Within a particular phylum, the specific endocrine regulators involved in metamorphosis appear to be similar. In most vertebrates, thyroid hormone signaling is a key endocrine pathway that regulates growth, development/morphogenesis and metabolism (Rabah et al., 2019). Thyroid hormone is produced and secreted from the thyroid gland and plays a chief role in metamorphosis in amphibians and fish (Gudernatsch, 1912). The main form of thyroid hormone secreted from the thyroid gland is thyroxine (T4), which is biologically inactive and is subsequently converted to the biologically active triiodothyronine (T3), which coordinates metamorphosis (Denver et al., 2002). This conversion is mediated by the enzyme type II iodothyronine deiodinase (Davey et al., 1995). In target tissues, thyroid hormone enters the cell and regulates the expression of target genes in several different ways. In vertebrates, T3 typically binds to the nuclear Thyroid hormone receptor (TR) (Sap et al., 1986; Weinberger et al., 1986), which together with the co-receptor retinoid co-receptor (RXR), bind to DNA and regulate transcription (Zhang and Kahl, 1993; Zhang and Lazar, 2000). The peak in thyroid hormone titers coincides with the beginning of metamorphosis and coordinates myriad morphological and physiological changes from resorption of the tail to growth of limbs and remodeling of the gut (Shi, 2000; Brown and Cai, 2007). Different tissues of a tadpole undergo metamorphic changes at distinct time points. For example, a metamorphosizing tadpole grows its limbs before losing its tail so that it can continue to swim while the limbs grow out. This tissue specific timing of metamorphosis is regulated by the distinct timing of appearance of mRNAs encoding TR, RXR and type II iodothyronine deiodinase (Yaoita and Brown, 1990; Kawahara et al., 1991; Wong and Shi, 1995; Shi et al., 1996; Cai and Brown, 2004). Thyroid hormone is both necessary and sufficient for metamorphosis in teleost fishes. For example, when flounder larvae are exposed to T4, they can accelerate metamorphosis, leading to small juveniles, whereas disruption of thyroid hormone production by thiourea leads to retention of larval traits (Inui and Miwa, 1985). Exogenous thyroid hormone is also sufficient to induce early metamorphosis in larvae of the grouper, Epinephelus coioides (de Jesus et al., 1998).

Thyroid hormone is part of the hypothalamic–pituitary–thyroid (HPT) axis (Figure 2A). As in mammals, thyroid stimulating hormone (TSH), which is secreted from the pituitary gland, stimulates the production of thyroid hormone. In amphibians, TSH release is in turn regulated by corticotropin releasing hormone (CRH) from the hypothalamus rather than the thyrotropin-releasing hormone as is the case in mammals (Denver, 1999). CRH is a potent regulator of metamorphosis and appears to overcome the negative feedback of thyroid hormone on TSH release (Manzon and Denver, 2004). In teleost fishes, the role of CRH in regulating thyroid production appears to be limited to some species (Larsen et al., 1998; Campinho et al., 2015).

The HPT axis interacts with the hypothalamic–pituitary–interrenal (HPI) axis, which responds to stress. The HPT axis begins with the hypothalamus releasing CRH, which stimulates the anterior pituitary to release adrenocorticotropic hormone (ACTH) (Figure 2A). ACTH acts on the interrenal glands to release corticosteroids, the key mediator of stress responses.

Corticosteroids also interact with the thyroid hormone pathway and regulate the developmental changes induced by thyroid hormone. The application of hydrocortisone accelerates T3- and T4-induced metamorphosis in Bufo bufo, Rana hechsheri and Rana pipiens (Frieden and Naile, 1955). Corticosterone was also found to stimulate T3-induced metamorphosis in Xenopus laevis (Gray and Janssens, 1990). Corticosteroids act on tissues by enhancing tissue sensitivity to thyroid hormone: Aldosterone and corticosterone increase T3 binding in tadpole tails (Niki et al., 1981; Suzuki and Kikuyama, 1983), and cultured tadpole tails exposed to corticosteroids express higher transcript levels of type II deiodinase and TR (Krain and Denver, 2004; Bonett et al., 2010). It is thus possible that the production of corticosteroids due to environmental stressor can accelerate metamorphosis by enhancing tissue sensitivity to thyroid hormone (Wada, 2008; Denver, 2021; Figure 2). The evidence for teleost fishes is more ambiguous: Although cortisol can enhance the impacts of T3 on fin-ray resorption of the Japanese flounder, Paralichthys olivaceus, in vitro, the timing of metamorphosis is not impacted by cortisol in vivo (de Jesus et al., 1990). The lack of in vivo effects may be because sufficient amount of cortisol is produced endogenously (de Jesus et al., 1990).

Thyroid hormone can play an essential role during metamorphosis of other Deuterostomes (Box 1), including several Echinoderm species (Chino et al., 1994; Heyland and Hodin, 2004; Heyland et al., 2006) and possibly also ascidians (Patricolo et al., 1981, 2001). Whether thyroid hormone acts via TR is not as well-established in these non-vertebrate Deuterostomes although TR is present in all Deuterostomes studied to date (Taylor and Heyland, 2017). Intriguingly, recent studies have also suggested the involvement of thyroid hormone signaling in accelerating molluscan metamorphosis (Fukazawa et al., 2001; Taylor and Heyland, 2017). Although regulators of corticosteroid action have been identified outside vertebrates (Baker, 2010), the role of corticosteroids during metamorphosis in these species remains unknown.

Before undergoing metamorphosis, most insects undergo several larval molts—the process involving the shedding of the exoskeleton to allow for growth. Within insects, the main developmental hormones are juvenile hormone (JH) and ecdysteroids (Nijhout, 1998; Truman, 2019; Figure 1). Generally, periodic surges of the 20-hydroxyecdysone (20E) trigger larval-larval molting as well as the initiation of metamorphosis. During the larval stage, JH prevents a larva from undergoing metamorphosis and therefore came to known as the “status quo hormone” (Riddiford, 1996). JH alters the effects of 20E action and inhibits metamorphic genes from being activated (Nijhout, 1998; Liu et al., 2009; Jindra et al., 2013, 2015). When bound to the Ecdysone receptor (EcR), 20E activates a transcriptional cascade of genes which induces molting (Riddiford et al., 2000) and adult tissue morphogenesis by activating a transcription factor called Ecdysone-induced protein 93 (E93) (Belles and Santos, 2014; Jindra, 2019; Truman and Riddiford, 2019). Conversely, JH binds to its receptor Methoprene-tolerant (Met) and induces the expression of Krüppel homolog 1 (Kr-h1), which represses E93 (Belles and Santos, 2014). Together, these regulators comprise the MEKRE93 pathway, which appears to be highly conserved across most insects studied to date (Belles, 2019, 2020). During metamorphosis, these regulators play critical roles in regulating metamorphic timing (Rountree and Bollenbacher, 1986; Mirth et al., 2005; Yamanaka et al., 2013; Hatem et al., 2015; Nijhout, 2015) and hence the final body size (Nijhout and Williams, 1974; Caldwell et al., 2005; Callier and Nijhout, 2013; Nijhout et al., 2014). We will discuss how hormones impact body size in Section “ENVIRONMENTAL IMPACTS ON METAMORPHIC HORMONES.” In addition, tissue proliferation and morphogenesis are also regulated by these hormones through their action on many target genes (Champlin and Truman, 1998a,b; Truman and Riddiford, 2002, 2007; Mirth et al., 2009; Herboso et al., 2015).

Just as the major metamorphic hormones of vertebrates are regulated by the brain, the production of metamorphic hormones in insects is also regulated by the brain, which can integrate various environmental cues (Figure 2B). Ecdysteroids production and release is regulated by prothoracicotropic hormone (PTTH), which is synthesized in the brain and released by the corpora cardiaca. JH synthesis and release is stimulated and inhibited by neuroendocrine factors called allatotropins and allatostatins, respectively, although the roles of these factors in the regulation of metamorphosis remain poorly understood (Goodman and Granger, 2005; Nijhout et al., 2014). Based on studies done in the silkworm, Bombyx mori, the allatostatins appear to act directly on the corpora allata whereas allatotropins appear to act indirectly by inhibiting Short neuropeptide F (sNPF), an inhibitor of JH biosynthesis that is produced in the corpora cardiaca (Kaneko and Hiruma, 2014; Figure 2B). JH activity is also modulated by JH degradation enzymes, JH esterase (JHE) and JH epoxide hydrolase (JHEH).

In addition to these two metamorphic hormones, Insulin-like peptides act on the Insulin/Target of rapamycin (TOR) signaling pathway and impact growth of insects (Koyama et al., 2020). This pathway plays an important role in regulating growth rate and determining the overall body size of the adult (Brogiolo et al., 2001; Geminard et al., 2009). Nutritional availability influences growth in almost all animals, and Insulin/TOR signaling pathway links growth of organisms to nutrient availability (Masumura et al., 2000; Ikeya et al., 2002; Geminard et al., 2009). In addition, this pathway plays a major role during metamorphosis to control tissue specific growth (Shingleton et al., 2005; Tang et al., 2011). Insulin-like peptides are often released in response to nutrients (Park et al., 2014) although in many cases, the interaction is indirect. For example, in fruitfly larvae, different tissues sense amino acids and sugars and release factors that then travel to cells that release insulin-like peptides (Figure 2B; Colombani et al., 2003; Geminard et al., 2009; Kim and Neufeld, 2015; Sano et al., 2015; Agrawal et al., 2016; Koyama and Mirth, 2016; Nässel and Broeck, 2016). Once released, insulin-like peptides travel to other parts of the body where they bind to the Insulin receptor, which activates a signal transduction cascade that ultimately leads to the phosphorylation of the forkhead transcription factor, Forkhead box O (FoxO), which regulates many developmentally and physiologically relevant genes (Koyama et al., 2020). The Insulin/TOR signaling pathway interacts with the ecdysteroid signaling pathway in a complex manner: Insulin/TOR signaling regulates the production of ecdysone, thus impacting the timing of metamorphosis (Mirth et al., 2005), while ecdysteroids also act to suppress Insulin signaling (Colombani et al., 2005; Mirth et al., 2014).

In vertebrates, various environmental cues have been shown to influence hormone titers. For example, T4, T3 and corticosteroid levels all increase rapidly when tadpoles of the Western spadefoot toad, Scapiopus hammondii encounter decreasing water levels (Denver, 1998). These environmental changes are sensed by the brain neurons, which trigger an increase in CRH release from the hypothalamus, activating the HPT axis (Denver, 1998; Boorse and Denver, 2003). These changes are correlated with an earlier onset of and small body size at metamorphosis (Denver et al., 1998).

Temperature also impacts T3 and corticosteroid levels. In leopard frog tadpoles, Lithobates pipiens, corticosteroid levels peak earlier and T3 levels are elevated at higher temperatures (Freitas et al., 2017). Similarly, and tadpoles of the American bullfrog, Lithobates catesbeianus, also have elevated T3 levels at higher temperatures (Freitas et al., 2016). Higher temperatures are associated with faster growth and earlier onset of metamorphosis and smaller sizes at metamorphosis (Smith-Gill and Berven, 1979; Leips and Travis, 1994; Álvarez and Nicieza, 2002). Although hormonal changes could explain some of these changes, it is also possible that the phenotypic effects could also result from increased rates of intrinsic biochemical reactions and an overall reduction in cell size (Atkinson and Sibly, 1997).

Nutrition also impacts the timing of metamorphosis of anurans. There is a critical size above which food deprivation accelerates metamorphosis (Leips and Travis, 1994) and leads to smaller body sizes at the time of metamorphosis (Denver et al., 1998; Nicieza, 2000). These impacts appear to be regulated by hormones. T3 and corticotropin-releasing hormone levels are increased in food restricted mid-prometamorphic S. hammondii tadpoles (Boorse and Denver, 2003), and thyroid glands from starved late pre- to early prometamorphic Rana catesbeiana tadpoles also produce significantly higher amounts of T4 (Wright et al., 1999).

In insects, a complex interaction between various endocrine regulators determines the timing of metamorphosis (Koyama et al., 2020). Within a particular species, the timing of metamorphosis can shift depending on environmental conditions, such as temperature and nutrient availability (Davidowitz et al., 2003). Both heritable differences in developmental time and plastic responses to the environment may involve alterations in endocrine regulators. In the lab, higher temperatures almost always lead to small adult body sizes by shortening the growth period (Davidowitz et al., 2003, 2004; Klok and Harrison, 2013). Observations in the field are much more complex and appear to depend on several factors including the number of generations, temperature, survival, and photoperiod (e.g., Roff, 1980; Atkinson, 1994; Imasheva et al., 1994; James et al., 1997; Horne et al., 2015).

Although studies have explored the cellular basis of temperature-dependent differences in body size (Partridge et al., 1994; Atkinson and Sibly, 1997; Zwaan et al., 2000), we still do not have a clear understanding of how temperature during the growth period impacts endocrine events that regulate life history transitions. However, the environment can impact hormones that regulate growth. A recent study on the cricket Modicogryllus siamensis demonstrated that higher rearing temperatures lead to enhanced Insulin/TOR signaling, leading to faster growth rate (Miki et al., 2020). Insulin/TOR signaling, however, does not impact the number of instars in M. siamensis; instead, the timing of JH decline impacts the duration of the juvenile growth period in a photoperiod-dependent manner (Miki et al., 2020). Thus, body size determination appears to rely on a complex interaction of endocrine regulators that respond differently to distinct environmental cues. Furthermore, we still do not understand how temperature influences the duration of larval stage, and more studies are needed to address this issue.

In addition, the environment can impact the timing of diapause and adult eclosion. Diapause is a dormant stage in insects that is equivalent to hibernation in vertebrates. Depending on the species, diapause can occur during different life history stages but metamorphic hormones often play prominent roles in regulating both the entry and duration of diapause (Chippendale and Yin, 1975; Zdarek and Denlinger, 1975; Sim and Denlinger, 2008, 2013). Environmental conditions, such as temperatures, can impact metamorphic hormones to influence the timing and duration of diapause (Turnock et al., 1986; Green and Kronforst, 2019; Cambron et al., 2021). We suspect that hormonal responses to environmental conditions are the norm, and that species can utilize these cues to coordinate life history transitions and phenotypic outcomes.

We end this section by discussing how hormones play prominent roles in polyphenisms. Polyphenic organisms can produce two or more distinct phenotypes from one genotype depending on the environment. A classic example of a polyphenism includes the polyphenisms of horned beetles where smaller male beetles have no horns on the head or the thorax, whereas larger male beetles grow horns (Kijimoto et al., 2013). These alternative morphs are both adaptive: Horned males use their horn as weapons to engage in male-male combat and guard the tunnels in which females are found, whereas hornless males “sneak by” the males by creating side-tunnels and gain access to the females (Emlen, 1997). Other examples of polyphenisms include the diet-induced polyphenisms of the caterpillars of Nemoria arizonaria, which can either develop into oak twig-resembling larvae or catkin-resembling morphs (Greene, 1989), and butterfly wing polyphenisms, where adult morphs adopt distinct wing color patterns depending on the season (Nijhout, 2003).

In polyphenisms, hormones play a salient role in instructing identical genomes to give rise to distinct adult morphologies that are adapted to particular environments (Nijhout, 1999, 2003). Because of the major effects developmental hormones have on adult tissue morphogenesis, small changes in the endocrine system can lead to profound changes during metamorphosis that results in distinct, and at times spectacular, adult phenotypes (Figure 3). In many polyphenisms, the endocrine centers integrate environmental stimuli encountered by the larva and adjusts the amount and timing of hormone production/release/response. For example, in the squinting bush brown butterfly Bicyclus anynana, the adult wing has eyespots that serve as defense against potential predators. Depending on the environment, both the ecdysteroid titers and the amount of ecdysone receptors expressed on the wing discs change (Monteiro et al., 2015) and impact the size of eyespots. In another butterfly, Precis coenia, the wings can be red and brown depending on the photoperiod and the temperature and their impacts on ecdysteroid levels during the early pupal stage (Rountree and Nijhout, 1995). Similarly, the alternative morphs of horned beetles are regulated by the titers of JH and ecdysteroids that are modulated by the amount of nutritional consumption (Emlen and Nijhout, 1999, 2001).

A glance at the organisms living around us highlights the diversity of body sizes across species. Although body sizes and hence the timing of metamorphosis can be impacted by environmental conditions, these differences can be explained by specific-specific differences: No matter how much a fruit fly larva eats, it will never grow as large as a bullfrog. At least some of the diversity of body size can be explained by genetic changes in the timing of metamorphosis (heterochrony).

Heterochronic shifts in the timing of thyroid hormone-mediated metamorphosis can impact adult sizes in Deuterostomes. In amphibians, premature exposure to thyroid hormone can cause the tadpole to initiate metamorphosis at a much smaller size than normally observed (Gudernatsch, 1912; Shi et al., 1996). In contrast, experimental ablation of thyroid glands can cause the tadpole to continue feeding and grow to an enormous size (Allen, 1916). An extreme case of heterochronic shifts has been documented in the direct-developing anuran, Eleutherodactylus coqui (Callery and Elinson, 2000). In this species, thyroid hormone production is initiated during embryogenesis such that the tadpole state is bypassed and a miniature adult frog hatches from the eggs. Shifting the timing of metamorphosis thus has profound impacts on the size of the adult. Similarly, exposure to thyroid hormone or thyroid hormone inhibitors can accelerate or delay, respectively, the timing metamorphosis in echinoderms (Heyland and Hodin, 2004).

In insects, body size can respond readily to artificial selection with corresponding shifts in the timing of metamorphosis (Grunert et al., 2015). In fact, it is the heritable changes in the endocrine response to the environment that often appears to be under selection and to underlie the divergent life history strategies. For instance, insects have evolved distinct responses to starvation depending on the feeding ecology (Callier and Nijhout, 2013; Hatem et al., 2015; Nijhout, 2015; Nagamine et al., 2016; Helm et al., 2017; Xu et al., 2020). In species that feed on ephemeral food sources, starvation often triggers an immediate switch to metamorphic induction by activating ecdysteroid production, ensuring that the larvae regardless of their size will initiate metamorphosis (Mirth et al., 2005; Helm et al., 2017). In species that have reliable food supply, starvation halts ecdysteroid synthesis, leading to a delay in the timing of metamorphosis (Nijhout, 2015; Xu et al., 2020). Moreover, different species have distinct threshold sizes, which is the size checkpoint that determines when a larva can metamorphose (Nijhout, 1975). Threshold size plays a critical role in the final size of the adult and does so by ultimately determining the timing of JH decline (Chafino et al., 2019; He et al., 2020).

In species with larvae that feed and grow, changes in the timing or rate of metamorphic hormone synthesis, release or sensitivity can influence final adult size (Figure 3). Because the endocrine regulators themselves do not change, such changes can occur without disrupting the process of metamorphosis itself. Thus, heterochronic shifts in the timing of metamorphosis, and hence the evolution of final adult size, may occur over just a few generations. We note that heterochronic changes can also occur at the level of individual tissues or behavior. Such heterochronic shifts can occur when traits become modularized and respond to hormones in a trait-specific manner (see next section).

Although the pleiotropic effects of metamorphic hormones might temporarily constrain evolution of metamorphic events, the sensitivity of target tissues to hormones may not be constrained in the same manner. Adaptive change in the sensitivity of tissues allows individual traits to be regulated independently from the rest of the body. Modularization (Box 1), or the evolution of a unique set of responses to hormones, releases the constraints imposed by the pleiotropic effects of endocrine regulators. Endocrine regulators can also be recruited to regulate new developmental event in a tissue specific manner (a process known as co-option) (True and Carroll, 2002).

The most obvious demonstration of modularization and/or co-option of hormonal pathways in adult development is seen in insect polyphenisms. A recent survey of nymphalid butterflies has demonstrated that 20E titers fluctuate in a thermally sensitive manner regardless of the effect on wing coloration (Bhardwaj et al., 2020). Thus, in polyphenic butterflies, the pigment specification and/or synthesis pathways appear to have co-opted the pre-existing thermally-sensitive ecdysteroid peak of metamorphosis so that the adult wing coloration can be modulated by the larval environment. This example suggests that (1) hormonal levels respond readily to the environment and (2) target tissues can evolve to respond uniquely to the fluctuating hormones.

In other polyphenic traits, hormones that regulate growth of the body can have an exaggerated effect on specific parts of the body. The impressive weapons of rhinoceros beetles grow larger because insulin signaling has an outsized effect on the growth of the head horns (Emlen et al., 2012). Similarly, the disproportionate growth of the horns and mandibles in some beetle species is regulated by localized effects of hormones that arise due to tissue specific sensitivities to metamorphic hormones (Emlen and Nijhout, 1999, 2001; Gotoh et al., 2011, 2014). Thus, when individual tissues acquire the ability to uniquely respond to hormones, phenotypes can overcome hormonal pleiotropy and diversify (Figure 4B). Such changes could arise, for example, by the increased production of the hormone receptor or by more efficient conversion of the prohormone to an active hormone in a particular tissue (Figure 4B).

Finally, we note that modularity facilitates heterochronic shifts of modules. Hormones can act on individual modularized traits and either speed up or slow down development relative to an ancestral trait. Thus, heterochronic changes and modularization can both facilitate phenotypic diversification. For example, changes in thyroid hormone have been suggested to underlie the diversification of barb species in Lake Tana: Experimental alterations of thyroid hormone levels in Lake Tana barbs Labeobarbus intermedius, for example, can accelerate or slow down craniogenesis and produce a bony skull that resembles that of Labeobarbus brevicephalus and Labeobarbus megastoma, respectively (Smirnov et al., 2012; Shkil and Smirnov, 2016). Thyroid hormone does not uniformly impact craniogenesis. Rather, different skull bones have distinct sensitivities to thyroid hormones, allowing thyroid hormone to heterochronically alter the development of skull bones in a modular fashion (Shkil et al., 2012; Shkil and Smirnov, 2016).