Nutritional control of developmental processes

For most animals, diet is the source of all nutrients and, ultimately, of all metabolites that sustain their growth, development and homeostasis. Although not all metabolic intermediates originate directly from diet, they are still synthesized by metabolic pathways from available nutrients. During development, metabolic networks are dynamically regulated to meet energy demands and supply the building blocks required for navigating life history transitions. These metabolic pathways actively respond to organismal nutritional status, enabling animals to cope with dietary fluctuations. The study of animal development must, therefore, involve understanding both the role of nutritional input and the organism's metabolic response in shaping developmental processes.

The primary purpose of nutrition is to fuel energy metabolism, which functions as an important permissive cue for developmental progression. When food is scarce, organisms suppress costly developmental processes in favor of survival. Cell proliferation, growth, mitochondrial activity and resource allocation are tightly regulated by signaling systems that integrate nutritional status with other cues, such as growth factors. Some of the best-studied examples of such regulators include mTOR, AMPK and metabolic FGF signaling (reviewed by Li, 2019; Valvezan and Manning, 2019; Hardie et al., 2016; Mantovani and Roy, 2011). The recent growth of metabolic studies in a developmental context has also brought increasing evidence that metabolic signals can instruct biological processes (Tu et al., 2023). Rather than merely running in the background, metabolism can shape developmental trajectories by participating in cellular signaling. Metabolites acting as direct signals or substrates for post-translational modifications, and ‘moonlighting’ metabolic enzymes have all been implicated in mediating the crosstalk between metabolism and developmental signaling (Baker and Rutter, 2023; Miyazawa et al., 2022; Evers et al., 2021; Boon et al., 2020; Miyazawa and Aulehla, 2018).

Ultimately, in the context of development or regeneration, the layers of permissive and instructive regulation must act in concert to ensure the correct deployment of morphogenetic processes. To achieve this goal, cells, tissues and organs cooperate across multiple temporal and spatial scales. Our understanding of these mechanisms is still rudimentary. In this Spotlight, we review the principles that underlie nutritional control of development, traversing organizational levels from cells to whole organisms. We highlight open questions and provide a perspective on how relatively simple multicellular systems can help us bridge the knowledge gap between nutrition, metabolism and development.

Depending on their function, different cell types have different metabolic demands and programs. White adipocytes, for example, specialize in storing nutrients, whereas highly catabolic muscle cells actively use nutrients to fuel contractions. Changes in cell state can also influence metabolic status, as demonstrated by the classic work of Otto Warburg on tumor metabolism (Warburg, 1925). It has since become clear that the hallmarks of rapidly proliferating cells, even in physiological conditions, are highly active glycolysis (irrespective of oxygen availability) and the use of tricarboxylic acid cycle intermediates for biosynthetic functions (reviewed by Vander Heiden et al., 2009), but can metabolic programs and nutrient signals impact cellular behaviors? A prime platform for this regulation is the intestine, where constant epithelial remodeling occurs under direct exposure to nutrients from ingested food. Here, food availability is a key driver of intestinal tissue plasticity across animal phyla (Piersma and Drent, 2003). Food-dependent regulation of intestinal plasticity is most prominent in hibernating animals (Hume et al., 2002), and infrequent feeders, such as certain snakes (Secor and Diamond, 1998), where the intestinal mucosa atrophies during periods of fasting and is quickly replenished upon re-feeding via cell proliferation and rearrangements (Starck, 2003). However, it also occurs to some extent even in frequent feeders, such as humans, where its defects may underlie the inability to absorb sufficient amounts of nutrients after intestinal resections (Drozdowski and Thomson, 2006). In Drosophila, intestinal plasticity is controlled by S-adenosylmethionine (SAM), among other signals (Fig. 1A). SAM is synthesized from dietary methionine and promotes stem cell division while suppressing cytokine signaling from enterocytes to stem cells. Under starvation, this repression is lifted and seems to prime the stem cells (along with cell-autonomous signals) for rapid proliferation after re-feeding (Obata et al., 2018). Similar regulation operates in other somatic stem cells, where self-renewal and proliferation capacities are controlled by different metabolites (reviewed by Meacham et al., 2022).

In pluripotent stem cells, nutrients influence not only self-renewal but also differentiation trajectories (reviewed by Lu et al., 2021). Withdrawing certain nutrients in cultured embryos or embryo-like models can either impact the proportions of differentiated cells or completely prevent the development of certain cell fates (e.g. Song et al., 2019; Chen et al., 2018). These varied outcomes are likely tightly linked to the underlying changes in metabolic programs during embryonic cell differentiation. For example, the two earliest distinct cell populations in the mouse embryo, trophectoderm and inner cell mass, execute distinct metabolic programs. Whereas establishing the trophectoderm fate requires glucose to be metabolized through the hexosamine and pentose phosphate pathways, glucose metabolism is dispensable for inner cell mass establishment (Chi et al., 2020). These two cell populations would thus react differently to glucose withdrawal. It is tempting to speculate that maternal control of nutrient amounts could, for example, modulate developmental timing in this way. Interestingly, concentrations of several nutrients differ markedly in the human oviductal versus uterine fluids (Gardner and Leese, 1999), leading some authors to propose that nutrients exert regulatory effects on the pre-implantation embryo as it travels to the uterus (Gardner et al., 2000). In addition, maternally derived metabolites could serve as indicators of external environmental conditions and may influence differentiation outcomes to prime the offspring for anticipated stress. Conversely, similar mechanisms may underlie pathologies in the offspring of mothers with unbalanced diets or suffering from metabolic diseases (Nakano et al., 2017; Moley et al., 1998).

Metabolic heterogeneity of tissues and organs also presents a challenge for nutrient allocation to satisfy various demands based on availability. Ultimately, organismal survival must be balanced with other functions, such as growth, reproduction and developmental progression. This prioritization requires the demands of different body parts to be integrated and their activities modulated based on the global nutritional status (for a conceptual framework on regulating resource allocation, see Ye and Medzhitov, 2019). One of the most remarkable examples of allocation control in action is cancer-associated organ wasting, in which a tumor drains resources from the organismal nutrient pool. In both fly and mammalian models of cancer-associated organ wasting, the brain is sustained for the longest, whereas other organs, such as muscles and adipose tissue, degenerate as a result of nutrient deprivation (Figueroa-Clarevega and Bilder, 2015; Kwon et al., 2015).

Nutritional status also controls the development and morphogenesis of new organs. A key example is the nutrient-dependent plasticity of the Drosophila tracheal system (Linneweber et al., 2014). Under food restriction, nutrient-responsive neurons shape the growth of specific tracheal subsets through the local and systemic release of insulin- and Pdf-like neuropeptides. This plasticity of tracheal architecture drives downstream metabolic changes, enabling flies to cope with poor nutritional conditions. Another striking example is the nutritional control of tentacle addition in the sea anemone Nematostella (Ikmi et al., 2020). The position of the future tentacles is already determined during embryonic development by clusters of cells expressing FGF receptor B (fgfrb), but post-embryonic tentacle development is permitted only when food intake is sufficient (Fig. 1B). The initiation of tentacle development is mediated by crosstalk between FGF and mTOR signaling that results in a local hyperactivation of mTOR around the fgfrb⁺ clusters, which exceeds the global activation level upon feeding.

Similar, local differences of mTOR, insulin and other metabolic signaling activities in response to nutrition also underlie body and organ scaling in other species (Casasa and Moczek, 2018; Koyama et al., 2013). For example, the adult body size of the Onthophagus horned dung beetles increases with higher food availability experienced during larval development. However, different organs do not scale with the same ratios. Whereas the size of genitalia seems to be largely constant, and wings grow proportionally to the body, the male ‘horns’ exhibit allometric hypertrophy – they grow much more than the rest of the body (House and Simmons, 2007; Emlen, 1994). These horns are used as weapons in competition for females and, as a costly trait, likely also function as a signal of partner quality. Wings, by contrast, must always be sized proportionally to the body for proper function (Emlen et al., 2006).

Regeneration of a missing body part is another process for which energetic costs must be balanced with the requirements for whole-body homeostasis. This balance might be mediated by storage organs, as in Drosophila wing disk regeneration, whereby intermediates of methionine and tryptophan metabolism from the fat body appear to act as permissive factors (Kashio and Miura, 2020; Kashio et al., 2016). To avoid growth mismatches and divert resources, the development of other body parts is also delayed during wing disk regeneration through the action of insulin-like peptides (Colombani et al., 2012; Garelli et al., 2012).

Importantly, intestinal microbiota is another source of nutrition-derived signals. Microbial metabolites, such as indoles derived from dietary tryptophan, have been implicated in regulating processes as distinct as whole-body regeneration in planarians (Lee et al., 2018) and liver regeneration in mammals (Zheng and Wang, 2021; Shavandi et al., 2020).

The impact of environmental nutrition input often goes beyond resource allocation and can have dramatic consequences on developmental trajectories. Developing organisms inevitably face fluctuating environmental conditions, including levels of nutrients, that require a coordinated whole-organism response. A variety of survival mechanisms have evolved as adaptations to starvation, which are often poorly understood. In the simplest form, these adaptations include adjusting developmental pace or modulating organismal size based on nutrient availability (e.g. González-Estévez et al., 2012; Layalle et al., 2008; Oviedo et al., 2003). Starvation can also divert the life cycle towards quiescent developmental stages, as demonstrated by the L1-arrested and dauer larvae of Caenorhabditis (reviewed by Rashid et al., 2021; Baugh, 2013; Sommer and Ogawa, 2011). In another nematode genus, Pristionchus, food shortage during larval development can additionally induce cannibalistic morphs, which exhibit morphological changes allowing the otherwise bacteria-eating larvae to successfully attack and consume other nematodes (Serobyan et al., 2013). One of the most extreme cases of life-cycle modification in response to starvation is ‘reverse development’ displayed by several cnidarians but most studied in the jellyfish Turritopsis (Matsumoto et al., 2019; Schmich et al., 2003),. Normal development of these animals proceeds from larvae to colonial polyps, which then give rise to medusae. However, in a stressful environment, medusae can revert to polyps through a larva-like stage. Despite its name, this alternative life cycle is not a direct reversal of normal developmental programs and involves massive cell re-specification (for a broader discussion of the phenomenon, see Piraino et al., 2004).

Nutritional modulation of development can also be part of the normal life cycle, such as with honeybees and other eusocial insects, whereby different diets induce changes in DNA-methylation patterns and juvenile hormone concentration, which are responsible for the alternative developmental trajectories of queens and workers (Abouheif, 2021; Kucharski et al., 2008; Foret et al., 2012). Although the mechanisms of nutrient-driven body-type and behavioral plasticity are undoubtedly complex, involving the previously mentioned pathways and many others, one unifying feature seems to be hormonal control, particularly through insulin and insulin-like peptides (reviewed by Romero and Eckel, 2021; Okamoto and Yamanaka, 2015; Liu et al., 2014).

Another underexplored aspect of organismal-level nutritional integration is the crosstalk between embryo and mother through nutrient provision. Embryos exhibit diverse modes of nutrition, such as placental transfer or yolk consumption, which lead to different possibilities for embryo-intrinsic and maternal regulation. Whereas lecithotrophic (yolk-dependent) embryos inherit the ‘past record’ of maternal nutrition, placentotrophic (placental) embryos experience maternal nutrition in real time. Interestingly, several studies of different lecithotrophic animal species, including fish, frogs, lizards and birds, suggest that the mother may be able to modulate the ratio between yolk volume and clutch size based on the environmental conditions as a ‘bet-hedging’ strategy to maximize offspring survival (Song et al., 2020; Morrongiello et al., 2012; Lips, 2001). However, whether such regulation exists for other parameters of embryonic nutrition, such as yolk composition, is still unclear (Van Dyke and Griffith, 2018). Because placentotrophic embryos develop inside the mother, they should be shielded from external environmental fluctuations. However, such buffering may be difficult to achieve, especially in animals with an invasive placenta (i.e. most mammals) where the embryonic and maternal bloodstreams are directly connected (Fig. 1C). This nutritional connection between the mother and embryo opens regulatory possibilities from the mother's side, but can also lead to pathologies in the offspring of metabolically unhealthy mothers or upon ingestion of harmful substances. However, embryo-intrinsic regulatory mechanisms that rely, for example, on changing transporter expression through insulin signaling have also been identified to buffer nutrient uptake and selection (Fowden et al., 2006; Constância et al., 2005).

Beyond elucidating the molecular mechanisms that explain how nutritional and metabolic signals control genetic and developmental events, understanding the underlying logic of this regulation is also an intriguing problem. Why and how do certain metabolic states guide developmental trajectories toward specific outcomes? Why do select metabolites have signaling functions? To what extent are these regulatory mechanisms adaptable to new nutrient sources under changing food-web dynamics? How can we direct the differentiation of pluripotent cells with nutritional input? The answers to these questions may also serve practical outcomes, such as facilitating the generation of artificial tissues or mitigating the effects of metabolic diseases.

To address these questions, we need to confront several neglected points and technical hurdles. One such challenge is understanding the regulatory roles of metabolic temporal dynamics. For example, changes of metabolic flux can be rapid and useful for regulating cellular events on finer timescales, but sensing long-term nutritional status likely requires different mechanisms. Moreover, a significant portion of the current knowledge still comes from in vitro cell culture experiments. Despite the value of their simplicity, the nutrient-rich environment of these systems and the lack of organismal framework may induce unusual metabolic states (McKnight, 2010). Thus, it is essential to integrate these findings into the context of tissues, organs and organisms. In mammalian models, various metabolic imaging modalities (e.g. Shan et al., 2022; Mächler et al., 2016; Rodrigues et al., 2014) and organ-specific metabolic flux measurements (e.g. Jang et al., 2019) have greatly enhanced our understanding of metabolism at the systemic level. However, owing to their size, morphological complexity and developmental time scales, keeping the whole organism in view during development is particularly challenging for mammals.

Relatively simple animals, such as worms, flies and sea anemones, offer great opportunities to measure and perturb nutritional metabolism during development and across multiple scales. Although possessing distinct tissues and organs, they are often small enough to be imaged as a whole, enabling the simultaneous interrogation of different levels of organization. Along with the ease and accessibility for targeted perturbations (e.g. optogenetics, genetics, microfluidics, etc.) without sacrificing the organismal context, they can be a powerful tool for understanding the principles of nutritional control of animal development. Recent studies have leveraged these advantages by combining metabolic sensors with targeted genetic perturbations to explore the metabolic interaction between the intestine and testes (Hudry et al., 2019), using high-throughput and -omics approaches to study the impact of diet on life history choices (Gandara et al., 2022; Watson et al., 2014), measuring body and organ size during development to uncover the principles of organ growth and size coupling (Stojanovski et al., 2022 preprint), and quantifying the metabolic costs of development and regeneration (Lewallen and Burggren, 2022; Rodenfels et al., 2019). Hence, invertebrate models can help us bridge the distance between easily tractable cell culture models and mammalian systems in which the conserved regulatory principles are deployed in more complex ways. This mechanistic insight will forge a path to a better understanding of how animals ‘become what they eat’, using nutrition as an environmental driver that informs and instructs developmental processes.

We apologize to colleagues whose work was not cited owing to space constraints. We thank Whitney Fropf and Jennifer Linn for proofreading the manuscript and two anonymous reviewers for their constructive suggestions.