The effects of insulin/TOR signaling on the final body size are masked by JH in the final instar
In this study, we have found that feeding rapamycin leads to a delay in the timing of metamorphosis in both wild-type and black mutant Manduca. However, the peak body size of wild-type larvae is unaltered by rapamycin, whereas black mutant larvae fed rapamycin grow to a significantly larger body size. Ecdysteroid biosynthesis has been shown to be impacted by TOR signaling in Manduca [18]. Furthermore, our findings show that by feeding rapamycin, the timing of initiation of metamorphosis is delayed, indicating that the timing of ecdysteroid release is impacted by TOR signaling. However, the presence or absence of JH determines whether the final size is impacted by the nutrient-sensitive pathways: in the wild-type Manduca, the JH-mediated delay masks the effects of the nutrition-sensitive insulin/TOR signaling pathway, and the final size of the animal is not influenced by nutrient-sensitive pathways. In contrast, when JH titers are low, inhibition of the insulin/TOR signaling pathway can significantly delay the timing of metamorphosis and impact the final body size. Our findings thus suggest that JH acts to protect larvae from nutritional disturbances. In animals with higher JH titers, fluctuations in nutritional-sensing pathways have minimal effects on the final body size, whereas in low JH animals, fluctuations in nutrient-sensing pathways can impact the final body size significantly.
In contrast, in Drosophila, the insulin signaling pathway directly impacts the timing of ecdysteroid biosynthesis at critical weight [11]. Furthermore, the terminal growth period has been shown to be lengthened by inhibition of TOR signaling in the prothoracic glands [14]. Thus, in Drosophila, the timing of metamorphosis relies primarily on nutritional inputs that directly modulate ecdysteroid biosynthesis.
Since rapamycin significantly delays the timing of ecdysteroid secretion in the black mutant Manduca and impacts the final body size, we think that insulin/TOR signaling also regulates the terminal growth period in the black mutants. In this case, the JH-mediated delay period is removed, and the final body size of the larva reflects a nutrition-sensitive pathway. Our study therefore shows that insulin/TOR signaling and JH signaling provide two distinct mechanisms for determining the final size of the larva and that, depending on the species, one or the other mechanism dominates. We can therefore envision a seesaw with either JH or insulin/TOR signaling dominating to regulate the timing of metamorphic onset (Fig. 7, right). In wild-type Manduca, JH clearly dominates and determines the timing of metamorphosis based on body size. In contrast, in Drosophila, insulin/TOR signaling dominates to set the timing of metamorphosis based on nutritional intake. In black mutant larvae, we presume that the roles of JH and insulin/TOR signaling are more or less equal, because the larvae still have some JH but inhibiting insulin/TOR signaling delays the timing of metamorphosis. Thus, neither body size nor nutritional intake dominates.
In penultimate instar Manduca larvae, insulin/TOR signaling plays a critical role in the timing of molting [18]. When larvae are fed rapamycin, ecdysteroid titers are dramatically suppressed and the initiation of a molt is delayed, so that the larva grows to a larger size before molting. Since the molt of a penultimate instar larva is relatively unaffected by JH [24], ecdysteroid production is purely dependent on the rate of ecdysteroid biosynthesis. This situation would be similar to what is observed in the final instar Drosophila and JH-deficient Manduca. In contrast, in the wild-type Manduca, ecdysteroid release is dependent on JH because the PTTH secretion from final instar brains becomes sensitive to JH [5, 25]. Thus, JH in the final instar wild-type Manduca plays a much more predominant role in regulating developmental transitions than in the earlier instars or in Drosophila.
The critical weight is not influenced by insulin/TOR signaling
We found that feeding rapamycin does not influence the critical weight in both wild-type and black mutant larvae. Our observations corroborate a previous finding that in wild-type Manduca the critical weight is regulated by a size-sensing pathway instead of a nutrient-sensing pathway. This size-sensing mechanism has been linked to oxygen levels [7]. In response to this size assessment cue, JH is cleared from the hemolymph, triggering the release of PTTH from the brain.
In contrast, in Drosophila, the critical weight is not linked to a body size-sensing mechanism. Instead, the critical weight is marked by a small increase in the ecdysteroid titer, which in turn is controlled by a nutrient-sensitive pathway [11, 26, 27]. Clearly, the underlying primary mechanism of critical weights differs between the two species. A previous study has shown that allatectomized Manduca larvae behave similarly to Drosophila in that if starved within the first 12 h of molting, the time of gut purge is delayed, whereas starvation after 24 h feeding no longer delays metamorphosis [28]. This switch in the response to starvation in the JH-less state is sensitive to a nutrition-dependent pathway and is similar to the situation in Drosophila. In contrast, in both the wild-type and black mutant Manduca, the critical weight observed is a distinct phenomenon and is clearly independent of the insulin/TOR signaling pathway.
The prothoracic gland size, however, was influenced by the depression of insulin/TOR signaling. Interestingly, in the wild-type Manduca, the prothoracic gland size at critical weight was drastically different between the rapamycin-fed and DMSO-fed animals (Fig. 6). It has previously been shown that the prothoracic glands could be an indicator for size in Drosophila. The prothoracic gland reaching a certain size is thought to act as a proxy for the overall body size and attainment of critical weight in Drosophila larvae [12]. It is clear that in wild-type Manduca, the prothoracic gland size is not correlated with the critical weight. Thus, these findings further support the notion that the switch in response to starvation is distinctly regulated in Manduca and Drosophila. We observed that the prothoracic gland growth in rapamycin-fed black mutant larvae was not disproportionately affected relative to the body size. This curious effect might imply that the scaling relationship between the body size and the prothoracic gland size is coupled in the absence of JH.
The black mutant larvae exhibit a bail-out response to starvation
In Drosophila, starvation after the critical weight leads to earlier onset of metamorphosis [10, 12]. These “bail-out mechanisms” have been observed in other insects, such as dung beetles that rely on ephemeral food supplies [28]. In the black mutant larvae, starvation post critical weight also led to earlier onset of metamorphosis, although the effect was only significant when larvae were fed rapamycin. Thus, in the absence of JH and insulin/TOR signaling, it appears that animals exhibit a bail-out response. We suggest that bail-out responses may be the default and that JH overrides these mechanisms to prolong the growth period in species where the costs of a small final body size outweigh the costs of delayed development.
The mechanism underlying this bail-out response is unknown at this time. We hypothesized that tissues in feeding larvae might grow and inhibit ecdysteroid release by secreting an inhibitory factor. In Drosophila, dILP8, secreted from growing imaginal discs, can delay the release of ecdysteroids from the prothoracic glands [23]. However, at least when wing imaginal discs were removed, the feeding larvae continued to delay metamorphosis relative to the starved larvae. This is perhaps not too surprising since discless mutant Drosophila also does not delay metamorphosis [29]. We therefore propose that some other factor, perhaps secreted from the fat body or other imaginal tissues besides wing discs, may be involved in the bail-out phenotype.
TOR signaling is a component of the molt timer
Recently, a developmental timer was identified in which allatectomized larvae initiate metamorphosis after four days as long as larvae feed on amino acids for one day and even when the size of the larva is substantially below the critical weight [19]. Given that TOR signaling is an amino acid-sensitive pathway, it is likely a part of the timer. Our finding that feeding rapamycin to JH-deficient black mutants significantly increases the delay period indicates that TOR signaling likely contributes to the timer. Whether or not TOR signaling is the sole regulator of the timer remains to be seen.