Backround and Significance
Metabolism and development, each crucial for the life of multicellular organisms, have long been viewed as separate entities. The notion of development-driven modifications of metabolic state, perhaps most prominently exemplified by the shift from glycolysis to oxidative metabolism during stem cell differentiation and back to glycolysis in iPS cells has rekindled interest in metabolism among developmental biologists (for review see Folmes et al., 2012; Shyh-Chang et al., 2013). At the same time the long-held assumption that metabolic enzymes perform the same housekeeping functions in all cells has been challenged (see Shyh- Chang et al., 2013) and physiology and biochemistry need to be considered in their developmental context. The impact of the environment and metabolic pathways on development are considered secondary compared to the deterministic role of the genome; yet metabolism is known to be highly reactive to the environment and plants and animal genomes have evolved to respond and adapt to its fluctuation.

For plants as photoautotrophic organisms that undergo life-long post-embryonic development, the inseparable connection between metabolism and development is well accepted. However, mechanistic insight is still scarce and it has been shown only recently that glucose–TOR signalling controls stem/ progenitor-cell proliferation through inter-organ nutrient coordination (Xiong et al., 2014). In animals, it is by now clear that the metabolic status of the cell modulates the activities of many signaling proteins and transcriptional regulators and examples of correlations between metabolic state shifts preceding changes in a developmental program are accumulating (Chell and Brand, 2010; Folmes et al., 2011). Although many studies suggest that metabolic shifts act as triggers of developmental programs, causality has only been demonstrated in few cases (Carey et al., 2014; Ryall et al., 2015).

With the recent advent of metabolomics and systems approaches, we have witnessed a shift in our appreciation of the role of metabolism in the control of development. Although the examples listed above clearly illustrate the importance of metabolic states for the proper execution of the genetic program underlying it, in many developmental contexts this has not been recognised yet. Expanding the repertoire of contexts in which we study the interactions between development and metabolism represents the first challenge that the field needs to address. The description of the metabolic profile associated with given developmental programs and shifts in cell fate is a necessary first step toward a mechanistic understanding of how metabolic control and developmental programs resonate with each other. In turn, this mechanistic understanding in a range of developmental contexts will permit the identification of general principles.

The second challenge is methodological in nature. The existence of interactions between metabolic and developmental processes has been until now often inferred based on transcriptional approaches. While informative about changes in the metabolic potential of the developing organ, such approaches frequently fail in providing a comprehensive view of metabolic pathways and in identifying candidate pathways and their key reactions that tune the developmental output. The current toolbox for the study of metabolites contains two very different, yet complementary set of methods. On the one hand, the combination of different analytical methods notably involving mass spectrometry detection (GC-MS and LC coupled to different MS detection systems) has become the gold standard for the quantification, and in some cases imaging for instance by MALDI-MS, of a large spectrum of small molecule classes out of population of thousands of pooled cells (Moussaieff et al., 2013). On the other hand, genetically encoded nanosensors are able to provide at high spatial and temporal resolution the dynamics of individual key metabolites. Reconciling these two opposite aspects is an important challenge. An increase in sensitivity of the untargeted analytical methods will allow to use fewer cells and lead to an increase in resolution (both in space & in time). In parallel, increasing the portfolio of available nanosensors for metabolites will boost our ability to report on the dynamics of the flow of these metabolites at the cellular scale.