- Email: [email protected]
Metabolite-Induced Protein Expression Guided by Metabolomics and Systems Biology
Cell Metabolism
Previews response to cellular stress. (3) The nonogt mouse model underlines how important NONO is for metabolic homeostasis. The lean phenotype of this mouse model is paired with glucose intolerance, which is interesting, as most of the animal models of glucose intolerance are based on high-fat-diet-induced obesity or genetic mutations that predispose mice to obesity. The nonogt mouse model should be therefore a valuable tool for future research beyond a circadian context. In accordance with many previous studies, this study demonstrates the crucial influence of temporal coordination between metabolic demand and expression of relevant genes. A key message here is that RNA binding proteins can play essential roles in this context, an important insight, e.g., for the development of novel treatment strategies for metabolic disorders. In summary, to prevent metabolic diseases, like obesity and diabetes, not only what you eat but also when you eat needs to be considered. Thus, going to the fridge at night is not
such a good idea if you want to keep a healthy metabolism. Benegiamo et al. (2018) helped us to understand why.
Koike, N., Yoo, S.H., Huang, H.C., Kumar, V., Lee, C., Kim, T.K., and Takahashi, J.S. (2012). Transcriptional architecture and chromatin landscape of the core circadian clock in mammals. Science 338, 349–354.
REFERENCES
Kowalska, E., Ripperger, J.A., Hoegger, D.C., Bruegger, P., Buch, T., Birchler, T., Mueller, A., Albrecht, U., Contaldo, C., and Brown, S.A. (2013). NONO couples the circadian clock to the cell cycle. Proc. Natl. Acad. Sci. USA 110, 1592–1599.
Bass, J., and Takahashi, J.S. (2010). Circadian integration of metabolism and energetics. Science 330, 1349–1354. Benegiamo, G., Mure, L.S., Erikson, G., Le, H.D., Moriggi, E., Brown, S.A., and Panda, S. (2018). The RNA-binding protein NONO coordinates hepatic adaptation to feeding. Cell Metab. 27, this issue, 404–418.
Prasanth, K.V., Prasanth, S.G., Xuan, Z., Hearn, S., Freier, S.M., Bennett, C.F., Zhang, M.Q., and Spector, D.L. (2005). Regulating gene expression through RNA nuclear retention. Cell 123, 249–263.
Brown, S.A., Ripperger, J., Kadener, S., FleuryOlela, F., Vilbois, F., Rosbash, M., and Schibler, U. (2005). PERIOD1-associated proteins modulate the negative limb of the mammalian circadian oscillator. Science 308, 693–696.
Scheer, F.A., Hilton, M.F., Mantzoros, C.S., and Shea, S.A. (2009). Adverse metabolic and cardiovascular consequences of circadian misalignment. Proc. Natl. Acad. Sci. USA 106, 4453–4458.
Damiola, F., Le Minh, N., Preitner, N., Kornmann, B., Fleury-Olela, F., and Schibler, U. (2000). Restricted feeding uncouples circadian oscillators in peripheral tissues from the central pacemaker in the suprachiasmatic nucleus. Genes Dev. 14, 2950–2961.
Torres, M., Becquet, D., Blanchard, M.P., Guillen, S., Boyer, B., Moreno, M., Franc, J.L., and Franc¸ois-Bellan, A.M. (2016). Circadian RNA expression elicited by 30 -UTR IRAlu-paraspeckle associated elements. eLife 5, e14837.
Knott, G.J., Bond, C.S., and Fox, A.H. (2016). The DBHS proteins SFPQ, NONO and PSPC1: a multipurpose molecular scaffold. Nucleic Acids Res. 44, 3989–4004.
Uren, P.J., Bahrami-Samani, E., Burns, S.C., Qiao, M., Karginov, F.V., Hodges, E., Hannon, G.J., Sanford, J.R., Penalva, L.O., and Smith, A.D. (2012). Site identification in high-throughput RNA-protein interaction data. Bioinformatics 28, 3013–3020.
Metabolite-Induced Protein Expression Guided by Metabolomics and Systems Biology Martin Giera,1,* Filipe Branco dos Santos,2 and Gary Siuzdak3,* 1Leiden
University Medical Center, Center for Proteomics and Metabolomics, Albinusdreef 2, 2300RC Leiden, the Netherlands Microbial Physiology Group, Faculty of Science, Swammerdam Institute for Life Sciences, University of Amsterdam, Science Park 904, 1098 XH Amsterdam, the Netherlands 3The Scripps Research Institute, Scripps Center for Metabolomics and Mass Spectrometry, La Jolla, CA, USA *Correspondence: [email protected] (M.G.), [email protected] (G.S.) https://doi.org/10.1016/j.cmet.2018.01.002 2Molecular
Metabolomics combined with systems biology can be used to identify endogenous metabolites that modulate protein expression. Recent examples include the 2-fold enhancements of pertussis toxin protein in vaccine production and myelin basic protein expression in oligodendrocyte maturation; both applied a metabolomics-systems strategy to identify active metabolites. Endogenous metabolites play essential roles in the regulation of cellular activity through their interactions with genes, transcripts, and proteins; however, our ability to identify active metabolites to modulate these interactions is relatively limited. Two recent examples represent notable exceptions in which metabolo-
mics and systems biology were combined to enhance protein expression: the first study identified metabolites that enhance the commercial production of a pertussis vaccine (Branco dos Santos et al., 2017), and the second identified a single metabolite that modulates an active protein for multiple sclerosis
270 Cell Metabolism 27, February 6, 2018 ª 2018 Elsevier Inc.
neuron remyelination (Beyer et al., 2018). Protein expression is intimately linked with the genome and the genetic information transmitted downstream; however, the role metabolomics and metabolites can play as a driving force in these upstream events is unappreciated (Figure 1, left). For example, the
Cell Metabolism
Previews
Figure 1. Metabolite-Induced Protein Expression (MIPE) An overview of metabolite-induced protein expression (left) as guided by systems biology and metabolomics with a schematic representation of the discussed examples: an increase of pertussis toxin production (right upper panel) was achieved using a combination of systems biology modeling and metabolomics. Lower panel presents how metabolomics combined with pathway analysis helped identify taurine and its ability to impact myelin basic protein expression, an important protein in multiple sclerosis neuron regeneration.
post-genome era is characterized by the rise of transcriptomics, proteomics, and metabolomics, all designed to elucidate the composition and, ultimately, the function behind their information content. In this context, the central dogma of biology (Crick, 1970) is that genomeencoded information is being transmitted downward to the transcriptome and proteome, leading to changes within the metabolic pool, the metabolome. Therefore, it is not surprising that the metabolome has been primarily linked with phenotypic changes while the transcriptome and proteome are seen as control centers. This downstream concept places metabolomics at the forefront of diagnostics and biomarker discovery, however, minimizing metabolites and their more significant role as initiators of bioactivity. Integrating metabolomics within a systems biology framework facilitates the elucidation of metabolites’ upstream impact and assists in the identification of biologically active metabolites that can promote protein expression and activity. To date, systems biology has not yet been effectively combined with metabolomics to identify metabolites that can modulate phenotype; a particularly interesting application is protein expression.
The recent study by Branco dos Santos et al. (2017) used a genome-wide metabolic model of the human pathogen Bordetella pertussis to increase pertussis vaccine production (Figure 1, upper right). Initially, the systems biologybased model of the pathogen revealed an apparent gap in the theoretical versus experimentally observed nitrogen metabolism. Constraint-based analysis techniques allowed a comprehensive accounting of mass flow in the organism, and a model for Bordetella pertussis was constructed and validated, revealing a gap in the experimentally determined nitrogen output. Characterization of theoretical optimal flux distributions through the metabolic network was used to predict all possible nitrogen sinks compatible with the data; this list was subsequently used as a filter for untargeted metabolomics analysis of the growth media. By performing an iterative cycle in which metabolomics was used to help decipher the mass balance of the simulations, the authors could improve/ curate the original computer model. In this way, quite unexpectedly, nucleobases and nucleosides were identified as substantial novel sinks of nitrogen, at concentrations ranging up to the millimolar level.
In addition to identifying the relevant nitrogen metabolite sinks, the models were also used to investigate sulfur balance. In the particular case of pertussis toxin (PT), it was known that sulfate inhibits PT production, and that sulfate accumulated in the traditional growth medium through excess methionine and glutathione (in the medium). The metabolic map provided an alternative pathway that was previously unknown for inorganic sulfur assimilation via thiosulphate; thus, using thiosulphate as an alternative, together with a careful balancing of amino acid requirements based on the metabolomics analysis of biomass composition, resulted in balanced growth conditions. Overall, metabolomics analysis of the fermentations, combined with a systems biology approach, led to a 2.4-fold increase in PT production from a simpler medium. In a second example, metabolomics was employed in combination with pathway analysis to identify metabolites that modulate the process of oligodendrocyte differentiation and/or maturation, an important regenerative process associated with demyelinating diseases (Beyer et al., 2018). Mass spectrometry-based metabolomics was used Cell Metabolism 27, February 6, 2018 271
Cell Metabolism
Previews to investigate the mechanism, and whether endogenous metabolites could impact oligodendrocyte precursor cell (OPC) differentiation. Taurine and creatine pathways were found to be the most highly altered events associated with OPC differentiation, where taurine and hypotaurine increased by over 20and 10-fold, respectively, and metabolites in the creatine pathway were also upregulated by at least an order of magnitude. Quantitative analysis of associated metabolites was performed on both the upstream and downstream parts of the taurine and creatine pathways. When added exogenously at physiologically relevant concentrations, taurine was found to dramatically enhance drug-induced OPC differentiation and facilitate the in vitro myelination of co-cultured axons. Mechanistically, taurine-induced OPC differentiationand myelination-enhancing activities appeared to be driven by taurine’s ability to increase serine levels, which is an initial building block required for the synthesis of the glycosphingolipid components of myelin. A key outcome in the taurine-induced differentiation process was the increased generation of myelin basic protein (Figure 1, lower right) by 2.5-fold, a protein directly associated with the myelination of neurons. The primary motivation for integrating metabolomics and systems biology has been in the generation of more comprehensive metabolic maps focusing primarily on altering metabolite levels; this coincides with recent reports indicating that our knowledge about metabolic biological activity is rapidly increasing (Husted et al., 2017). A first step in the direction of providing metabolic activity has recently been described
272 Cell Metabolism 27, February 6, 2018
and incorporated into XCMS Online, where metabolite selection is made on multiple selection criteria including statistical significance, pathway analysis, and multi-omics screening. However, as we are in the early stages a more sophisticated level of metabolite selection will likely evolve. Current systems biology models concentrate on production of metabolites and flux with criteria that have yet to be fully understood, including how proteins are biosynthesized, their folding requirements, and what the production inhibitory effects are. Ultimately, metabolomics-guided systems biologybased models will provide the necessary framework for the integration of all aforementioned approaches, in that it should allow predicting metabolic requirements for the production of key active metabolites. In turn, this will allow us to perturb metabolism and concurrently shape the proteomic landscape to our needs. The examples described above underscore the utility of integrating metabolomics and systems biology to decipher and advance metabolite-induced protein expression. Metabolites are a logical source to enhance and understand these biological systems, especially as their bioactivities are becoming more apparent (C. Guijas et al., unpublished data), as is the systems-wide role they are playing (Huan et al., 2017). A significant challenge will be in deciphering which metabolites are best suited to alter the system, a challenge that will likely be solved by combining seemingly disparate information, including statistical analysis, pathway data, and other ‘omic technologies, and integrating this knowledge into existing databases. This is a task well suited for machine learning-type approaches, once enough data are available to provide
meaningful information (Sastry et al., 2017). For example, combining genomewide metabolic models with experimental metabolomics data will likely lead to improved models and the identification of metabolites with unique activities. Most intriguing to us is that unlike other ‘omic approaches, the beauty of metabolomics and activity testing is that metabolites are readily commercially accessible and generally inexpensive, and can directly impact the system quickly. This is also intriguing because as technologists, we are no longer just passive observers but instead active participants. REFERENCES Beyer, B.A., Fang, M., Sadrian, B., MontenegroBurke, J.R., Plaisted, W.C., Kok, B.P.C., Saez, E., Kondo, T., Siuzdak, G., and Lairson, L.L. (2018). Metabolomics-based discovery of a metabolite that enhances oligodendrocyte maturation. Nat. Chem. Biol. 14, 22–28. Branco dos Santos, F., Olivier, B.G., Boele, J., Smessaert, V., De Rop, P., Krumpochova, P., Klau, G.W., Giera, M., Dehottay, P., Teusink, B., and Goffin, P. (2017). Probing the genome-scale metabolic landscape of Bordetella pertussis, the causative agent of whooping cough. Appl. Environ. Microbiol. Published online August 25, 2017. https://doi.org/10.1128/AEM.01528-17. Crick, F. (1970). Central dogma of molecular biology. Nature 227, 561–563. Huan, T., Forsberg, E.M., Rinehart, D., Johnson, C.H., Ivanisevic, J., Benton, H.P., Fang, M., Aisporna, A., Hilmers, B., Poole, F.L., et al. (2017). Systems biology guided by XCMS Online metabolomics. Nat. Methods 14, 461–462. Husted, A.S., Trauelsen, M., Rudenko, O., Hjorth, S.A., and Schwartz, T.W. (2017). GPCRmediated signaling of metabolites. Cell Metab. 25, 777–796. Sastry, A., Monk, J., Tegel, H., Uhlen, M., Palsson, B.O., Rockberg, J., and Brunk, E. (2017). Machine learning in computational biology to accelerate high-throughput protein expression. Bioinformatics 33, 2487–2495.
Previews response to cellular stress. (3) The nonogt mouse model underlines how important NONO is for metabolic homeostasis. The lean phenotype of this mouse model is paired with glucose intolerance, which is interesting, as most of the animal models of glucose intolerance are based on high-fat-diet-induced obesity or genetic mutations that predispose mice to obesity. The nonogt mouse model should be therefore a valuable tool for future research beyond a circadian context. In accordance with many previous studies, this study demonstrates the crucial influence of temporal coordination between metabolic demand and expression of relevant genes. A key message here is that RNA binding proteins can play essential roles in this context, an important insight, e.g., for the development of novel treatment strategies for metabolic disorders. In summary, to prevent metabolic diseases, like obesity and diabetes, not only what you eat but also when you eat needs to be considered. Thus, going to the fridge at night is not
such a good idea if you want to keep a healthy metabolism. Benegiamo et al. (2018) helped us to understand why.
Koike, N., Yoo, S.H., Huang, H.C., Kumar, V., Lee, C., Kim, T.K., and Takahashi, J.S. (2012). Transcriptional architecture and chromatin landscape of the core circadian clock in mammals. Science 338, 349–354.
REFERENCES
Kowalska, E., Ripperger, J.A., Hoegger, D.C., Bruegger, P., Buch, T., Birchler, T., Mueller, A., Albrecht, U., Contaldo, C., and Brown, S.A. (2013). NONO couples the circadian clock to the cell cycle. Proc. Natl. Acad. Sci. USA 110, 1592–1599.
Bass, J., and Takahashi, J.S. (2010). Circadian integration of metabolism and energetics. Science 330, 1349–1354. Benegiamo, G., Mure, L.S., Erikson, G., Le, H.D., Moriggi, E., Brown, S.A., and Panda, S. (2018). The RNA-binding protein NONO coordinates hepatic adaptation to feeding. Cell Metab. 27, this issue, 404–418.
Prasanth, K.V., Prasanth, S.G., Xuan, Z., Hearn, S., Freier, S.M., Bennett, C.F., Zhang, M.Q., and Spector, D.L. (2005). Regulating gene expression through RNA nuclear retention. Cell 123, 249–263.
Brown, S.A., Ripperger, J., Kadener, S., FleuryOlela, F., Vilbois, F., Rosbash, M., and Schibler, U. (2005). PERIOD1-associated proteins modulate the negative limb of the mammalian circadian oscillator. Science 308, 693–696.
Scheer, F.A., Hilton, M.F., Mantzoros, C.S., and Shea, S.A. (2009). Adverse metabolic and cardiovascular consequences of circadian misalignment. Proc. Natl. Acad. Sci. USA 106, 4453–4458.
Damiola, F., Le Minh, N., Preitner, N., Kornmann, B., Fleury-Olela, F., and Schibler, U. (2000). Restricted feeding uncouples circadian oscillators in peripheral tissues from the central pacemaker in the suprachiasmatic nucleus. Genes Dev. 14, 2950–2961.
Torres, M., Becquet, D., Blanchard, M.P., Guillen, S., Boyer, B., Moreno, M., Franc, J.L., and Franc¸ois-Bellan, A.M. (2016). Circadian RNA expression elicited by 30 -UTR IRAlu-paraspeckle associated elements. eLife 5, e14837.
Knott, G.J., Bond, C.S., and Fox, A.H. (2016). The DBHS proteins SFPQ, NONO and PSPC1: a multipurpose molecular scaffold. Nucleic Acids Res. 44, 3989–4004.
Uren, P.J., Bahrami-Samani, E., Burns, S.C., Qiao, M., Karginov, F.V., Hodges, E., Hannon, G.J., Sanford, J.R., Penalva, L.O., and Smith, A.D. (2012). Site identification in high-throughput RNA-protein interaction data. Bioinformatics 28, 3013–3020.
Metabolite-Induced Protein Expression Guided by Metabolomics and Systems Biology Martin Giera,1,* Filipe Branco dos Santos,2 and Gary Siuzdak3,* 1Leiden
University Medical Center, Center for Proteomics and Metabolomics, Albinusdreef 2, 2300RC Leiden, the Netherlands Microbial Physiology Group, Faculty of Science, Swammerdam Institute for Life Sciences, University of Amsterdam, Science Park 904, 1098 XH Amsterdam, the Netherlands 3The Scripps Research Institute, Scripps Center for Metabolomics and Mass Spectrometry, La Jolla, CA, USA *Correspondence: [email protected] (M.G.), [email protected] (G.S.) https://doi.org/10.1016/j.cmet.2018.01.002 2Molecular
Metabolomics combined with systems biology can be used to identify endogenous metabolites that modulate protein expression. Recent examples include the 2-fold enhancements of pertussis toxin protein in vaccine production and myelin basic protein expression in oligodendrocyte maturation; both applied a metabolomics-systems strategy to identify active metabolites. Endogenous metabolites play essential roles in the regulation of cellular activity through their interactions with genes, transcripts, and proteins; however, our ability to identify active metabolites to modulate these interactions is relatively limited. Two recent examples represent notable exceptions in which metabolo-
mics and systems biology were combined to enhance protein expression: the first study identified metabolites that enhance the commercial production of a pertussis vaccine (Branco dos Santos et al., 2017), and the second identified a single metabolite that modulates an active protein for multiple sclerosis
270 Cell Metabolism 27, February 6, 2018 ª 2018 Elsevier Inc.
neuron remyelination (Beyer et al., 2018). Protein expression is intimately linked with the genome and the genetic information transmitted downstream; however, the role metabolomics and metabolites can play as a driving force in these upstream events is unappreciated (Figure 1, left). For example, the
Cell Metabolism
Previews
Figure 1. Metabolite-Induced Protein Expression (MIPE) An overview of metabolite-induced protein expression (left) as guided by systems biology and metabolomics with a schematic representation of the discussed examples: an increase of pertussis toxin production (right upper panel) was achieved using a combination of systems biology modeling and metabolomics. Lower panel presents how metabolomics combined with pathway analysis helped identify taurine and its ability to impact myelin basic protein expression, an important protein in multiple sclerosis neuron regeneration.
post-genome era is characterized by the rise of transcriptomics, proteomics, and metabolomics, all designed to elucidate the composition and, ultimately, the function behind their information content. In this context, the central dogma of biology (Crick, 1970) is that genomeencoded information is being transmitted downward to the transcriptome and proteome, leading to changes within the metabolic pool, the metabolome. Therefore, it is not surprising that the metabolome has been primarily linked with phenotypic changes while the transcriptome and proteome are seen as control centers. This downstream concept places metabolomics at the forefront of diagnostics and biomarker discovery, however, minimizing metabolites and their more significant role as initiators of bioactivity. Integrating metabolomics within a systems biology framework facilitates the elucidation of metabolites’ upstream impact and assists in the identification of biologically active metabolites that can promote protein expression and activity. To date, systems biology has not yet been effectively combined with metabolomics to identify metabolites that can modulate phenotype; a particularly interesting application is protein expression.
The recent study by Branco dos Santos et al. (2017) used a genome-wide metabolic model of the human pathogen Bordetella pertussis to increase pertussis vaccine production (Figure 1, upper right). Initially, the systems biologybased model of the pathogen revealed an apparent gap in the theoretical versus experimentally observed nitrogen metabolism. Constraint-based analysis techniques allowed a comprehensive accounting of mass flow in the organism, and a model for Bordetella pertussis was constructed and validated, revealing a gap in the experimentally determined nitrogen output. Characterization of theoretical optimal flux distributions through the metabolic network was used to predict all possible nitrogen sinks compatible with the data; this list was subsequently used as a filter for untargeted metabolomics analysis of the growth media. By performing an iterative cycle in which metabolomics was used to help decipher the mass balance of the simulations, the authors could improve/ curate the original computer model. In this way, quite unexpectedly, nucleobases and nucleosides were identified as substantial novel sinks of nitrogen, at concentrations ranging up to the millimolar level.
In addition to identifying the relevant nitrogen metabolite sinks, the models were also used to investigate sulfur balance. In the particular case of pertussis toxin (PT), it was known that sulfate inhibits PT production, and that sulfate accumulated in the traditional growth medium through excess methionine and glutathione (in the medium). The metabolic map provided an alternative pathway that was previously unknown for inorganic sulfur assimilation via thiosulphate; thus, using thiosulphate as an alternative, together with a careful balancing of amino acid requirements based on the metabolomics analysis of biomass composition, resulted in balanced growth conditions. Overall, metabolomics analysis of the fermentations, combined with a systems biology approach, led to a 2.4-fold increase in PT production from a simpler medium. In a second example, metabolomics was employed in combination with pathway analysis to identify metabolites that modulate the process of oligodendrocyte differentiation and/or maturation, an important regenerative process associated with demyelinating diseases (Beyer et al., 2018). Mass spectrometry-based metabolomics was used Cell Metabolism 27, February 6, 2018 271
Cell Metabolism
Previews to investigate the mechanism, and whether endogenous metabolites could impact oligodendrocyte precursor cell (OPC) differentiation. Taurine and creatine pathways were found to be the most highly altered events associated with OPC differentiation, where taurine and hypotaurine increased by over 20and 10-fold, respectively, and metabolites in the creatine pathway were also upregulated by at least an order of magnitude. Quantitative analysis of associated metabolites was performed on both the upstream and downstream parts of the taurine and creatine pathways. When added exogenously at physiologically relevant concentrations, taurine was found to dramatically enhance drug-induced OPC differentiation and facilitate the in vitro myelination of co-cultured axons. Mechanistically, taurine-induced OPC differentiationand myelination-enhancing activities appeared to be driven by taurine’s ability to increase serine levels, which is an initial building block required for the synthesis of the glycosphingolipid components of myelin. A key outcome in the taurine-induced differentiation process was the increased generation of myelin basic protein (Figure 1, lower right) by 2.5-fold, a protein directly associated with the myelination of neurons. The primary motivation for integrating metabolomics and systems biology has been in the generation of more comprehensive metabolic maps focusing primarily on altering metabolite levels; this coincides with recent reports indicating that our knowledge about metabolic biological activity is rapidly increasing (Husted et al., 2017). A first step in the direction of providing metabolic activity has recently been described
272 Cell Metabolism 27, February 6, 2018
and incorporated into XCMS Online, where metabolite selection is made on multiple selection criteria including statistical significance, pathway analysis, and multi-omics screening. However, as we are in the early stages a more sophisticated level of metabolite selection will likely evolve. Current systems biology models concentrate on production of metabolites and flux with criteria that have yet to be fully understood, including how proteins are biosynthesized, their folding requirements, and what the production inhibitory effects are. Ultimately, metabolomics-guided systems biologybased models will provide the necessary framework for the integration of all aforementioned approaches, in that it should allow predicting metabolic requirements for the production of key active metabolites. In turn, this will allow us to perturb metabolism and concurrently shape the proteomic landscape to our needs. The examples described above underscore the utility of integrating metabolomics and systems biology to decipher and advance metabolite-induced protein expression. Metabolites are a logical source to enhance and understand these biological systems, especially as their bioactivities are becoming more apparent (C. Guijas et al., unpublished data), as is the systems-wide role they are playing (Huan et al., 2017). A significant challenge will be in deciphering which metabolites are best suited to alter the system, a challenge that will likely be solved by combining seemingly disparate information, including statistical analysis, pathway data, and other ‘omic technologies, and integrating this knowledge into existing databases. This is a task well suited for machine learning-type approaches, once enough data are available to provide
meaningful information (Sastry et al., 2017). For example, combining genomewide metabolic models with experimental metabolomics data will likely lead to improved models and the identification of metabolites with unique activities. Most intriguing to us is that unlike other ‘omic approaches, the beauty of metabolomics and activity testing is that metabolites are readily commercially accessible and generally inexpensive, and can directly impact the system quickly. This is also intriguing because as technologists, we are no longer just passive observers but instead active participants. REFERENCES Beyer, B.A., Fang, M., Sadrian, B., MontenegroBurke, J.R., Plaisted, W.C., Kok, B.P.C., Saez, E., Kondo, T., Siuzdak, G., and Lairson, L.L. (2018). Metabolomics-based discovery of a metabolite that enhances oligodendrocyte maturation. Nat. Chem. Biol. 14, 22–28. Branco dos Santos, F., Olivier, B.G., Boele, J., Smessaert, V., De Rop, P., Krumpochova, P., Klau, G.W., Giera, M., Dehottay, P., Teusink, B., and Goffin, P. (2017). Probing the genome-scale metabolic landscape of Bordetella pertussis, the causative agent of whooping cough. Appl. Environ. Microbiol. Published online August 25, 2017. https://doi.org/10.1128/AEM.01528-17. Crick, F. (1970). Central dogma of molecular biology. Nature 227, 561–563. Huan, T., Forsberg, E.M., Rinehart, D., Johnson, C.H., Ivanisevic, J., Benton, H.P., Fang, M., Aisporna, A., Hilmers, B., Poole, F.L., et al. (2017). Systems biology guided by XCMS Online metabolomics. Nat. Methods 14, 461–462. Husted, A.S., Trauelsen, M., Rudenko, O., Hjorth, S.A., and Schwartz, T.W. (2017). GPCRmediated signaling of metabolites. Cell Metab. 25, 777–796. Sastry, A., Monk, J., Tegel, H., Uhlen, M., Palsson, B.O., Rockberg, J., and Brunk, E. (2017). Machine learning in computational biology to accelerate high-throughput protein expression. Bioinformatics 33, 2487–2495.