Evaluating the Potential of Applying Machine Learning Tools to Metabolic Pathway Optimization

Successful engineering of a microbial host for efficient production of a target product from a given substrate can be viewed as an extensive optimization task. Such a task involves the selection of high activity enzymes as well as their gene expression regulatory control elements (i.e., promoters and ribosome binding sites). Finally, there is also the need to tune expression of multiple genes along a heterologous pathway to relieve constraints from rate-limiting step and help reduce metabolic burden on cells from unnecessary over-expression of high activity enzymes. While the aforementioned tasks could be performed through combinatorial experiments, such an approach incurs significant cost, time and effort, which is a handicap that can be relieved by application of modern machine learning tools. Such tools could attempt to predict high activity enzymes from sequence, but they are currently most usefully applied in classifying strong promoters from weaker ones as well as combinatorial tuning of expression of multiple genes. This perspective reviews the application of machine learning tools to aid metabolic pathway optimization through identifying challenges in metabolic engineering that could be overcome with the help of machine learning tools.

Control of Gene Expression With Quercetin-Responsive Modular Circuits

Control of gene expression is crucial for several biotechnological applications, especially for implementing predictable and controllable genetic circuits. Such circuits are often implemented with a transcriptional regulator activated by a specific signal. These regulators should work independently of the host machinery, with low gratuitous induction or crosstalk with host components. Moreover, the signal should also be orthogonal, recognized only by the regulator with minimal interference with the host operation. In this context, transcriptional regulators activated by plant metabolites as flavonoids emerge as candidates to control gene expression in bacteria. However, engineering novel circuits requires the characterization of the genetic parts (e.g., genes, promoters, ribosome binding sites, and terminators) in the host of interest. Therefore, we decomposed the QdoR regulatory system of B. subtilis, responsive to the flavonoid quercetin, and reassembled its parts into genetic circuits programmed to have different levels of gene expression and noise dependent on the concentration of quercetin. We showed that only one of the promoters regulated by QdoR worked well in E. coli, enabling the construction of other circuits induced by quercetin. The QdoR expression was modulated with constitutive promoters of different transcriptional strengths, leading to low expression levels when QdoR was highly expressed and vice versa. E. coli strains expressing high and low levels of QdoR were mixed and induced with the same quercetin concentration, resulting in two stable populations expressing different levels of their gene reporters. Besides, we demonstrated that the level of QdoR repression generated different noise levels in gene expression dependent on the concentration of quercetin. The circuits presented here can be exploited in applications requiring adjustment of gene expression and noise using a highly available and natural inducer as quercetin.

Analysis of Orientia tsutsugamushi promoter activity

ABSTRACT Orientia tsutsugamushi is an obligate intracellular bacterium that causes scrub typhus, a potentially fatal rickettsiosis, and for which no genetic tools exist. Critical to addressing this technical gap is to identify promoters for driving expression of antibiotic resistance and fluorescence reporter genes in O. tsutsugamushi. Such promoters would need to be highly conserved among strains, expressed throughout infection, and exhibit strong activity. We examined the untranslated regions upstream of O. tsutsugamushi genes encoding outer membrane protein A (ompA), 22-kDa type-specific antigen (tsa22) and tsa56. The bacterium transcribed all three during infection of monocytic, endothelial and epithelial cells. Examination of the upstream noncoding regions revealed putative ribosome binding sites, one set of predicted −10 and −35 sequences for ompA and two sets of −10 and −35 sequences for tsa22 and tsa56. Comparison of these regions among geographically diverse O. tsutsugamushi patient isolates revealed nucleotide identities ranging from 84.8 to 100.0%. Upon examination of the candidates for the ability to drive green fluorescence protein expression in Escherichia coli, varying activities were observed with one of the tsa22 promoters being the strongest. Identification and validation of O. tsutsugamushi promoters is an initial key step toward genetically manipulating this important pathogen.

Engineering a dynamic, controllable infectivity switch in bacteriophage T7

Transcriptional repressors play an important role in regulating phage genomes. Here, we examined how synthetic regulation based on repressors can be to create a dynamic, controllable infectivity switch in bacteriophage T7. We engineered T7 by replacing a large region of the early phage genome with combinations of ligand-responsive promoters and ribosome binding sites (RBS) designed to control the phage RNA polymerase. Phages with the engineered switch showed virulence comparable to wildtype when not repressed, indicating the phage can be engineered without a loss of fitness. When repressed, the most effective switch used a TetR promoter and a weak RBS, resulting in a two-fold increase in latent period (time to lyse host) and change in phage titer. Further, phage activity could be tuned by varying inducer concentrations. Our study provides a proof of concept for a simple circuit for user control over phage infectivity.

Tuning a bi-enzymatic cascade reaction in Escherichia coli to facilitate NADPH regeneration for ε-caprolactone production

Abstractε-Caprolactone is a monomer of poly(ε-caprolactone) which has been widely used in tissue engineering due to its biodegradability and biocompatibility. To meet the massive demand for this monomer, an efficient whole-cell biocatalytic approach was constructed to boost the ε-caprolactone production using cyclohexanol as substrate. Combining an alcohol dehydrogenase (ADH) with a cyclohexanone monooxygenase (CHMO) in Escherichia coli, a self-sufficient NADPH-cofactor regeneration system was obtained. Furthermore, some improved variants with the better substrate tolerance and higher catalytic ability to ε-caprolactone production were designed by regulating the ribosome binding sites. The best mutant strain exhibited an ε-caprolactone yield of 0.80 mol/mol using 60 mM cyclohexanol as substrate, while the starting strain only got a conversion of 0.38 mol/mol when 20 mM cyclohexanol was supplemented. The engineered whole-cell biocatalyst was used in four sequential batches to achieve a production of 126 mM ε-caprolactone with a high molar yield of 0.78 mol/mol.

Optimized photochemistry enables efficient analysis of dynamic RNA structuromes and interactomes in genetic and infectious diseases

Direct determination of RNA structures and interactions in living cells is critical for understanding their functions in normal physiology and disease states. Here, we present PARIS2, a dramatically improved method for RNA duplex determination in vivo with >4000-fold higher efficiency than previous methods. PARIS2 captures ribosome binding sites on mRNAs, reporting translation status on a transcriptome scale. Applying PARIS2 to the U8 snoRNA mutated in the neurological disorder LCC, we discover a network of dynamic RNA structures and interactions which are destabilized by patient mutations. We report the first whole genome structure of enterovirus D68, an RNA virus that causes polio-like symptoms, revealing highly dynamic conformations altered by antiviral drugs and different pathogenic strains. We also discover a replication-associated asymmetry on the (+) and (−) strands of the viral genome. This study establishes a powerful technology for efficient interrogation of the RNA structurome and interactome in human diseases.

Programmable cross-ribosome-binding sites to fine-tune the dynamic range of transcription factor-based biosensor

Currently, predictive translation tuning of regulatory elements to the desired output of transcription factor (TF)-based biosensors remains a challenge. The gene expression of a biosensor system must exhibit appropriate translation intensity, which is controlled by the ribosome-binding site (RBS), to achieve fine-tuning of its dynamic range (i.e. fold change in gene expression between the presence and absence of inducer) by adjusting the translation level of the TF and reporter. However, existing TF-based biosensors generally suffer from unpredictable dynamic range. Here, we elucidated the connections and partial mechanisms between RBS, translation level, protein folding and dynamic range, and presented a design platform that predictably tuned the dynamic range of biosensors based on deep learning of large datasets cross-RBSs (cRBSs). In doing so, a library containing 7053 designed cRBSs was divided into five sub-libraries through fluorescence-activated cell sorting to establish a classification model based on convolutional neural network in deep learning. Finally, the present work exhibited a powerful platform to enable predictable translation tuning of RBS to the dynamic range of biosensors.

Evaluating the Potential of Applying Machine Learning Tools to Metabolic Pathway Optimization

Successful engineering of a microbial host for efficient production of a target product from a given substrate can be viewed as an extensive optimization task. Such a task involves the selection of high activity enzymes as well as their gene expression regulatory control elements (i.e., promoters and ribosome binding sites). Finally, there is also the need to tune expression of multiple genes along a heterologous pathway to relieve constraints from rate-limiting step and help reduce metabolic burden on cells from unnecessary over-expression of high activity enzymes. While the aforementioned tasks could be performed through combinatorial experiments, such an approach incurs significant cost, time and effort, which is a handicap that can be relieved by application of modern machine learning tools. Such tools could attempt to predict high activity enzymes from sequence, but they are currently most usefully applied in classifying strong promoters from weaker ones as well as combinatorial tuning of expression of multiple genes. This perspective reviews the application of machine learning tools to aid metabolic pathway optimization through identifying challenges in metabolic engineering that could be overcome with the help of machine learning tools.

The discovery of a recombinant SARS2-like CoV strain provides insights into SARS and COVID-19 pandemics

In December 2019, the world awoke to a new zoonotic strain of coronavirus named severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2). In the present study, we identified key recombination regions and mutation sites cross the SARS-CoV-2, SARS-CoV and SARS-like CoV clusters of betacoronavirus subgroup B. Based on the analysis of these recombination events, we proposed that the Spike protein of SARS-CoV-2 may have more than one specific receptor for its function. In addition, we reported—for the first time—a recombination event of ORF8 at the whole-gene level in a bat and ultimately determined that ORF8 enhances the viral replication. In conjunction with our previous discoveries, we found that receptor binding abilities, junction furin cleavage sites (FCSs), strong first ribosome binding sites (RBSs) and enhanced ORF8s are main factors contributing to transmission, virulence and host adaptability of CoVs. Junction FCSs and enhanced ORF8s increase the efficiencies in viral entry into cells and replication, respectively while strong first RBSs enhance the translational initiation. The strong recombination ability of CoVs integrated these factors to generate multiple recombinant strains, two of which evolved into SARS-CoV and SARS-CoV-2 by nature selection, resulting in the SARS and COVID-19 pandemics.

Turning the screw: engineering extreme pH resistance in Escherichia coli through combinatorial synthetic operons

Adoption of microorganisms as platforms for sustainable biobased production requires host cells to be able to withstand harsh industrial conditions, which are usually far from the ones where these organisms are naturally adapted to thrive. However, novel survival mechanisms unearthed by the study of microbiomes from extreme habitats may be exploited to enhance microbial robustness under the strict conditions needed for different applications. In this work, synthetic biology approaches were used to engineer enhanced acidic tolerance in Escherichia coli under extreme conditions through the characterization of a library of twenty-seven unique operons composed of combinatorial assemblies of three novel genes from an extreme environment and three synthetic ribosome binding sites. The results here presented illustrate the efficacy of combining different metagenomic genes for tolerance in truly synthetic genetic operons, as expression of these gene clusters increased hundred-fold the survival percentage of cells exposed to an acidic shock in minimal media at pH 1.9 under aerobic conditions.