cyamopsis tetragonoloba (L.) Taub] is a legume grown worldwide, tolerant of hot, dry, and saline conditions. Phenotyping is a key bottleneck in guar improvement, with belowground root nodule phenes being largely overlooked. To address this bottleneck, a greenhouse study was conducted to evaluate phenotypic variability and associations in 14 nodule and plant growth parameters in 50 guar genotypes over 48 days. Nodule mass per plant differed among genotypes and had moderate heritability (P = 0.011, H2 = 0.40), while nodule number did not differ and had low heritability (P = 0.247, H2 = 0.14), and the parameters were not correlated. Nodule mass was correlated with total plant N (r = 0.56), plus most biomass parameters, indicating phenotypic links between plant capacity to support nodulation, N2 accumulation, and plant biomass production. Stem diameter may be useful as an indirect, high-throughput selection trait for increased nodulation in guar, as there was a positive association between stem diameter and nodule mass. Branch number and stem diameter, which have great agronomic significance, had wide phenotypic variability and were highly heritable. Reproductive development was delayed in branched genotypes, which may have implications in breeding guar for environments that differ in growing season length. Despite low heritability, there was a wide range in specific leaf area (SLA) that may be exploited by breeders to optimize drought tolerance and N use efficiency traits. These analyses give novel and confirmatory insights on the nature and associations of phenotypic character traits in guar, particularly nodule traits, providing added tools for breeders in developing desired and beneficial plant ideotypes.

Drug discovery from traditional Chinese herbal medicine using high content imaging technology

Traditional Chinese medicine (TCM) has been widely used in China and other Asia countries for thousands of years to treat or prevent human diseases. Chinese herbal medicine, one of the most important components of TCM, has unique diversities in chemical components, and thus results in a wide range of biological activities. However, pharmaceutical industry is facing a major challenge to develop a large population of novel natural products and drugs, and considerable efforts have not resulted in high-volume of novel drug discovery and productivity. At present, increasing attention has been paid to Chinese herb medicine modernization in combination with the cutting-age technologies of drug discovery, especially the high throughput selection. High content imaging is an image-based high throughput screening method by using automated microscopy and image analysis software to capture and analyze phenotypes at a large scale to investigate multiple biological features simultaneously in the biological complex. Here, we described the pipeline of the state-of-the-art high content imaging technology, summarized the applications of the high content imaging technology in drug discovery from traditional Chinese herbal medicine, and finally discussed the current challenges and future perspectives for development of high throughput image-based screening technology in novel drug research and discovery.

Directed evolution of alditol oxidase for the production of optically pure D-glycerate from glycerol in the engineered Escherichia coli.

D-glycerate is an attractive chemical for a wide variety of pharmaceutical, cosmetic, biodegradable polymers, and other applications. Now several studies have been reported about the synthesis of glycerate by different biotechnological and chemical routes from glycerol or other feedstock. Here, we present the construction of an Escherichia coli engineered strain to produce optically pure D-glycerate by oxidizing glycerol with an evolved variant of alditol oxidase (AldO) from Streptomyces coelicolor. This is achieved by starting from a previously reported variant mAldO and employing three rounds of directed evolution, as well as the combination of growth-coupled high throughput selection with colorimetric screening. The variant eAldO3-24 displays a higher substrate affinity toward glycerol with 5.23-fold than the wild-type AldO, and a 1.85-fold increase of catalytic efficiency (kcat/KM). Then we introduced an isopropyl-β-D-thiogalactopyranoside (IPTG)-inducible T7 expression system in E. coli to overexpress the variant eAldO3-24, and deleted glucosylglycerate phosphorylase encoding gene ycjM to block the consumption of D-glycerate. Finally, the resulting strain TZ-170 produced 30.1 g/l D-glycerate at 70 h with a yield of 0.376 mol/mol in 5-l fed-batch fermentation.

Yeast Synthetic Minimal Biosensors for Evaluating Protein Production.

The unfolded protein response (UPR) is a highly conserved cellular response in eukaryotic cells to counteract endoplasmic reticulum (ER) stress, typically triggered by unfolded protein accumulation. In addition to its relevance to human diseases like cancer, the induction of the UPR has a significant impact on the recombinant protein production in eukaryotic cell factories, including the industrial workhorseSaccharomyces cerevisiae. Being able to accurately detect and measure this ER stress response in single cells, enables the rapid optimization of protein production conditions and high-throughput strain selection strategies. Current methodologies to monitor the UPR in S. cerevisiae are often temporally and spatially removed from the cultivation stage or lack updated systematic evaluation. To this end, we constructed and systematically evaluated a series of high-throughput UPR sensors by different designs, incorporating either yeast native UPR promoters or novel synthetic minimal UPR promoters. The native promoters of DER1 and ERO1 were identified to have suitable UPR biosensor properties and served as an expression level guide for orthogonal sensor benchmarking. Our best synthetic minimal sensor is only 98 bp in length, has minimal homology to other native yeast sequences and displayed superior sensor characteristics. The synthetic minimal UPR sensor was able to accurately distinguish between cells expressing different heterologous proteins and between the different secretion levels of the same protein. This work demonstrated the potential of synthetic UPR biosensors as high-throughput tools to predict the protein production capacity of strains, interrogate protein properties hampering their secretion, and guide rational engineering strategies for optimal heterologous protein production.

Ultradilute Measurements of Self-Association for the Identification of Antibodies with Favorable High-Concentration Solution Properties.

There is significant interest in formulating antibody therapeutics as concentrated liquid solutions, but early identification of developable antibodies with optimal manufacturability, stability, and delivery attributes remains challenging. Traditional methods of identifying developable mAbs with low self-association in common antibody formulations require relatively concentrated protein solutions (>1 mg/mL), and this single challenge has frustrated early-stage and large-scale identification of antibody candidates with drug-like colloidal properties. Here, we describe charge-stabilized self-interaction nanoparticle spectroscopy (CS-SINS), an affinity-capture nanoparticle assay that measures colloidal self-interactions at ultradilute antibody concentrations (0.01 mg/mL), and is predictive of antibody developability issues of high viscosity and opalescence that manifest at four orders of magnitude higher concentrations (>100 mg/mL). CS-SINS enables large-scale, high-throughput selection of developable antibodies during early discovery.

Development of Single-Walled Carbon Nanotube-Based Optical Sensors Via Data Analytics

The photoluminescence of single-walled carbon nanotubes (SWCNTs) offers unique advantages for the development of biological sensors, including the potential to sensitize emission to many classes of analytes. Methods to selectively transduce the local concentration of an analyte via an optical signal often involve the attachment of existing molecular recognition elements to the nanotube surface, or the development of selectivities by high-throughput screening or selection techniques. Due to the existence of several mechanisms of optical modulation of SWCNTs, as well as multiple carbon nanotube species with differing bandgaps, and the increasing number of excipients and polymers that can be used for colloidal suspension and non-covalent functionalization, data analytics can be used to facilitate selective recognition of bioanalytes to detect and process multi-parametric spectroscopic fingerprints. These strategies will be discussed for the detection of single and multiplexed analytes.

Specific bile salt hydrolase genes in Lactobacillus plantarum AR113 and relationship with bile salt resistance

The resistance of probiotics to bile salts is critical to their survive in the intestinal tract. Bile salt hydrolase (BSH) catalyzes the deconjugation of conjugated bile salts and is thought to be involved in bile salt resistance. The aim of this study was to explore the roles of different bsh genes in the bile salt resistance of Lactobacillus plantarum AR113. CRISPR-Cas9 gene editing was used to construct single, double, and triple bsh knockout strains. The functions of the bsh 1, and bsh 3 genes were found to be critical for BSH activity in L. plantarum AR113. Examination of six bile salts revealed that glycine-conjugated bile salts were the preferred substrate of BSH. The growth of strains with bsh1 or bsh3 knockout was significantly inhibited when treated with glycine-conjugated bile salts. However, strains with bsh2 or bsh4 knockout were not affected. These results suggested that bsh 1 and bsh 3 are closely related to bile salt resistance. This result will be useful for the rational, high-throughput selection of strains with high tolerance to bile salts.

Optical plasticity of mammalian cells.

Transparency is widespread in nature, ranging from transparent insect wings to ocular tissues that enable you to read this text, and transparent marine vertebrates. And yet, cells and tissue models in biology are usually strongly light scattering and optically opaque, precluding deep optical microscopy. Here we describe the directed evolution of cultured mammalian cells toward increased transparency. We find that mutations greatly diversify the optical phenotype of Chinese Hamster Ovary cells, a cultured mammalian cell line. Furthermore, only three rounds of high-throughput optical selection and competitive growth are required to yield fit cells with greatly improved transparency. Based on 15 monoclonal cell lines derived from this directed evolution experiment, we find that the evolved transparency frequently goes along with a reduction of nuclear granularity and physiological shifts in gene expression profiles. In the future this optical plasticity of mammalian cells may facilitate genetic clearance of living tissues for in vivo microscopy.

Inducible plasmid display system for high-throughput selection of proteins with improved solubility

Soluble expression of enzymes inside the cell is a prerequisite for the successful biotransformation of valuable products. Some key enzymes involved in biotransformation processes, however, are hardly expressed in their soluble forms. Here, we propose an inducible plasmid display, which is a molecular evolution strategy coupled with a high-throughput screening and/or selection method, as a simple and powerful tool for improving the solubility of target enzymes. Specifically, the Oct-1 DNA-binding domain and intein (i.e., auto-processing domain) were employed as anchoring and protein trans-splicing motifs to develop the system, in which the probability of protein trans-splicing is dependent on the soluble property of target proteins. The applicability of inducible plasmid display was investigated using an α-1,2-fucosyltransferase (FucT2) from Helicobacter pylori, a highly insoluble and unstable enzyme in the cytoplasmic space of Escherichia coli, as a model protein. One round of the overall inducible plasmid display process, which consists of in vivo production of FucT2 mutants and in vitro screening, enabled soluble expression of FucT2 and selection of plasmids containing the corresponding genetic information. The inducible plasmid display developed in this study will contribute to the rapid and efficient screening and/or selection of soluble proteins.

Genetically encoded multimode reporter of adaptor complex 3 traffic in budding yeast.

AP-3 (adaptor complex 3) mediates traffic from the late Golgi or early endosomes to late endosomal compartments. In mammals, mutations in AP-3 cause Hermansky-Pudlak syndrome type 2, cyclic neutropenias, and a form of epileptic encephalopathy. In budding yeast, AP-3 carries cargo directly from the trans-Golgi to the lysosomal vacuole. Despite the pathway's importance and its discovery two decades ago, rapid screens and selections for AP-3 mutants have not been available. We now report GNSI, a synthetic, genetically encoded reporter that allows rapid plate-based assessment of AP-3 functional deficiency, using either chromogenic or growth phenotype readouts. This system identifies defects in both the formation and consumption of AP-3 carrier vesicles and is adaptable to high-throughput screening or selection in both plate array and liquid batch culture formats. Episomal and integrating plasmids encoding GNSI have been submitted to the Addgene repository.

Isoleucyl-tRNA synthetase mutant based whole-cell biosensor for high-throughput selection of isoleucine overproducers

Whole-cell amino acid biosensors can sense the concentrations of certain amino acids and output easily detectable signals, which are important for construction of microbial producers. However, many reported biosensors have poor specificity because they also sense non-target amino acids. Besides, biosensors for many amino acids are still unavailable. In this study, we proposed a new strategy for constructing whole-cell biosensors based on aminoacyl-tRNA synthetases (aaRSs), which take the advantage of their universality and intrinsically specific binding ability to corresponding amino acids. Taking isoleucine biosensor as an example, we first mutated the isoleucyl-tRNA synthetase in Escherichia coli to dramatically decrease its affinity to isoleucine. The engineered cells specifically sensed isoleucine and output isoleucine dose-dependent cell growth as an easily detectable signal. To further expand the sensing range, an isoleucine exporter was overexpressed to enhance excretion of intracellular isoleucine. Since cells equipped with the optimized whole-cell biosensor showed accelerated growth when cells produced higher concentrations of isoleucine, the biosensor was successfully applied in high-throughput selection of isoleucine overproducers from random mutation libraries. This work demonstrates the feasibility of engineering aaRSs to construct a new kind of whole-cell biosensors for amino acids. Considering all twenty proteinogenic and many non-canonical amino acids have their specific aaRSs, this strategy should be useful for developing biosensors for various amino acids.

Leveraging Oxidative Stress to Regulate Redox Balance-Based, In Vivo Growth Selections for Oxygenase Engineering.

Directed evolution methods based on high-throughput growth selection enable efficient discovery of enzymes with improved function in vivo. High-throughput selection is particularly useful when engineering oxygenases, which are sensitive to structural perturbations and prone to uncoupled activity. In this work, we combine the principle that reactive oxygen species (ROS) produced by uncoupled oxygenase activity are detrimental to cell fitness with a redox balance-based growth selection method for oxygenase engineering that enables concurrent advancement in catalytic activity and coupling efficiency. As a proof-of-concept, we engineered P450-BM3 for degradation of acenaphthene (ACN), a recalcitrant environmental pollutant. Selection of site-saturation mutagenesis libraries in E.coli strain MX203 identified P450-BM3 variants GVQ-AL and GVQ-D222N, which have both improved coupling efficiency and catalytic activity compared to the starting variant. Computational modeling indicates that the discovered mutations cooperatively optimize binding pocket shape complementarity to ACN, and shift the protein's conformational dynamics to favor the lid-closed, catalytically competent state. We further demonstrated that the selective pressure on coupling efficiency can be tuned by modulating cellular ROS defense mechanisms.

Methods for Automated Single Cell Isolation and Sub-Cloning of Human Pluripotent Stem Cells.

Advances in human pluripotent stem cell (hPSC) techniques have led them to become a widely used and powerful tool for a vast array of applications, including disease modeling, developmental studies, drug discovery and testing, and emerging cell-based therapies. hPSC workflows that require clonal expansion from single cells, such as CRISPR/Cas9-mediated genome editing, face major challenges in terms of efficiency, cost, and precision. Classical sub-cloning approaches depend on limiting dilution and manual colony picking, which are both time-consuming and labor-intensive, and lack a real proof of clonality. Here we describe the application of three different automated cell isolation and dispensing devices that can enhance the single-cell cloning process for hPSCs. In combination with optimized cell culture conditions, these devices offer an attractive alternative compared to manual methods. We explore various aspects of each device system and define protocols for their practical application. Following the workflow described here, single cell-derived hPSC sub-clones from each system maintain pluripotency and genetic stability. Furthermore, the workflows can be applied to uncover karyotypic mosaicism prevalent in bulk hPSC cultures. Our robust automated workflow facilitates high-throughput hPSC clonal selection and expansion, urgently needed in the operational pipelines of hPSC applications. © 2020 The Authors. Basic Protocol: Efficient automated hPSC single cell seeding and clonal expansion using the iotaSciences IsoCell platform Alternate Protocol 1: hPSC single cell seeding and clonal expansion using the Cellenion CellenONE single-cell dispenser Alternate Protocol 2: hPSC single cell seeding and clonal expansion using the Cytena single-cell dispenser Support Protocol 1: Coating cell culture plates with Geltrex Support Protocol 2: hPSC maintenance in defined feeder-free conditions Support Protocol 3: hPSC passaging in clumps Support Protocol 4: Laminin 521 coating of IsoCell plates and 96-well/384-well plates Support Protocol 5: Preparation of medium containing anti-apoptotic small molecules Support Protocol 6: 96- and 384-well target plate preparation prior to single cell seeding Support Protocol 7: Single cell dissociation of hPSCs Support Protocol 8: IsoCell-, CellenONE-, and Cytena-derived hPSC clone subculture and expansion.

In Vivo, High-Throughput Selection of Thermostable Cyclohexanone Monooxygenase (CHMO).

Cyclohexanone monooxygenase (CHMO) from Acinetobacter sp. NCIMB 9871 is characterized as having wide substrate versatility for the biooxidation of (cyclic) ketones into esters and lactones with high stereospecificity. Despite industrial potential, CHMO usage is restricted by poor thermostability. Limited high-throughput screening tools and challenges in rationally engineering thermostability have impeded CHMO engineering efforts. We demonstrate the application of an aerobic, high-throughput growth selection platform in Escherichia coli (strain MX203) for the discovery of thermostability enhancing mutations for CHMO. The selection employs growth for the easy readout of CHMO activity in vivo, by requiring nicotinamide adenine dinucleotide phosphate (NADPH)-consuming enzymes to restore cellular redox balance. In the presence of the native substrate cyclohexanone, variant CHMO GV (A245G-A288V) was discovered from a random mutagenesis library screened at 42 °C. This variant retained native activity, exhibited ~4.4-fold improvement in residual activity after 30 °C incubation, and demonstrated ~5-fold higher cyclohexanone conversion at 37 °C compared to the wild type. Molecular modeling indicates that CHMO GV experiences more favorable residue packing and supports additional backbone hydrogen bonding. Further rational design resulted in CHMO A245G-A288V-T415C with improved thermostability at 45 °C. Our platform for oxygenase evolution enabled the rapid engineering of protein stability critical for industrial scalability.

Integration of genotypic, hyperspectral, and phenotypic data to improve biomass yield prediction in hybrid rye

Key messageHyperspectral and genomic data are effective predictors of biomass yield in winter rye. Variable selection procedures can improve the informativeness of reflectance data.Integrating cutting-edge technologies is imperative to sustainably breed crops for a growing global population. To predict dry matter yield (DMY) in winter rye (Secale cereale L.), we tested single-kernel models based on genomic (GBLUP) and hyperspectral reflectance-derived (HBLUP) relationship matrices, a multi-kernel model combining both matrices and a bivariate model fitted with plant height as a secondary trait. In total, 274 elite rye lines were genotyped using a 10 k-SNP array and phenotyped as testcrosses for DMY and plant height at four locations in Germany in two years (eight environments). Spectral data consisted of 400 discrete narrow bands ranging between 410 and 993 nm collected by an unmanned aerial vehicle (UAV) on two dates on each environment. To reduce data dimensionality, variable selection of bands was performed, resulting in the least absolute shrinkage and selection operator (Lasso) as the best method in terms of predictive abilities. The mean heritability of reflectance data was moderate (h^{2} = 0.72) and highly variable across the spectrum. Correlations between DMY and single bands were generally significant (p < 0.05) but low (≤ 0.29). Across environments and training set (TRN) sizes, the bivariate model showed the highest prediction abilities (0.56–0.75), followed by the multi-kernel (0.45–0.71) and single-kernel (0.33–0.61) models. With reduced TRN, HBLUP performed better than GBLUP. The HBLUP model fitted with a set of selected bands was preferred. Within and across environments, prediction abilities increased with larger TRN. Our results suggest that in the era of digital breeding, the integration of high-throughput phenotyping and genomic selection is a promising strategy to achieve superior selection gains in hybrid rye.

Empowering a Methanol-Dependent Escherichia coli via Adaptive Evolution Using a High-Throughput Microbial Microdroplet Culture System.

Recently, a methanol-essential Escherichia coli was constructed; this strain is highly dependent on a supply of gluconate as a co-substrate for growth. Adaptive laboratory evolution is commonly applied to obtain mutants with specific phenotypes under certain selected pressure. However, conventional adaptive evolution approaches are not only laborious and time consuming, but they also come with lower throughput and inefficiency. In order to empower the aforementioned E. coli with reduced gluconate usage and enhanced growth rate, an irrational strategy based on a microbial microdroplet culture (MMC) platform was developed in this study. Given the automatic high-throughput selection of the MMC, a three-stage regime of an adaptive evolution experiment via gradually decreasing the availability of gluconate during the cultivation was performed for 50 days continuously in order to obtain the mutations. Finally, a candidate mutant was obtained with a 3-fold faster growth rate, a 43% shorter lag phase, and 40% less gluconate usage compared with the starting strain. Moreover, the gene mutations of gntU, idnT, edd, and pckA were identified by analyzing the whole-genome sequencing of mutants, which are strongly associated with the efficiency of gluconate uptake and cell growth. In conclusion, we have successfully demonstrated the feasibility of using MMC platform to empower the target strain with specific requirements in a manner of labor, time efficiency, and directed evolution.

A Growth-Based, High-Throughput Selection Platform Enables Remodeling of 4-Hydroxybenzoate Hydroxylase Active Site.

We report an aerobic, growth-based selection platform founded on NADP(H) redox balance restoration in Escherichia coli, and we demonstrate its application in the high-throughput evolution of an oxygenase. A single round of selection followed by a facile growth assay enabled Pseudomonas aeruginosa 4-hydroxybenzoate hydroxylase (PobA) to efficiently hydroxylate both 4-hydroxybenzoic acid (4-HBA) and 3,4-dihydroxybenzoic acid (3,4-DHBA), two consecutive steps in gallic acid biosynthesis. Structural modeling suggests precise reorganization of active site hydrogen bond network, which is difficult to obtain without deep navigation of combinatorial sequence space. We envision universal application of this selection platform in engineering NADPH-dependent oxidoreductases.