Reducing the vicissitudes of heterologous prochiral substrate catalysis by alcohol dehydrogenases through machine learning algorithms.

Alcohol dehydrogenases (ADHs) are popular catalysts for synthesizing chiral synthons a vital step for active pharmaceutical intermediate (API) production. They are grouped into three superfamilies namely, medium-chain (MDRs), short-chain dehydrogenase/reductases (SDRs), and iron-containing alcohol dehydrogenases. The former two are used extensively for producing various chiral synthons. Many studies screen multiple enzymes or engineer a specific enzyme for catalyzing a substrate of interest. These processes are resource-intensive and intricate. The current study attempts to decipher the ability to match different ADHs with their ideal substrates using machine learning algorithms. We explore the catalysis of 284 antibacterial ketone intermediates, against MDRs and SDRs to demonstrate a unique pattern of activity. To facilitate machine learning we curated a dataset comprising 33 features, encompassing 4 descriptors for each compound. Subsequently, an ensemble of machine learning techniques viz. Partial Least Squares (PLS) regression, k-Nearest Neighbors (kNN) regression, and Support Vector Machine (SVM) regression, was harnessed. Moreover, the assimilation of Principal Component Analysis (PCA) augmented precision and accuracy, thereby refining and demarcating diverse compound classes. As such, this classification is useful for discerning substrates amenable to diverse alcohol dehydrogenases, thereby mitigating the reliance on high-throughput screening or engineering in identifying the optimal enzyme for specific substrate.

A curated list of targeted optimized promiscuous ketoreductases (TOP-K).

Enzymes are either specific or promiscuous catalysts in nature. The latter is portrayed by protein families like CYP450Es, Aldo-ketoreductases and short/medium-chain dehydrogenases which participate in detoxification or secondary metabolite production. However, enzymes are evolutionarily 'blind' to an ever-increasing synthetic substrate library. Industries and laboratories have circumvented this by high-throughput screening or site-specific engineering to synthesize the product of interest. However, this paradigm entails cost and time-intensive one-enzyme, one-substrate catalysis model. One of the superfamilies regularly used for chiral alcohol synthesis are short-chain dehydrogenases/reductases (SDRs). Our objective is to determine a superset of promiscuous SDRs that can catalyze multiple ketones. They are typically classified into shorter 'Classical' and longer 'Extended' type ketoreductases. However, current analysis of modelled SDRs reveals a length-independent conserved N-terminus Rossmann-fold and a variable substrate-binding C-terminus substrate-binding region for both categories. The latter is recognized to influence the enzyme's flexibility and substrate promiscuity and we hypothesize these properties are directly linked with each other. We tested this by catalyzing ketone intermediates with the essential and specific enzyme: FabG_E, as well as non-essential SDRs such as UcpA and IdnO. The experimental results confirmed this biochemical-biophysical association, making it an interesting filter for ascertaining promiscuous enzymes. Hence, we created a dataset of physicochemical properties derived from the protein sequences and employed machine learning algorithms to examine potential candidates. This resulted in 24 targeted optimized ketoreductases (TOP-K) from 81 014 members. The experimental validation of select TOP-Ks demonstrated the correlation between the C-terminal lid-loop structure, enzyme flexibility and turnover rate on pro-pharmaceutical substrates.

Stairway to Stereoisomers: Engineering Short- and Medium-Chain Ketoreductases To Produce Chiral Alcohols.

The short- and medium-chain dehydrogenase/reductase superfamilies are responsible for most chiral alcohol production in laboratories and industries. In nature, they participate in diverse roles such as detoxification, housekeeping, secondary metabolite production, and catalysis of several chemicals with commercial and environmental significance. As a result, they are used in industries to create biopolymers, active pharmaceutical intermediates (APIs), and are also used as components of modular enzymes like polyketide synthases for fabricating bioactive molecules. Consequently, random, semi-rational and rational engineering have helped transform these enzymes into product-oriented efficient catalysts. The rise of newer synthetic chemicals and their enantiopure counterparts has proved challenging, and engineering them has been the subject of numerous studies. However, they are frequently limited to the synthesis of a single chiral alcohol. The study attempts to defragment and describe hotspots of engineering short- and medium-chain dehydrogenases/reductases for the production of chiral synthons.

Industrial light at the end of the iron-containing (group III) alcohol dehydrogenase tunnel.

There are three prominent alcohol dehydrogenases superfamilies: short-chain, medium-chain, and iron-containing alcohol dehydrogenases (FeADHs). Many members are valuable catalysts for producing industrially relevant products such as active pharmaceutical intermediates, chiral synthons, biopolymers, biofuels, and secondary metabolites. However, FeADHs are the least explored enzymes among the superfamilies for commercial tenacities. They portray a conserved structure having a "tunnel-like" cofactor and substrate binding site with particular functions, despite representing high sequence diversity. Interestingly, phylogenetic analysis demarcates enzymes catalyzing distinct native substrates where closely related clades convert similar molecules. Further, homologs from various mesophilic and thermophilic microbes have been explored for designing a solvent and temperature-resistant enzyme for industrial purposes. The review explores different iron-containing alcohol dehydrogenases potential engineering of the enzymes and substrates helpful in manufacturing commercial products.

Validated In Silico Population Model of Escherichia coli.

Flux balance analysis (FBA) and ordinary differential equation models have been instrumental in depicting the metabolic functioning of a cell. Nevertheless, they demonstrate a population's average behavior (summation of individuals), thereby portraying homogeneity. However, living organisms such as Escherichia coli contain more biochemical reactions than engaging metabolites, making them an underdetermined and degenerate system. This results in a heterogeneous population with varying metabolic patterns. We have formulated a population systems biology model that predicts this degeneracy by emulating a diverse metabolic makeup with unique biochemical signatures. The model mimics the universally accepted experimental view that a subpopulation of bacteria, even under normal growth conditions, renders a unique biochemical state, leading to the synthesis of metabolites and persister progenitors of antibiotic resistance and biofilms. We validate the platform's predictions by producing commercially important heterologous (isobutanol) and homologous (shikimate) metabolites. The predicted fluxes are tested in vitro resulting in 32- and 42-fold increased product of isobutanol and shikimate, respectively. Moreover, we authenticate the platform by mimicking a bacterial population in the presence of glyphosate, a metabolic pathway inhibitor. Here, we observe a fraction of subsisting persisters despite inhibition, thus affirming the signature of a heterogeneous populace. The platform has multiple uses based on the disposition of the user.

Structure and function relationship of OqxB efflux pump from Klebsiella pneumoniae.

OqxB is an RND (Resistance-Nodulation-Division) efflux pump that has emerged as a factor contributing to the antibiotic resistance in Klebsiella pneumoniae. OqxB underwent horizontal gene transfer and is now seen in other Gram-negative bacterial pathogens including Escherichia coli, Enterobacter cloacae and Salmonella spp., further disseminating multi-drug resistance. In this study, we describe crystal structure of OqxB with n-dodecyl-ß-D-maltoside (DDM) molecules bound in its substrate-binding pocket, at 1.85 Å resolution. We utilize this structure in computational studies to predict the key amino acids contributing to the efflux of fluoroquinolones by OqxB, distinct from analogous residues in related transporters AcrB and MexB. Finally, our complementation assays with mutated OqxB and minimum inhibitory concentration (MIC) experiments with clinical isolates of E. coli provide further evidence that the predicted structural features are indeed involved in ciprofloxacin efflux.

Correction to: FabG: from a core to circumstantial catalyst.

In the original publication of the article, under section Polyketide synthesis: a pathway similar to fatty acid synthesis, the sentences "Phylogeny of KS domains and proteins of FAS and PKS, inferred by Bayesian estimation. Numbers above branches indicate posterior clade probability values." and "Branch length indicates number of inferred amino acid changes per position." in the first paragraph were included inadvertently.

FabG: from a core to circumstantial catalyst.

Core biochemical pathways such as Fatty-acid synthesis II (FAS II) is ascribed to the synthesis of fatty-acids, biotin and lipoic acid in prokaryotes. It has two dehydrogenases namely, FabG and FabI which interact with the fatty-acid chain bound to Acyl-carrier protein (ACP), a well-studied enzyme which binds to substrates of varying lengths. This protein-protein interaction 'broadens' the active site of these dehydrogenases thus, contributing to their flexible nature. This property is exploited for catalysing numerous chiral synthons, alkanes, long-chain alcohols and secondary metabolites in industries especially with FabG. FASI relegates FASII in eukaryotes making it a 'relic gene pool' and an antibacterial drug target with diverse inhibitor and substrate markush. FabG often substitutes other dehydrogenases for producing secondary metabolites in nature. This redundancy is probably due to gene duplication or addition events possibly making FabG, a progenitor to some of the complex short-chain dehydrogenases used in organisms and industries today.

A strategy to identify a ketoreductase that preferentially synthesizes pharmaceutically relevant (S)-alcohols using whole-cell biotransformation.

INTRODUCTION: Chemical industries are constantly in search of an expeditious and environmentally benign method for producing chiral synthons. Ketoreductases have been used as catalysts for enantioselective conversion of desired prochiral ketones to their corresponding alcohol. We chose reported promiscuous ketoreductases belonging to different protein families and expressed them in E. coli to evaluate their ability as whole-cell catalysts for obtaining chiral alcohol intermediates of pharmaceutical importance. Apart from establishing a method to produce high value (S)-specific alcohols that have not been evaluated before, we propose an in silico analysis procedure to predict product chirality. RESULTS: Six enzymes originating from Sulfolobus sulfotaricus, Zygosaccharomyces rouxii, Hansenula polymorpha, Corynebacterium sp. ST-10, Synechococcus sp. PCC 7942 and Bacillus sp. ECU0013 with reported efficient activity for dissimilar substrates are compared here to arrive at an optimal enzyme for the method. Whole-cell catalysis of ketone intermediates for drugs like Aprepitant, Sitagliptin and Dolastatin using E. coli over-expressing these enzymes yielded (S)-specific chiral alcohols. We explain this chiral specificity for the best-performing enzyme, i.e., Z. rouxii ketoreductase using in silico modelling and MD simulations. This rationale was applied to five additional ketones that are used in the synthesis of Crizotinib, MA-20565 (an antifungal agent), Sulopenem, Rivastigmine, Talampanel and Barnidipine and predicted the yield of (S) enantiomers. Experimental evaluation matched the in silico analysis wherein ~ 95% (S)-specific alcohol with a chemical yield of 23-79% was obtained through biotransformation. Further, the cofactor re-cycling was optimized by switching the carbon source from glucose to sorbitol that improved the chemical yield to 85-99%. CONCLUSIONS: Here, we present a strategy to synthesize pharmaceutically relevant chiral alcohols by ketoreductases using a cofactor balanced whole-cell catalysis scheme that is useful for the industry. Based on the results obtained in these trials, Zygosaccharomyces rouxii ketoreductase was identified as a proficient enzyme to obtain (S)-specific alcohols from their respective ketones. The whole-cell catalyst when combined with nutrient modulation of using sorbitol as a carbon source helped obtain high enantiomeric and chemical yield.

Nitrothiophene carboxamides, a novel narrow spectrum antibacterial series: Mechanism of action and Efficacy.

The mechanism of efflux is a tour-de-force in the bacterial armoury that has thwarted the development of novel antibiotics. We report the discovery of a novel chemical series with potent antibacterial properties that was engineered to overcome efflux liability. Compounds liable to efflux specifically via the Resistance Nodulation and cell Division (RND) pump, AcrAB-TolC were chosen for a hit to lead progression. Using structure-based design, the compounds were optimised to lose their binding to the efflux pump, thereby making them potent on wild-type bacteria. We discovered these compounds to be pro-drugs that require activation in E. coli by specific bacterial nitroreductases NfsA and NfsB. Hit to lead chemistry led to the generation of compounds that were potent on wild-type and multi-drug resistant clinical isolates of E. coli, Shigella spp., and Salmonella spp. These compounds are bactericidal and efficacious in a mouse thigh infection model.

RAPD assisted selection of black gram (Vigna mungo L. Hepper) towards the development of multiple disease resistant germplasm.

Black gram (Vigna mungo L. Hepper), is an extensively studied food crop which is affected by many abiotic and biotic factors, especially diseases. The yield potential of Black gram is shallow due to lack of genetic variability and biotic stress susceptibility. Core biotic stress factors include mung bean yellow mosaic virus (MYMV), urdbean leaf crinkle virus (UCLV), wilt (Fusarium oxysporum) and powdery mildew (Erysiphe polygoni DC). Although many studies determine resistant varieties to a particular disease, however, it is often complimented by low yield and susceptibility to other diseases. Hence, this study focuses on investigating the genetic relationships among three varieties and nine accessions of black gram having disease resistance to previously described diseases and susceptibility using random amplified polymorphic deoxyribonucleic acid (RAPD) markers. A total of 33 RAPD primers were used for diversity analysis and yielded 206 fragments. Number of amplified fragments ranged from two (OPN-1) to 13 (OPF-1). The highest similarity coefficient was observed between IC-145202 and IC-164118 (0.921), while lowest similarity was between PU-31 and IC-145202 (0.572). The genetic diversity obtained in this study along with disease analysis suggests PU31as a useful variety for the development of markers linked to MYMV, UCLV, wilt and powdery mildew resistance by marker-assisted back cross breeding and facilitates the production of crosses with multiple disease resistance.

Delineating Substrate Diversity of Disparate Short-Chain Dehydrogenase Reductase from Debaryomyces hansenii.

Short-chain dehydrogenase reductases (SDRs) have been utilized for catalyzing the reduction of many aromatic/aliphatic prochiral ketones to their respective alcohols. However, there is a paucity of data that elucidates their innate biological role and diverse substrate space. In this study, we executed an in-depth biochemical characterization and substrate space mapping (with 278 prochiral ketones) of an unannotated SDR (DHK) from Debaryomyces hansenii and compared it with structurally and functionally characterized SDR Synechococcus elongatus. PCC 7942 FabG to delineate its industrial significance. It was observed that DHK was significantly more efficient than FabG, reducing a diverse set of ketones albeit at higher conversion rates. Comparison of the FabG structure with a homology model of DHK and a docking of substrate to both structures revealed the presence of additional flexible loops near the substrate binding site of DHK. The comparative elasticity of the cofactor and substrate binding site of FabG and DHK was experimentally substantiated using differential scanning fluorimetry. It is postulated that the loop flexibility may account for the superior catalytic efficiency of DHK although the positioning of the catalytic triad is conserved.

Keratin protein property based classification of mammals and non-mammals using machine learning techniques.

Keratin protein is ubiquitous in most vertebrates and invertebrates, and has several important cellular and extracellular functions that are related to survival and protection. Keratin function has played a significant role in the natural selection of an organism. Hence, it acts as a marker of evolution. Much information about an organism and its evolution can therefore be obtained by investigating this important protein. In the present study, Keratin sequences were extracted from public data repositories and various important sequential, structural and physicochemical properties were computed and used for preparing the dataset. The dataset containing two classes, namely mammals (Class-1) and non-mammals (Class-0), was prepared, and rigorous classification analysis was performed. To reduce the complexity of the dataset containing 56 parameters and to achieve improved accuracy, feature selection was done using the t-statistic. The 20 best features (parameters) were selected for further classification analysis using computational algorithms which included SVM, KNN, Neural Network, Logistic regression, Meta-modeling, Tree Induction, Rule Induction, Discriminant analysis and Bayesian Modeling. Statistical methods were used to evaluate the output. Logistic regression was found to be the most effective algorithm for classification, with greater than 96% accuracy using a 10-fold cross validation analysis. KNN, SVM and Rule Induction algorithms also were found to be efficacious for classification.