The essential role of bursicon during Drosophila development

Abstract

Background: The protective external cuticle of insects does not accommodate growth during development. To compensate for this, the insect life cycle is punctuated by a series of molts. During the molt, a new and larger cuticle is produced underneath the old cuticle. Replacement of the smaller, old cuticle culminates with ecdysis, a stereotyped sequence of shedding behaviors. Following each ecdysis, the new cuticle must expand and harden. Studies from a variety of insect species indicate that this cuticle hardening is regulated by the neuropeptide bursicon. However, genetic evidence from Drosophila melanogaster only supports such a role for bursicon after the final ecdysis, when the adult fly emerges. The research presented here investigates the role that bursicon has at stages of Drosophila development which precede adult ecdysis.

Results: We addressed the mechanism and timing of hormonal release from bursicon-positive motor neurons at the larval neuromuscular junction. Our findings indicate that vesicle membrane proteins which are required for classical neurotransmitter release are also expressed at these peptidergic motor neuron terminals; and that these terminals secrete hormones including bursicon at the neuromuscular junction, coinciding with larval ecdysis. This release surprisingly occurs in two waves, indicating bursicon secretion preceding and following the ecdysis sequence. Next, we addressed the functional significance of bursicon signaling during development, by disrupting the expression of its receptor, rickets, in different target tissues. We determined that rickets is developmentally required in the epidermis and imaginal discs for proper formation of the prepupa. It is also required to harden the pharate adult cuticle before eclosion. Significantly, we have also found that the available rickets mutants are not genetic nulls as previously believed, which necessitated the use of targeted RNA interference to disrupt rickets expression.

Conclusions: Our results are consistent with the view that bursicon is the insect tanning hormone. However, this is the first study to rigorously demonstrate both its release and function during development. Importantly, we provide new evidence that bursicon release can precede the initiation of larval ecdysis, and that bursicon tans the puparium. Our results firmly establish bursicon signaling as essential to insect growth and development.

Figure 1

Bursicon is expressed in type III boutons at the larval NMJ. Bursicon immunoreactivity (BURS-IR) has previously been shown in the periphery [10], without confirming its distribution pattern. (Aa) BURS-IR at a representative muscle 12 at the larval NMJ. (Ab) HRP immunoreactivity labels all bouton types at NMJ 12. (Ac) Bursicon distribution is found in some, but not all boutons at NMJ 12. (B) To verify if bursicon is expressed in type III boutons, we used transgenic larvae expressing the fluorescent neuropeptide marker ANF-EMD with a CCAP-GAL4 promoter. (Ba) BURS-IR at a representative muscle 12. (Bb) Vesicles with ANF-EMD are distributed in type III boutons. (Bc) BURS co-localizes with the ectopic ANF-EMD marker. Scale bars = 10 μm. An antibody which recognizes the bursicon-α subunit was consistently used for BURS-IR.

Figure 2

Bursicon co-localizes with the pre-synaptic markers N-SYB and CSP at the NMJ. (A) Neuronal synaptobrevin (N-SYB) expression at NMJ 12 includes type III boutons. (Aa) BURS-IR in type III boutons. (Ab) Distribution of N-SYB-IR at NMJ 12. (Ac) N-SYB is expressed in multiple bouton types, including type III. (B) Cysteine string protein expression pattern at NMJ 12 includes type III boutons. (Ba) BURS-IR as a marker for type III boutons. (Bb) CSP-IR labels multiple boutons at NMJ 12. (C) BURS and CSP patterns co-localize in type III boutons. The expression of CSP immunoreactivity in type III boutons appears to be weaker than in other boutons at NMJ 12. In (A) and (B), arrows indicate representative boutons that co-express BURS and the relevant pre-synaptic marker, whereas arrowheads indicate boutons that do not express bursicon. Scale bars = 10 μm.

Figure 3

Two phases of vesicle release from type III boutons overlap with larval ecdysis. We observed fluorescence changes in the type III boutons of CCAP>ANF-EMD larvae. Six stages were chosen that broadly extend over the duration of the 2^nd larval ecdysis: 'DMH', which precedes the onset of ecdysis; 'DVP', which coincides with the initiation of ETH release; 'FE', which signifies the completion of ecdysis; '+2 hours' after FE; '+3 hours' after FE; and 'L3', wandering 3^rd instar larvae. For DVP, we selected animals at the earliest point when the new vertical plates were recognizable. At each stage, the fluorescence intensity of all visible type III boutons from 4 animals was measured and converted to fluorescence intensity averages (see Methods), in arbitary units (top panel). Representative type III boutons from each stage are shown below the corresponding fluorescence intensity measurements with accompanying asterisks representing single-factor ANOVA results for consecutive stages (bottom panels). The results significantly show two waves of ANF-EMD release, before DVP and after FE, as shown by an asterisk. For *, p < 0.005. Error bars indicate +/- SEM. Scale bar = 10 μm.

Figure 4

Hormone release from type III boutons at the pupal NMJ coincides with pupal ecdysis. Changes in fluorescence intensity are also observed in type III boutons of CCAP>ANF-EMD animals at pupal ecdysis. (A) P3 stage pupa, preceding pupal ecdysis by several hours. (B) Pupa at end of P4(i) stage, shortly after initiation of the ecdysis sequence. (C) P4(ii) stage pupa immediately following head eversion (occurring during pupal ecdysis). (D) P5(i) pupa following elongation of legs, approximately 1/2 hour after head eversion. Average fluorescence intensity is reported on a log scale. For (D), no detectable fluorescence intensity is reported. Identification of pupal stages is described in the Methods. Error bars indicate +/- SEM. Scale bar = 10 μm.

Figure 5

Knock-down of burs transcripts by RNAi impedes progression to the pharate adult stage. (A) Development of Act5C>burs RNAi animals proceeds until the pharate adult stage, when all progeny die trapped within the puparium. (B) BURS-IR comparisons of Act5C>burs RNAi larvae and driver-less UAS-burs RNAi control larvae confirms that ubiquitous expression of burs RNAi severely limits the expression of bursicon at the NMJ, as quantified by fluorescence intensity. For each data point, larval NMJs (n = 16) from two animals were observed. Representative 'RNAi' and 'Control' NMJs accompany the data points (in lower panels). Error bars indicate +/- SEM.

Figure 6

Lethality results from ubiquitous expression of UAS-rk RNAi. (A) Within standard food vials, Act5C>rk RNAi progeny all die before the end of the 3^rd larval stage. These larvae often exhibit the double vertical plates phenotype, indicative of failed larval ecdysis. New 3^rd instar mouthparts are labeled with the arrow; the 2^nd instar mouthparts which failed to shed are labeled with the arrowhead. (B) If Act5C>rk RNAi larvae are rescued from food vials and allowed to develop on grape juice agar plates, they progress to the pupal stage. Lethality is 100% during this stage, but the timing and phenotype are highly variable. Note the pale color and flattened shape of the puparia. All pupae are shown at the same scale, from the ventral side. (C) A second Act5C-GAL4 stock was also used, referred to here as Act5C(II). Although most Act5C(II)>rk RNAi progeny die as pupae (data not shown), adults occasionally eclose. Their wings never expand and their cuticle never tans. Panels A, B, and C are not shown at the same scale. (D) RT-PCR with pupal cDNA templates was performed to confirm that ubiquitous expression of UAS-rk RNAi knocks down rk transcript levels. Lanes 1 and 3: UAS-rk RNAi (control with no driver). Lanes 2 and 4: Act5C>rk RNAi. Open arrow: expected genomic band size for rk. Closed arrow: expected cDNA band size for rk. Note the absence of an appropriate sized rk cDNA band for Act5C>rk RNAi in lane 2. Feathered arrow: expected cDNA size for the positive control, RpS26.

Figure 7

Analysis of leg phenotypes reveals that rk¹ is not a null mutant. The homozygous rk¹ stock displays incomplete penetrance of leg deformities. Leg deformities in rk¹ homozygotes are classified as resembling wild-type legs (A), moderately kinked (B), or severely kinked (C). A moderate leg deformity is defined by the presence of a kink in the first tarsal segment (arrow). Severe leg deformities additionally exhibit a bulbous tarsal segment (arrowhead) and a rotated tarsal segment (feathered arrowhead). The corresponding percentage of leg deformities is given (A-C) for a random sample of adult metathoracic legs from the rk¹ stock (n = 150). (D) When hemizygous, 100% of rk¹/Df(2L)BSC252 animals exhibit the most severe leg deformities (n = 64). This enhanced penetrance indicates that the rk¹ allele is not a null. Scale bar = 0.1 mm.

Figure 8

Phenotypes resulting from different GAL4 drivers of UAS-rk RNAi. (A) All T76>rk RNAi progeny die with small, soft, untanned puparia before pupal ecdysis. The puparia often have a slight crescent bend to them, rather than a straight orientation along the anterior-posterior axis. (B) In contrast, T80>rk RNAi develop into pharate adults, but all are then unable to eclose from their puparia and die. (C) A limited number of 69B>rk RNAi flies are able to eclose, but their legs are flaccid and collapse under the weight of the body. All flies die within 24 hours, before ever expanding their wings or tanning their cuticle. (D) C855a>rk RNAi flies successfully eclose without any obvious problems. However, all adults have drooping, partially extended wings. This appears to be a result of the successful deployment of wing expansion behaviors in the absence of cuticular tanning. Only A and B are shown at the same scale, to emphasize the smaller size of T76>rk RNAi pupae.

Figure 9

Expression patterns of different GAL4 drivers in larval tissues. The membrane-bound GFP reporter UAS-mCD8::GFP was expressed with different drivers to assay for strength of expression and pattern in CNS (brain), ventral nervous system (VNS), imaginal discs, and epidermal tissue. Compared to all other drivers, Act5C(II) expresses strongly in all tissue with solid patterns. T76 expresses strongly in CNS and epidermis. Expression in the imaginal discs is faint, but uniform. T80 expresses strongly in CNS and imaginal discs, but is absent from epidermal tissue. 69B expression is strong in CNS and imaginal discs. It can also be detected in a weak, very restricted pattern in the epidermal tissue. C855a expression is noticeably restricted in CNS and imaginal discs, as compared to Act5C(II). Expression is completely absent in the epidermis. MJ33a expression is restricted in the brain, extremely weak in the epidermis and discs, and absent from the VNS. For imaginal discs, examples from wing, haltere or metathoracic leg disc are shown. Epidermis refers to the epidermal tissue of the larval body wall.

Figure 10

Comparisons between GAL4 expression patterns and UAS-rk RNAi phenotypes. (A) Key to the schematics. Assayed tissues are labeled as brain, VNS, imaginal discs and epidermis. (B) Correspondence of T76 expression pattern with T76>rk RNAi phenotype. Uniform expression in all tissues assayed appears to prevent the progression of prepupal development, resulting in small, untanned puparia. (C) Correspondence of 69B expression pattern with 69B>rk RNAi phenotype. Solid expression in brain, VNS, and imaginal discs, but weak pattern in epidermis, appears to result in flies with flaccid legs and unexpanded wings. (D) Correspondence of C855a expression pattern with C855a>rk RNAi phenotype. Extremely limited expression in all tissues assayed still results in adult flies whose wings cannot expand. The limited expression pattern likely allows rk expression required for wing expansion behaviors, but subsequent tanning of the wing is not possible, resulting in the sagging wing appearance. In the panels, solid green tissue symbolizes a uniform GFP pattern as observed in Figure 9, whereas the green punctae symbolize a restricted pattern of GFP expression.