Algorithms and flowchart for the design of synthetic biochemical networks

New drugs are needed to assure effective therapies for previously untreated diseases, emerging diseases, and personalized medicine, but the process of drug development is complex, costly, and time-consuming. This is especially problematic considering that 90% of drug candidates in clinical trials are discarded due to unexpected toxicity or other secondary effects. This inefficiency threatens our health care system and economy. Despite the advances in the cellular metabolism, our knowledge of the mechanisms governing enzymatic biotransformations in cells is far from complete, in particular regarding degradation pathways, mode of action, or side effects of drugs. Examining the mechanisms of enzymatic reactions at the cellular scale could improve our fundamental understanding of their catalytic capability, and facilitate identifying and filling the knowledge gaps. The scale and the complexity of metabolic data is ever-expanding, requiring scientists to apply more advanced computational methods to systematically store, explore, and interpret the enzymatic potential of cells. The first step toward simulating enzymes in silico is to learn from their biochemical reactions in nature. To do this, we use distilled knowledge of known biochemistry in the form of generalized enzymatic reaction rules. Enzymatic rules are mathematical representations of enzymatic action mimicking the catalytic function of enzymes. They are formulated in a less specific manner (more promiscuous) to act on a broad range of substrates. In addition to reconstructing known biochemistry, the application of these reaction rules paves the way toward the discovery of novel enzymatic interactions. In this thesis, I developed computational models, tools, and methodologies to facilitate the study of metabolism and catalytic action of enzymes. We analyzed different aspects of metabolism through five distinct studies: In a first study, in order to provide a holistic view of currently known biochemistry, we gathered biochemical data from 14 sources, covering the known metabolic networks of all species. We integrated all biological data into a high-performance database based on ontology, named LCSB DB. We further expanded the scope of LCSB DB to cover all bioactive and chemicals. LCSB DB offers fast and efficient searching of biochemical data and serves as a platform for sharing, storing, and analyzing biochemical data. In a second study, we used enzymatic reaction rules to predict all theoretically possible metabolic reactions between biological and bioactive compounds in LCSB DB. In a third study, we developed a method to find enzymes are able to catalyze orphan and predicted reactions, called BridgIT. BridgIT uses the knowledge of reactive sites on substrates to find the most similar, known biochemical reactions. We then validated the utility of BridgIT in enzyme discovery for the design of de novo synthetic pathways producing tetrahydropalmatine and adipic acid. In the last study, we propose a workflow for rational drug design and systems-level analysis of drug metabolism, called NICEdrug.ch. NICEdrug.ch allows large-scale computational analysis of drug biochemistry (metabolic precursors or prodrugs and metabolic fate or degradation), enzymatic targets, and toxicity in the context of cellular metabolism. Finally, in the conclusion chapter, we discuss the contribution and the potential further applications of the computational tools that were developed in this thesis.

Modeling, predicting and mining metabolism at atom-level resolution

Jasmin Maria Hafner

Living organisms can catalyze many thousands of biochemical reactions that they use to convert energy and matter, which provides them with the essentials for life. The sum of these chemical reactions happening in an organism is called metabo-lism. Understanding metabolism is crucial to elucidate the fundamental principles of biology, and further to enable us to redesign it for the sustainable, biosynthetic pro-duction of bio-based fuels, commodity chemicals and medicines. Mathematical mod-els are essential to organize and understand the complexity of metabolism. They usu-ally represent metabolism as a network of reactions, but they tend to neglect the exact molecular structure of the metabolites. To redesign metabolic reactions, how-ever, a mechanistic understanding of metabolic reactions and their catalysts, proteins called enzymes, is essential. In this work, a mathematical description of enzymatic reaction mechanism, called generalized reaction rules, is applied to computationally simulate and predict meta-bolic processes at the level of atoms. Each reaction rule describes the catalytic activi-ty of an enzyme, or a group of enzymes, at the mechanistic level by encoding the re-arrangement of atoms in the reaction. The reaction rules are called âgeneralizedâ, because they mimic the ability of a single enzyme to catalyze multiple reactions by acting on a range of substrates. Using these reaction rules, we first developed a computational representation of me-tabolism that allowed tracking single atoms throughout complex metabolic reaction networks. The principle of atom-tracking was then used to develop a graph-theory based method to represent and analyze metabolic networks, and to reliably identify of metabolic pathways for the biosynthesis of chemicals. Next, we applied the gener-alized reaction rules to predict all possible novel, hypothetical reactions from known biological compounds, and we stored the five million generated novel reactions them in a database called ATLAS. Finally, the developed tools and resources were applied to specific engineering and research problems, such as the biosynthetic pathway design for the biofuel bisabolene and the plastic precursor 1,4-butanediol. We further pre-dicted a biosynthesis route for the pharmaceutical tetrahydropalmatine and engi-neered a yeast strain to produce it. Finally, we show that our tools can be used to mine available genome sequences to find organisms that can degrade xenobiotics. Our findings suggest that the atom-level representation of metabolism can greatly contribute to its understanding, exploration and prediction. Given the complexity of atom-level modeling of metabolic processes, we propose metrics that can approxi-mate the atom-level information to conserve the information at the level of big, hy-pothetical metabolic networks like ATLAS. This database plus the developed pathway search techniques form a valuable resource for scientist to help characterizing un-known biosynthesis pathways towards secondary metabolites, and for metabolic engi-neers for designing novel bioproduction pathways for chemicals. Hopefully, these considerations will contribute to a better understanding of metabolism, advance the exploration of the bioproduction of drugs and other valuable molecules, and acceler-ate metabolic engineering efforts to realize the switch from a petroleum-based chem-ical industry towards a more sustainable, bio-based production of societyâs chemical needs.

EPFL2020

Enzyme annotation for orphan and novel reactions using knowledge of substrate reactive sites

Noushin Hadadi, Vassily Hatzimanikatis, Ljubisa Miskovic, Marianne Seijo

Thousands of biochemical reactions with characterized activities are “orphan,” meaning they cannot be assigned to a specific enzyme, leaving gaps in metabolic pathways. Novel reactions predicted by pathway-generation tools also lack associated sequences, limiting protein engineering applications. Associating orphan and novel reactions with known biochemistry and suggesting enzymes to catalyze them is a daunting problem. We propose the method BridgIT to identify candidate genes and catalyzing proteins for these reactions. This method introduces information about the enzyme binding pocket into reaction-similarity comparisons. BridgIT assesses the similarity of two reactions, one orphan and one well-characterized nonorphan reaction, using their substrate reactive sites, their surrounding structures, and the structures of the generated products to suggest enzymes that catalyze the most-similar nonorphan reactions as candidates for also catalyzing the orphan ones. We performed two large-scale validation studies to test BridgIT predictions against experimental biochemical evidence. For the 234 orphan reactions from the Kyoto Encyclopedia of Genes and Genomes (KEGG) 2011 (a comprehensive enzymatic-reaction database) that became nonorphan in KEGG 2018, BridgIT predicted the exact or a highly related enzyme for 211 of them. Moreover, for 334 of 379 novel reactions in 2014 that were later cataloged in KEGG 2018, BridgIT predicted the exact or highly similar enzymes. BridgIT requires knowledge about only four connecting bonds around the atoms of the reactive sites to correctly annotate proteins for 93% of analyzed enzymatic reactions. Increasing to seven connecting bonds allowed for the accurate identification of a sequence for nearly all known enzymatic reactions.

PNAS2019

ATLAS of Biochemistry: A Repository of All Possible Biochemical Reactions for Synthetic Biology and Metabolic Engineering Studies

Noushin Hadadi, Jasmin Maria Hafner, Vassily Hatzimanikatis, Adrian Shajkofci, Aikaterini Zisaki

Because the complexity of metabolism cannot be intuitively understood or analyzed, computational methods are indispensable for studying biochemistry and deepening our understanding of cellular metabolism to promote new discoveries. We used the computational framework BNICE.ch along with cheminformatic tools to assemble the whole theoretical reactome from the known metabolome through expansion of the known biochemistry presented in the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. We constructed the ATLAS of Biochemistry, a database of all theoretical biochemical reactions based on known biochemical principles and compounds. ATLAS includes more than 130 000 hypothetical enzymatic reactions that KEGG connect two or more KEGG metabolites through novel enzymatic reactions that have never been reported to occur in living organisms. Moreover, ATLAS reactions integrate 42% of KEGG metabolites that are not currently present in any KEGG reaction into one or more novel enzymatic reactions. The generated repository of information is organized in a Web-based database (http://lcsb-databases.epfl.ch/atlas/) that allows the user to search for all possible routes from any substrate compound to any product. The resulting pathways involve known and novel enzymatic steps that may indicate unidentified enzymatic activities and provide potential targets for protein engineering. Our approach of introducing novel biochemistry into pathway design and associated databases will be important for synthetic biology and metabolic engineering.

Amer Chemical Soc2016

Modeling, predicting and mining metabolism at atom-level resolution

Jasmin Maria Hafner

EPFL2020