Homa Mohammadi-Peyhani1,2 & Jasmin Hafner1,2, Anastasia Sveshnikova1, Victor Viterbo1, Vassily Hatzimanikatis1*
1 Laboratory of Computational Systems Biotechnology, École Polytechnique Fédérale de Lausanne, EPFL, Lausanne, Switzerland
2 These authors contributed equally
*Corresponding author: vassily [kropka] hatzimanikatisepfl [kropka] ch
Metabolic dark matter describes the gaps in today’s knowledge of metabolic processes, which has been accumulated in the past decades of biochemical research. The continuous growth of biochemical reaction databases, the sustained discovery of novel natural products and the difficulty to predict the behavior of cellular metabolism strongly indicate that many metabolic components are currently missing from our biochemical record. Yet, these unknowns not only undermine our understanding of metabolism, they also hamper synthetic biology and metabolic engineering endeavors and therefore slow down the shift of the chemical industry towards greener, more sustainable biosynthesis processes. To our knowledge, no one has ever attempted to systematically map and fill the knowledge gaps in metabolism at the scale of global biochemical knowledge.
We herein present a comprehensive map of known, well-characterized and novel, predicted reactions between millions of known chemical, biochemical and bioactive compounds. It is the first attempt to systematically map and fill the knowledge gaps in metabolism at the scale of global biochemical reaction networks. Our work is based on the reaction prediction tool BNICE.ch, which condenses available biochemical knowledge about enzymatic reactions into generalized “reaction rules”. We used these rules to integrate almost 2 million molecules from public databases (e.g. ChEMBL, PubChem, HMDB) including plant natural products, pharmaceuticals, and other molecules that are part of or interact with biological systems, into a global biochemical reaction network called ATLASx. Our technology can further propose biosynthetic routes for chemicals designed by synthetic chemistry. We additionally integrated our existing enzyme prediction tool BridgIT1 to propose enzymes that may catalyze the novel, predicted reactions. A network analysis of ATLASx demonstrates how our predictions increase the interconnectivity of metabolic knowledge. To underline the biological relevance of our predictions, we present some practical examples that can be easily reconstructed using the ATLASx online platform. While this work is purely computational, several publications have demonstrated the application of our reaction prediction for the design of de novo synthetic pathways producing diverse benzylisoquinoline alkaloids2, tropane alkaloid derivatives3, and novel one-carbon assimilation pathways4.
Overall, this work contributes to the fundamental understanding of biochemical processes in a global and systematic way, providing the scientific community with a rational estimate of the unexplored possibilities in biochemical research. ATLASx and the associated computational methods are the first collection of resources and tools of such scale and application range for exploring the known and predicted biochemical reaction space, and it should be of a useful and valuable resource for synthetic biology and metabolic engineering.
- Hadadi, N., MohammadiPeyhani, H., Miskovic, L., Seijo, M. & Hatzimanikatis, V. Enzyme annotation for orphan and novel reactions using knowledge of substrate reactive sites. PNAS 116, 7298–7307 (2019).
- Hafner, J., Payne, J., MohammadiPeyhani, H., Hatzimanikatis, V. & Smolke, C. A computational workflow for the expansion of heterologous biosynthetic pathways to natural product derivatives. Nature Communications 12, 1760 (2021).
- Srinivasan, P. & Smolke, C. D. Engineering cellular metabolite transport for biosynthesis of computationally predicted tropane alkaloid derivatives in yeast. Proc Natl Acad Sci USA 118, e2104460118 (2021).
- Yang, X., Yuan, Q., Luo, H., Li, F., Mao, Y., Zhao, X., Du, J., Li, P., Ju, X., Zheng, Y., Chen, Y., Liu, Y., Jiang, H., Yao, Y., Ma, H. & Ma, Y. Systematic design and in vitro validation of novel one-carbon assimilation pathways. Metabolic Engineering 56, 142–153 (2019).
Prof. Dr. Vassily Hatzimankatis
Dr. Vassily Hatzimanikatis is Associate Professor of Chemical Engineering, Chemistry and Bioengineering at Ecole Polytechnique Federale de Lausanne (EPFL), in Lausanne, Switzerland. Dr Hatzimanikatis' research interests are on systems and synthetic biology, with focus on evolution and design of metabolism, integration of omics data, bioenergetics, and biochemical and biophysical kinetics. Dr Hatzimanikatis is a fellow of the American Institute for Medical and Biological Engineering (2010), he was a DuPont Young Professor (2001-2004), and he has also received the Jay Bailey Young Investigator Award in Metabolic Engineering (2000), the ACS Elmer Gaden Award (2011), and the Metabolic Engineering Award from the International Society of Metabolic Engineering (2014). He also serves as associate editor of the journals PLOS Computational Biology and Biotechnology & Bioengineering and, and Senior Editor of Biotechnology Journal.