Large scale active-learning-guided exploration for in vitro protein production optimization.

Lysate-based cell-free systems have become a major platform to study gene expression but batch-to-batch variation makes protein production difficult to predict. Here we describe an active learning approach to explore a combinatorial space of ~4,000,000 cell-free buffer compositions, maximizing protein production and identifying critical parameters involved in cell-free productivity. We also provide a one-step-method to achieve high quality predictions for protein production using minimal experimental effort regardless of the lysate quality.

Reinforcement Learning for Bioretrosynthesis.

Metabolic engineering aims to produce chemicals of interest from living organisms, to advance toward greener chemistry. Despite efforts, the research and development process is still long and costly, and efficient computational design tools are required to explore the chemical biosynthetic space. Here, we propose to explore the bioretrosynthesis space using an artificial intelligence based approach relying on the Monte Carlo Tree Search reinforcement learning method, guided by chemical similarity. We implement this method in RetroPath RL, an open-source and modular command line tool. We validate it on a golden data set of 20 manually curated experimental pathways as well as on a larger data set of 152 successful metabolic engineering projects. Moreover, we provide a novel feature that suggests potential media supplements to complement the enzymatic synthesis plan.

Metabolic perceptrons for neural computing in biological systems.

Synthetic biological circuits are promising tools for developing sophisticated systems for medical, industrial, and environmental applications. So far, circuit implementations commonly rely on gene expression regulation for information processing using digital logic. Here, we present a different approach for biological computation through metabolic circuits designed by computer-aided tools, implemented in both whole-cell and cell-free systems. We first combine metabolic transducers to build an analog adder, a device that sums up the concentrations of multiple input metabolites. Next, we build a weighted adder where the contributions of the different metabolites to the sum can be adjusted. Using a computational model fitted on experimental data, we finally implement two four-input perceptrons for desired binary classification of metabolite combinations by applying model-predicted weights to the metabolic perceptron. The perceptron-mediated neural computing introduced here lays the groundwork for more advanced metabolic circuits for rapid and scalable multiplex sensing.

Plug-and-play metabolic transducers expand the chemical detection space of cell-free biosensors.

Cell-free transcription-translation systems have great potential for biosensing, yet the range of detectable chemicals is limited. Here we provide a workflow to expand the range of molecules detectable by cell-free biosensors through combining synthetic metabolic cascades with transcription factor-based networks. These hybrid cell-free biosensors have a fast response time, strong signal response, and a high dynamic range. In addition, they are capable of functioning in a variety of complex media, including commercial beverages and human urine, in which they can be used to detect clinically relevant concentrations of small molecules. This work provides a foundation to engineer modular cell-free biosensors tailored for many applications.

Custom-made transcriptional biosensors for metabolic engineering.

Transcriptional biosensors allow screening, selection, or dynamic regulation of metabolic pathways, and are, therefore, an enabling technology for faster prototyping of metabolic engineering and sustainable chemistry. Recent advances have been made, allowing for routine use of heterologous transcription factors, and new strategies such as chimeric protein design allow engineers to tap into the reservoir of metabolite-binding proteins. However, extending the sensing scope of biosensors is only the first step, and computational models can help in fine-tuning properties of biosensors for custom-made behavior. Moreover, metabolic engineering is bound to benefit from advances in cell-free expression systems, either for faster prototyping of biosensors or for whole-pathway optimization, making it both a means and an end in biosensor design.

Models for Cell-Free Synthetic Biology: Make Prototyping Easier, Better, and Faster.

Cell-free TX-TL is an increasingly mature and useful platform for prototyping, testing, and engineering biological parts and systems. However, to fully accomplish the promises of synthetic biology, mathematical models are required to facilitate the design and predict the behavior of biological components in cell-free extracts. We review here the latest models accounting for transcription, translation, competition, and depletion of resources as well as genome scale models for lysate-based cell-free TX-TL systems, including their current limitations. These models will have to find ways to account for batch-to-batch variability before being quantitatively predictive in cell-free lysate-based platforms.

Enzyme Discovery: Enzyme Selection and Pathway Design.

In this protocol, we describe in silico design methods that can assist in the engineering of production pathways that are based on enzymatic transformations. The described protocols are the basis for automated processes to be integrated into an iterative Design-Build-Test-Learn cycle in synthetic biology for chemical production. Selecting the right enzyme sequence for a desired biocatalytic activity from the extensive catalogue of sequences available in databases is challenging and can dramatically influence the success of bioproducing chemical compounds. A method for enzyme selection is presented that helps identifying candidate enzyme sequences through a scoring approach that considers not only sequence homology but also reaction similarity. Selecting a viable biochemical pathway for compound production requires screening large sets of reactions in a process involving combinatorial complexity. A method for pathway design using retrosynthesis is presented. The protocol allows the discovery of alternative chemical pathways leading to the final product by using reaction rules of selectable degree of specificity. The protocols can be reversed through clustering discovery and product identification processes. The integration of these protocols into a general pipeline provides a toolbox for enhanced automated synthetic biology design and metabolic engineering.

Building a minimal and generalizable model of transcription factor-based biosensors: Showcasing flavonoids.

Progress in synthetic biology tools has transformed the way we engineer living cells. Applications of circuit design have reached a new level, offering solutions for metabolic engineering challenges that include developing screening approaches for libraries of pathway variants. The use of transcription-factor-based biosensors for screening has shown promising results, but the quantitative relationship between the sensors and the sensed molecules still needs more rational understanding. Herein, we have successfully developed a novel biosensor to detect pinocembrin based on a transcriptional regulator. The FdeR transcription factor (TF), known to respond to naringenin, was combined with a fluorescent reporter protein. By varying the copy number of its plasmid and the concentration of the biosensor TF through a combinatorial library, different responses have been recorded and modeled. The fitted model provides a tool to understand the impact of these parameters on the biosensor behavior in terms of dose-response and time curves and offers guidelines to build constructs oriented to increased sensitivity and or ability of linear detection at higher titers. Our model, the first to explicitly take into account the impact of plasmid copy number on biosensor sensitivity using Hill-based formalism, is able to explain uncharacterized systems without extensive knowledge of the properties of the TF. Moreover, it can be used to model the response of the biosensor to different compounds (here naringenin and pinocembrin) with minimal parameter refitting.

A dataset of small molecules triggering transcriptional and translational cellular responses.

The aim of this dataset is to identify and collect compounds that are known for being detectable by a living cell, through the action of a genetically encoded biosensor and is centred on bacterial transcription factors. Such a dataset should open the possibility to consider a wide range of applications in synthetic biology. The reader will find in this dataset the name of the compounds, their InChI (molecular structure), the publication where the detection was reported, the organism in which this was detected or engineered, the type of detection and experiment that was performed as well as the name of the biosensor. A comment field is also provided that explains why the compound was included in the dataset, based on quotes from the reference publication or the database it was extracted from. Manual curation of ACS Synthetic Biology abstracts (Volumes 1 to 6 and Volume 7 issue 1) was performed as well as extraction from the following databases: Bionemo v6.0 (Carbajosa et al., 2009) [1], RegTransbase r20120406 (Cipriano et al., 2013) [2], RegulonDB v9.0 (Gama-Castro et al., 2016) [3], RegPrecise v4.0 (Novichkov et al., 2013) [4] and Sigmol v20180122 (Rajput et al., 2016) [5].

Molecular structures enumeration and virtual screening in the chemical space with RetroPath2.0.

BACKGROUND: Network generation tools coupled with chemical reaction rules have been mainly developed for synthesis planning and more recently for metabolic engineering. Using the same core algorithm, these tools apply a set of rules to a source set of compounds, stopping when a sink set of compounds has been produced. When using the appropriate sink, source and rules, this core algorithm can be used for a variety of applications beyond those it has been developed for. RESULTS: Here, we showcase the use of the open source workflow RetroPath2.0. First, we mathematically prove that we can generate all structural isomers of a molecule using a reduced set of reaction rules. We then use this enumeration strategy to screen the chemical space around a set of monomers and predict their glass transition temperatures, as well as around aminoglycosides to search structures maximizing antibacterial activity. We also perform a screening around aminoglycosides with enzymatic reaction rules to ensure biosynthetic accessibility. We finally use our workflow on an E. coli model to complete E. coli metabolome, with novel molecules generated using promiscuous enzymatic reaction rules. These novel molecules are searched on the MS spectra of an E. coli cell lysate interfacing our workflow with OpenMS through the KNIME Analytics Platform. CONCLUSION: We provide an easy to use and modify, modular, and open-source workflow. We demonstrate its versatility through a variety of use cases including molecular structure enumeration, virtual screening in the chemical space, and metabolome completion. Because it is open source and freely available on MyExperiment.org, workflow community contributions should likely expand further the features of the tool, even beyond the use cases presented in the paper.

TRP channels in brown and white adipogenesis from human progenitors: new therapeutic targets and the caveats associated with the common antibiotic, streptomycin.

Transient receptor potential (TRP) channels are polymodal cell sensors responding to diverse stimuli and widely implicated in the developmental programs of numerous tissues. The evidence for an involvement of TRP family members in adipogenesis, however, is scant. We present the first comprehensive expression profile of all known 27 human TRP genes in mesenchymal progenitors cells during white or brown adipogenesis. Using positive trilineage differentiation as an exclusion criterion, TRP polycystic (P)3, and TPR melastatin (M)8 were found to be uniquely adipospecific. Knockdown of TRPP3 repressed the expression of the brown fat signature genes uncoupling protein (UCP)-1 and peroxisome proliferator-activated receptor γ coactivator (PGC)-1α as well as attenuated forskolin-stimulated uncoupled respiration. However, indices of generalized adipogenesis, such as lipid droplet morphology and fatty acid binding protein (FAPB)-4 expression, were not affected, indicating a principal mitochondrial role of TRPP3. Conversely, activating TRPM8 with menthol up-regulated UCP-1 expression and augmented uncoupled respiration predominantly in white adipocytes (browning), whereas streptomycin antagonized TRPM8-mediated calcium entry, downregulated UCP-1 expression, and mitigated uncoupled respiration; menthol was less capable of augmenting uncoupled respiration (thermogenesis) in brown adipocytes. TRPP3 and TRPM8 hence appear to be involved in the priming of mitochondria to perform uncoupled respiration downstream of adenylate cyclase. Our results also underscore the developmental caveats of using antibiotics in adipogenic studies.-Goralczyk, A., van Vijven, M., Koch, M., Badowski, C., Yassin, M. S., Toh, S.-A., Shabbir, A., Franco-Obregón, A., Raghunath, M. TRP channels in brown and white adipogenesis from human progenitors: new therapeutic targets and the caveats associated with the common antibiotic, streptomycin.