Juvenile Hormone Is Required to Couple Imaginal Disc Formation with Nutrition in Insects

See allHide authors and affiliations

Science  02 Jun 2006:
Vol. 312, Issue 5778, pp. 1385-1388
DOI: 10.1126/science.1123652


In starved larvae of the tobacco hornworm moth Manduca sexta, larval and imaginal tissues stop growing, the former because they lack nutrient-dependent signals but the latter because of suppression by juvenile hormone. Without juvenile hormone, imaginal discs form and grow despite severe starvation. This hormone inhibits the intrinsic signaling needed for disc morphogenesis and does so independently of ecdysteroid action. Starvation and juvenile hormone treatments allowed the separation of intrinsic and nutrient-dependent aspects of disc growth and showed that both aspects must occur during the early phases of disc morphogenesis to ensure normal growth leading to typical-sized adults.

Research on growth control in insects has largely concentrated on imaginal discs because such structures determine the size and morphology of the adult (1). Imaginal disc growth has both extrinsic and intrinsic components (2, 3). Extrinsic components link growth to nutritional state via endocrine factors [such as the insulin-like peptides and ecdysone (47)], whereas instrinsic components associate growth with local morphogens (such as wingless and dpp) that establish patterns directing morphogenetic growth (2). Besides possessing growth-promoting factors, insects have a potential growth-inhibiting factor, the juvenile hormone (JH). JH is a sesquiterpene hormone that is released from the corpora allata (CA) and is responsible for directing the action of the ecdysteroid molting hormones ecdysone and 20-hydroxyecdysone. JH is characterized as a “status quo” hormone (8) because, in its presence, ecdysteroids cause molting larvae to repeat the most recent stage. However, the association of JH with growth control has generally been ascribed to this modulation of ecdysteroid action because the disappearance of JH in the last larval stage initiates the endocrine changes that cause metamorphosis (9).

The concepts of imaginal disc growth and morphogenesis are based on highly derived “early-forming” discs, such as those of Drosophila and the wing discs of Lepidoptera, such as Manduca sexta. The cells for these discs are set apart from the larval tissues during embryogenesis. They play no functional role in the larva but grow in concert with it and begin morphogenesis late in larval life (1, 10). In contrast, an ancestral strategy of disc formation is seen in the development of lepidopteran eye, leg, and antennal discs (11, 12). These are “late-forming” discs that arise from the primordia (Fig. 1A, red), which are fields of diploid cells that contribute to the larval body but retain the embryonic potential to form an imaginal disc (1115). Early in the last larval stage, cells at discrete sites in each primordium (Fig. 1A, blue) begin to transform, becoming columnar and starting to invaginate. Cellular morphogenesis and proliferation then spread from the sites through the primordium, forming a prominent invaginated disc by the start of metamorphosis 3 days later (Fig. 1C). The formation and growth of these late-forming discs depended on feeding and did not occur when last-stage larvae were starved (16, 17) (Fig. 1D). Starvation also results in elevated levels of JH (18) rather than the JH decline typical of the last larval stage. When this JH increase was prevented by the removal of the CA, we found that the discs formed and grew despite starvation (Fig. 1, E and G).

Fig. 1.

Interaction of hormone and nutrient manipulations on imaginal disc organization and growth in larval Manduca sexta. (A) Drawings showing the location of the eye and leg primordia (red) and wing imaginal discs (green) at the onset of the last larval stage. Blue shows the sites that start the disc transformation. (B to F) Propidium iodide–stained projections of confocal Z-stacks showing the state of the primordia and discs after different treatments. (B) At the start of the last instar (0 h, 0 hours), the primordia were quiescent and the wing disc was small. (C) After 4 days of feeding (4d), the primordia formed invaginated discs (arrows) and the wing disc grew and established the wing veins (inset, arrowhead). (D) Larvae starved for 4 days showed neither eye nor leg disc formation nor wing disc growth. (E) Larvae lacking their CA (CAX) and starved for 4 days showed wing disc growth and vein formation (inset, arrowhead), and both the eye and leg discs formed (arrows). (F) Treatment of starved-CAX larvae with JHM (pyriproxifen, 10 μg per larva) suppressed disc formation and wing disc growth and vein formation (inset). The red dashed line in the legs in (B) to (F) shows the muscle insertion on the dorsal distal tibia. (G and H) Quantification of the effects of hormone and nutrient manipulations on (G) proliferation in the tibial portion of the leg disc and (H) growth of the early-forming wing disc. Cell division was assessed by the number of phosphohistone H3 (PH3)–positive cells (those in mitosis), and growth was assessed by the dorsal area (total pixels × 100) of the wing disc. Circles indicate intact larvae that were fed normal diet (black), fed diet and treated with 10 μg of pyriproxifen per larva (red), or starved (open). Diamonds indicate CAX larvae that were starved (open), starved and treated with pyriproxifen (red), or fed a normal diet after 4 days of starvation (black/blue diamonds). Symbols are the average (±SEM) for four to five preparations.

The earliest known manifestation of the commitment to metamorphosis is the appearance of the broad transcription factor, a member of the Broad complex, Tramtrak, and Bric à brac/Pox virus and zinc finger (BTB/POZ) family (17, 19, 20). Broad-Z2 transcripts appeared in the leg primordium of feeding last-stage larvae by 12 hours after ecdysis and increased through 48 hours (Fig. 2A). Broad transcripts did not appear in starved larvae unless the larvae were also lacking their CA. Both the broad-Z2 expression (Fig. 2B) and disc formation and growth (Fig. 1, F and G) seen in starved allatectomised (CAX) larvae were completely suppressed by treatment with pyriproxifen, a stable JH mimic (JHM).

Fig. 2.

Interaction of JHM and nutrient manipulations on the expression of the broad gene in small pieces of larval epidermis containing the leg or eye primordium. The levels of broad-Z2 mRNA were determined by reverse transcription polymerase chain reaction and referenced to levels in fed larvae at 24 hours after ecdysis. (A) The time course of broad-Z2 appearance in the leg primordium of fed (solid circles) or starved (open circles) larvae. (B) The effects of feeding, CA removal (CAX), and treatment with pyriproxifen (JHM) on broad-Z2 expression in the leg and eye primordia. Values are at 24 hours after ecdysis except where indicated and are the average (±SEM) of three determinations per point.

An early-forming disc, the wing disc, showed similar growth responses to starvation and JH. During the last instar, the wing discs increased markedly in area and also established the sites of the future veins (Fig. 1C, inset) (21). Neither response occurred in starved larvae (Fig. 1, D and H) unless the larvae lacked their CA (Fig. 1, E and H). The wing disc growth and morphogenesis seen in starved CAX larvae were both blocked by treatment with JHM (Fig. 1F).

Because the developmental effects of JH are normally attributed to its modulating the actions of ecdysteroids (9), we used in vivo body ligations to test the involvement of ecdysteroids in disc formation and growth. Neck ligation decapitated the larva and deprived the thorax of cephalic endocrine factors, including JH. Decapitated larvae showed the formation and growth of leg (Fig. 3A) and wing discs (not shown in the figure) similar to that seen in starved CAX larvae, and the development of both discs was suppressed by JHM (Fig. 3B). A ligature applied behind the second thoracic segment additionally isolated the T3 primordia from ecdysteroids coming from the prothoracic glands in the anterior thorax. Nevertheless, the ecdysteroid-free portions posterior to the ligature still showed the organization and growth of their T3 discs (Fig. 3C). Therefore, ecdysteroids are not required for the formation and early proliferation of the discs.

Fig. 3.

The effects of treatment with pyriproxifen (JHM) on disc formation and growth. (A and B) Newly ecdysed last-stage larvae that were decapitated by neck ligation showed well-formed leg disc invaginations (arrow) by 4 days after ligation (A), but this morphogenesis was inhibited by treatment with 10 μg of pyriproxifen (B). (C) Ligation behind segment T2 to remove the influence of the prothoracic glands also resulted in disc formation (arrow). The red dashed lines in the legs in (A) to (C) show the muscle insertion landmark. (D and E) Localized treatment of one leg of a decapitated larva with pyriproxifen by covering it with wax impregnated with JHM suppressed disc formation in the treated leg (E) while the contralateral disc formed normally (D), as indicated by the arrow. (F) Pupae formed from larvae that were starved for 4 days and then fed ad libitum until metamorphosis. Left, CAX individual; right, control. (G and H) The dorsal region of the pre-wandering eye imaginal disc in a normally fed individual (H) and a CAX larva that had been starved for 4 days and then fed for 60 hours (G). In the latter, the growing eye disc failed to incorporate the small cells in the dorsal and anterior regions of the eye primordium (included in the red dashed line). (I) Scheme summarizing the role of JH in regulating the growth of adult primordia during larval life. 20E, 20-hydroxyecdysone. (A) to (E), (G), and (H) show propidium iodide–stained projections from confocal Z-stacks.

The growth suppression that accompanies starvation is generally attributed to the lack of growth-promoting factors (such as insulin-related peptides) (22, 23). Therefore, JH could be acting systemically to suppress the release of such factors. We tested for systemic versus local effects of JH by topically applying JHM to a single leg of neck-ligatured larvae [see the supporting online material (SOM)]. The treated leg showed no disc formation (Fig. 3E), whereas the remaining legs formed their discs normally (Fig. 3D). In vitro experiments also supported the direct action of JH in suppressing the organization and growth of the late-forming discs. Pieces of the fourth instar head cuticle that contained the eye primordium formed an eye disc when cultured without pyriproxifen, but disc formation and proliferation were blocked by the addition of JHM. The median effective dose for the suppression of disc proliferation was about 10–10 M pyriproxifen (fig. S1A). In vivo studies showed that JHM could act within 8 to 12 hours to suppress proliferation in discs that were just starting to form (fig. S1B).

Although the onset of metamorphosis is usually explained in terms of the loss of JH, there is clearly another factor involved. JHM treatment suppressed disc formation and growth in starved CAX larvae but not in larvae that were feeding (Fig. 1G) (16, 17). Similarly, treatment with JHM suppressed broad-Z2 induction in the primordia of starved CAX larvae but not in feeding larvae (Fig. 2B). This failure of JHM to suppress disc formation in feeding last-stage larvae cannot be explained by the enhanced clearance of the applied hormone, because topical treatment of the leg primordia with JHM was also ineffective in feeding larvae. Therefore, nutritional cues apparently result in the release of a metamorphosis-initiating factor (MIF) that overrides the suppression of disc formation by JH (16, 17). The release of a MIF in response to feeding appears to be unique to the last larval stage, because JH also inhibits disc formation in earlier larval instars but feeding does not overcome this suppression. The endocrine interactions during the last larval stage are known to be species-specific (24, 25), and although a MIF is an additional player in this interaction, its nature is unknown.

Although feeding overrides JH suppression of disc formation in the last instar, there are still JH-sensitive aspects of subsequent disc growth. Manduca larvae treated with JHM early in the last instar neither molted nor metamorphosed and grew to monstrous sizes (over 18 g) (26). In such larvae, leg disc proliferation and wing disc growth were normal through the first 2 days but then both processes declined markedly (Fig. 1, G and H). The growing discs started to differentiate on day 2, as demonstrated by the establishment of veins in wing discs (21) (Fig. 1C, inset), but this early differentiation was abnormal in feeding larvae treated with JHM. Hence, JH appears to affect two aspects of disc growth: the initial formation and growth of the disc and the subsequent differentiation of the simple disc epithelium into a complex structure. Nutrient-dependent factors can overcome the first aspect of JH suppression but apparently not the second.

The discs that grew in the starved CAX larvae illustrate the growth potential of the primordium in the absence of additional nutritional input. We have termed this type of growth, which is due to intrinsic signaling within the primordium, “morphogenetic growth.” Morphogenetic growth sets the lower size limits for discs that form from their respective primordia. These discs may correspond to the tiny discs formed in chico mutants of Drosophila that cannot respond to nutrient-related cues because of a mutation in the insulin-signaling pathway (27). The size difference between these discs and those formed by normally fed individuals is the extrinsic contribution of “nutrient-dependent growth.” Normally, intrinsic and extrinsic components act cooperatively during the early portion of the last instar as larvae are feeding. The use of CAX larvae, however, permitted us to examine disc growth when these two components were separated in time.

CAX larvae that were starved for 4 days and then given food formed severely stunted pupae as compared to the pupae of similarly treated intact larvae (Fig. 3F; see the SOM). The dramatic size difference reflected the size of imaginal discs made by CAX larvae versus those formed by intact larvae. For example, in normal feeding larvae, all cells of the eye primordium had been incorporated into the eye disc by the start of metamorphosis (Fig. 3H). The mature discs of the CAX larvae, in contrast, were bordered on their anterior and dorsal margins by small cells that were part of the eye primordium (Fig. 3G), and these boundaries were similar to those established by the end of starvation (Fig. 1E). We think that, in normal feeding larvae, the recruitment of cells into the forming disc spreads rapidly through the primordium because of the coexistence of nutrient-dependent and morphogenic signaling. In the starved CAX larvae, in contrast, the lack of the nutrient-dependent component resulted in slower recruitment, leaving many primordium cells outside of the nascent disc at the end of starvation. Because the disc boundaries moved little, if at all, in the CAX larvae during subsequent feeding, there appears to be a limited temporal window for the recruitment of cells into the nascent disc. This window had closed by the end of the starvation period, with only a fraction of the competent cells having been recruited into the disc. When nutrient-dependent mechanisms were subsequently invoked by feeding, no more recruitment of cells was possible and growth was confined to the smaller disc that contributed to a stunted pupa. Whereas recent emphasis on the study of disc growth and size focuses on nutrient-dependent mechanisms working through the insulin-like peptides (4), our study shows that these extrinsic signals must work in concert with intrinsic control systems for normal growth to occur. These intrinsic mechanisms seem especially important early in disc growth, and their occurrence in the absence of nutrition results in stunted pupae despite the subsequent access of larvae to abundant high-quality food. This finding may have parallels to human conditions in which early infant malnutrition results in long-term effects on growth, including stunting and obesity (28).

JH has long been known to be active at the start of molts to modulate ecdysteroid action (9). Here we have shown that JH is also necessary throughout the intermolt periods (Fig. 3I) to allow the isomorphic growth of the primordia together with that of the larva as a whole, while suppressing morphogenetic signaling systems that would transform the primordia into imaginal discs. This larval action of JH in suppressing morphogenesis is an extension of that seen in embryos of more basal insects, in which premature exposure to JH suppresses embryonic patterning and induces precocious terminal differentiation (12, 29). Hence, an ancient developmental role of this hormone appears to be in switching tissues between programs of growth and/or morphogenesis. JH-like molecules occur in taxa other than the insects (30, 31), where they likely have similar developmental roles. Molecules with such actions might be exploited as antitumor agents in the future.

Supporting Online Material

Materials and Methods

SOM Text

Fig. S1


References and Notes

View Abstract

Navigate This Article