No molecule in a living organism exists in a vacuum. Indeed, each interacts with thousands of other molecules. In any given organism, native interactions have been harnessed to achieve cellular objectives. Similarly, other interactions have been avoided. Thus, through evolution, genes and proteins have evolved to increase the chance that a given living system will be perpetuated (Lewis, et al. MSB, 2010). That is, each component has evolved in the context of all other components in the cell or organism and in the environment (Nam*, Lewis*, Science, 2012). Furthermore, their true function occurs in this complex cellular context, and this function may or may not be seen in vitro. While enzyme functions and biomolecular interactions have classically been studied in vitro or in artificial systems (e.g., enzyme assays or yeast 2-hybrid screens), technological and theoretical advances have recently arisen to elucidate the biochemistry of many individual enzymes in parallel in the context of the complex environment of the cell. Over the coming years, I aim to develop and apply systems analyses (Lewis, et al. Nat Rev Microb 2012) and multiplexed high throughput screening technologies for a few key applications to accelerate the characterization of individual proteins, improve the success of synthetic biology and metabolic engineering designs, and provide deep fundamental insights into the molecular basis of life. First, approaches will be developed and applied to characterize the functions of enzyme post-translational modifications and other enzyme-level properties. Second, systems modeling and genome editing will be harnessed to modulate and study complex post-translational modifications, such as glycosylation. Third, computational models of metabolism and protein secretion will be constructed and used for predictive modeling and omics data analysis to drive improvements in biotherapeutic development. A recent podcast describes our work on CHO cells.
Systems Glycobiology and Protein Secretion
The protein secretion pathway is essential to eukaryotic function and drug development. Through this pathway a diverse range of molecules are processed and released into the extracellular space. This process ensures that proteins are translated, folded, subjected to quality control, modified to enhance their function, and ultimately released. Such proteins include components the extracellular matrix which keeps tissues and organs together. Many other proteins such as homone signals and proteases are released by cells through this pathway. Alterations in protein secretion often accompany disease, such as cancer where changes in the protein secretion pathway influences tumor growth and enables metastasis. In our group we are taking a systems biology approach to study protein secretion in order to understand how changes in this system influences cell physiology. Using whole-genome sequencing (Xu et al., Nat Biotech, 2011; Lewis, et al., Nat Biotech, 2013; Robasky*,Lewis* and Church, Nat Rev Genetics, 2014), transcriptomics, proteomics (Baycin, et al., J Prot Res, 2012), metabolomics, and many other tools, we are harnessing our analyses to study disease and to guide synthetic biology designs through which we engineer the pathway to enhance the quantity and quality of protein-based drug production (Spahn and Lewis, Curr Opin Biotech, 2014). For more on these efforts and their application to drug production in Chinese hamster ovary cells, check out these podcasts: The Chinese hamsters that helped birth biotech and CHO Cells and Computational Models
Proteins constitute most of the molecular machinery that drives the various functions of a cell. Proteins are frequently modified with small chemical moieties to control their function. However, since thousands of different proteins can exist in a single cell at any given moment, it is not immediately clear how one post-translational modification (PTM) affects the entire cell. Furthermore, it is often unclear how diversity in complex PTMs, such as glycosylation, is controlled and how we could change glycosylation for therapeutic purposes or drug development. To approach these questions, we utilize systems biology models, in conjunction with protein structure models to study how a PTM on one protein can impact all other molecules in a cell and therefore influence phenotype (Spahn, et al. Metabolic Engineering, 2016).
Genome-scale modeling of metabolism
Through evolution, life has emerged based on the ability of certain molecules to react and extract energy from its environment. This energy can stem from light or the vast repertoire of chemicals in nature. Metabolism is the process through which a cell extracts energy from its surroundings, synthesizes more cell parts, and modifies its environment to improve fitness. A cell’s metabolism can include thousands of chemical reactions, each of which are linked together in pathways. Ultimately, metabolism can take simple molecules from nature, such as the sugar glucose, and turn it into all of its cell parts and the energy it needs to function and grow. In our group we leverage diverse data types to build genome-scale metabolic network models, which are computational models that contain all known chemical reactions in a cell. These models can be employed for many applications, such as studying evolution (Nam*, Lewis*, Science, 2012), developing cell factories (Hefzi and Lewis, 2014), or studying disease (Lewis, Nat Biotech, 2010; Kumar, et al., PLoS One, 2014). These models allow us to understand how all of the components in a cell contribute to an entire cell’s function.
Systems biology of neurological diseases
While some systems like metabolism have been heavily studied for decades, many biological processes remain poorly characterized, and so the molecular etiology of complex pathologies, such as neurological disorders, remains unclear. Fortunately, top-down statistical and network analysis approaches are helping to remedy those challenges. Numerous genome-wide studies are identifying genomic and epigenetic factors that correlate with neurological disorders, such as autism spectrum disorders (ASD) and Alzheimer’s disease, and network analysis approaches being used to study their genetic basis. Using clinical metrics, blood transcriptomes, and brain imaging data, we aim to use systems biology approaches (e.g., Lewis, Nat Biotech, 2010, Busskamp*, Lewis*, Mol Syst Bio, 2014), coupled with stem cell biology (Busskamp*, Lewis*, Mol Syst Bio, 2014)) and genome editing techniques to gain a mechanistic understanding of processes underlying neurological diseases.