Arthropods possess a hard exoskeleton that is articulated in many regions for bending and flexibility. This outer cuticle is the secreted product of the underlying hypodermal cells. Its properties, from soft and pliable to stiff and solid, are determined by multiple factors including chitin architecture, the degree of sclerotization, the degree of hydration, and, most importantly, by the composition of proteins in the cuticular matrix (Andersen et al., 1995; Andersen,2003). Resilin is one of the more unusual cuticular proteins. It is present in high concentrations at locations where specialized mechanical functions are required, especially where there is a need for repetitive strain, elasticity, or energy storage. In insect flight systems, resilin acts as an important kinetic energy-storing system during wing strokes (Weis-Fogh,1960). The presence of resilin in wings may assist in their folding, prevent material damage during repeated folding and unfolding, and allow the wings to be deformed by aerodynamic forces, as observed in the flexible joints of hind wing veins (Haas et al.,2000a), the wing membrane in beetles (Pachnoda marginata and Coccinella septempunctata), and in the wings of earwigs (Forficula auricularia) (Haas et al.,2000b) and damselflies (Enallagma cyathigerum) (Gorb,1999). In fleas, click beetles, and froghoppers, resilin plays a role in storing energy via prolonged muscular action and releasing this energy very rapidly when needed (Sannasi,1969; Rothschild and Schlein,1975; Burrows et al.,2008). Resilin-bearing cuticle also acts to counteract movement caused by muscles, such as in the opening mechanism of the spiracle in locusts (Miller,1960), the sound-producing tymbals in cicadas (Young and Bennet-Clark,1995), and the joints connecting segments in the legs of the cockroach, Periplaneta americana (Neff et al.,2000).
The development of biomimetic resilin-like compounds (Elvin et al.,2005; Qin et al.,2009) and their possible applications (Kopecek,2007) means that the methods by which insects utilise resilin will come under more intense scrutiny. Most studies that identify resilin-rich regions rely on the optical and chemical properties of resilin, combined with an elasticity or energy storage function. Methods include fluorescence emission under ultraviolet light, known as auto-fluorescence, pH dependence of the auto-fluorescence, and positive staining with toluidine blue and methylene blue. These techniques have been shown to be the most efficient ways to detect the presence of resilin in insects and crustaceans (Weis-Fogh,1960; Neff et al.,2000) but doubt can remain, as it is rare for all methods to be applied. While Drosophila pro-resilin was the first insect resilin gene to be completely sequenced, to date there has been no direct confirmation that resilin is indeed expressed as a cuticular protein in Drosophila as inferred from the chitin-binding domain (R&R Consensus) (Andersen,2010). We use an antibody generated against exon 1 of recombinant Drosophila resilin (Elvin et al.,2005) to show that resilin is expressed in the developing larval epidermis of Drosophila and we identify sites of expression in the developing adult cuticle of late-stage pupae. We show that the antibody cross-reacts, binding to sites of suspected or known resilin expression in other insect species, and therefore has wide application in confirming the presence of resilin and allowing investigation of its spatial and temporal expression.
Resilin in Drosophila melanogaster
Embryonic expression of resilin.
Using conventional methods of detection, embryos were observed under ultraviolet illumination but there were no indications of regions showing the strong sapphire-blue auto-fluorescence characteristic of the dityrosine crosslinks found in resilin. Resilin expression was then investigated using the anti-Rec1 antibody (Elvin et al.,2005), a polyclonal antibody raised in rabbits to a recombinant protein composed of the first exon of the Drosophila melanogaster resilin gene, CG15920. Initial staining of embryos of different ages showed that resilin detection was restricted to late embryos. At or near the 19-hr time point, most of the embryos had emerged as larvae and were, therefore, not amenable to antibody staining because of the barrier to antibody diffusion presented by the mature cuticle.
Embryos were fixed and immunostained at 20-min intervals from 17 to 19 hr of development. In whole mounts, epidermal expression patches were seen to develop in waves in a dynamic, repeatable pattern (Fig. 1A). Expression first appeared at the anterior of the embryo. The anterior patches persisted through to 19 hr. In the thoracic and abdominal segments, initially three to four bilateral patches were visible, each concentrated mid-laterally. Next, ventro-lateral patches were seen. The extent of staining expanded to cover progressively more segments between 17 and 19 hr. The 20-min gap between collection points allowed classification based on the number of expressed patches on the epidermis of the embryo (Fig. 1A). At time-point 17 hr, in general 1 to 3 pairs of expression patches were seen. At 17 hr 20 min, most embryos displayed embryos with three to five pairs of patches. At 17 hr 40 min, most had four, five, and six expression patches. At 18 hr, more than five patches were seen and, in general, the patches became more dispersed rather than focused (Fig. 1A). A control was performed to accompany one of the immunostains by eliminating the primary antibody. No signal was detected (not shown). Patches of expression appeared to be restricted primarily to the hypodermal cells but expression was also found in the cells of chordotonal stretch receptors (lch5) in the embryo (Fig. 1B,C) that lie immediately beneath the hypodermis (Hartenstein,1988). The strongest staining in epidermal patches and subepidermal chordotonal sense organ cells is intracellular (Fig. 1C,E) and may be associated with the perinuclear endoplasmic reticulum.
Resilin expression in pupae and adults.
Having confirmed the immunoreactivity of the anti-Rec1 resilin antibody in whole-mount embryos, wings and legs of Drosophila pharate adults were next stained with the antibody. Pharates were chosen because antibody diffusion was inhibited in tissues bearing mature cuticle. Resilin expression was seen at the base of sensilla in pupae, with marked concentration in the socket region of the articulated sense organs (Fig. 2A,B).
Using auto-fluorescence under ultra-violet illumination as a detection method, the largest single concentration of putative resilin in adult Drosophila was at the wing bases (Fig. 3A). Auto-fluorescence in a pair of pads at the wing bases was normal at pH 7 and pH12 and was quenched at pH 2 (Fig. 3B). The reactivity of the anti-Rec 1 antibody was tested by embedding pupal wing-bases in resin followed by sectioning and immunostaining, rather than by immunostaining whole tissues. The antibody bound to the pads at the base of the pupal wings at the same location as the pads identified by fluorescence microscopy (Fig. 3C).
Resilin in Other Insects: Anti-Rec1 Antibody Cross-Reactivity in Bactrocera, Flea, and Dragonfly
To assess whether the antibody cross-reacts with resilin in different species and to ascertain whether the embryonic expression pattern seen in Drosophila is more widespread, a similar experiment was performed on a distant relative of Drosophila, the tephritid fruitfly Bactrocera tryoni. B. tryoni embryos were collected, fixed and immunostained in accordance with the Drosophila protocol. Staining patterns observed in the embryos were similar to those seen in Drosophila embryos (Fig. 1D,E). Expression developed in waves, starting with one segmental pair of epidermal patches, expanding to more than six, as observed in Drosophila embryos. Morphological features of the gut and body wall indicated that the B. tryoni embryos were at a similar developmental stage as the Drosophila embryos when comparing staining patterns.
A swollen, elastic section of the tendon of the pleuro-subalar muscle of dragonflies is perhaps the largest known site of internalised resilin (Andersen and Weis-Fogh,1964). Four such resilin-bearing flight tendons are present in dragonflies, one at the base of each wing, attached to the base of the wings via tendons. They were identified by Weis-Fogh (1960) as concentrations of almost pure resilin and used to test resilin's mechanical and structural properties including its elasticity (Weis-Fogh,1961). The tendon shows the sapphire-blue auto-fluorescence under UV illumination used to characterise resilin (Fig. 4). Immunostaining of sectioned dragonfly tendons with anti-Rec1 antibody showed cross-reactivity of the antibody. Staining was restricted to the swollen putative resilin-containing component rather than muscle or adjacent fibrous tendon (Fig. 4B,C).
Another well-known site of resilin is the pad in the pleural arch within the thorax of fleas that stores energy for release when the flea jumps (Rothschild and Schlein,1975). First, we confirmed that the pleural arch of the cat flea, Ctenocephaledes felis, shows strong blue auto-fluorescence under UV illumination (Fig. 4D). Pleural arches were fixed, embedded, and sectioned, then immunostained with anti-Rec1 antibody. A crescent of resilin-bearing cuticle in the pleural arch stained brown upon immunostaining using a HRP-conjugated secondary antibody (Fig. 4E). A control with no primary antibody produced no comparable staining (not shown).
Anti-Rec1 Resilin Binds to Resilin Across a Variety of Insects
The anti-Rec1-resilin antibody generated against the first exon of the putative Drosophila resilin CG15920 has been used here to detect and localize expression of resilin in embryos, pupae, and adults of Drosophila melanogaster. The binding of the antibody at sites in Drosophila where resilin-bearing structures have been previously reported in other insects such as pads at the wing bases (Andersen and Weis-Fogh,1964; Neff et al.,2000) and the socket region of sensilla (Thurm,1964; Chevalier,1969) provides confirmatory evidence that CG15920 is a resilin-encoding gene of Drosophila. The antibody binds at sites in insects other than Drosophila where resilin-bearing structures have been reported. Two canonical sites of resilin-rich pads are the pleural arch of the flea (Rothschild et al.,1975) and the tendon connecting dragonfly flight muscle to the thoracic hypodermis (Weis-Fogh,1960). Both sites bind the anti-Rec1 antibody and both sites show the strong blue autofluorescence characteristic of resilin (Neff et al.,2000). No binding of the anti-Rec1 antibody to adjacent cuticles that do not show the strong auto-fluorescence was evident, indicating a correlation between the autofluorescence characteristic of resilin and antibody binding to those sites.
Resilin Expression in Developing Touch Receptors
The antibody revealed two additional sites of staining in Drosophila: the bases of touch receptors in developing adults and the lateral body wall of developing larvae. The strong anti-Rec1 binding at the base of the touch-receptive sensilla on the Drosophila pupal leg confirms the presence of resilin in the joint membrane region of insect mechanoreceptors, postulated in the past using various diagnostic methods. Thurm (1964) showed that the cuticle at the base of each sensillum of the honey bee hair plate is elastic and stains positively with methylene blue, one of Weis-Fogh's (1960) diagnostic characteristics for identification of resilin. Subsequently, transmission electron microscopy has shown that the flexible socket cuticle, termed the joint membrane, is composed of a homogeneous material, possibly resilin (Matsumoto and Farley,1978; Keil and Steinbrecht,1984; Keil,1997), embedded in a matrix of fibrous cuticle. Touch sensilla are likely to undergo many repeated deflections in an insect's lifetime. For example, Drosophila adults spend a considerable proportion of time undergoing stereotyped grooming behaviours, involving rubbing the legs repeatedly over body parts and over each other (Szebenyi,1969), bending the tactile mechanoreceptors. Functional tactile mechanoreceptors are a prerequisite for efficient grooming behaviour as shown through mutant analysis (Phillis et al.,1993; Melzig et al.,1996) so resilin is likely to provide the return force necessary when hairs are deflected at the socket.
Resilin Expression in Developing Epidermis
Given that embryos and larvae of Drosophila, along with many other larval Diptera, tend to have a very thin cuticle, resilin expression was unexpected. However, a pattern of expression in epidermal patches was seen during late embryogenesis, in both Drosophila and another fly, Bactrocera. Expression is first seen in a single epidermal expression patch 17 hr after egg deposition and by 20 hr the epidermal expression patches become more widespread and less discrete. The narrow developmental window in which strong expression is seen could reflect a time when pro-resilin is being actively secreted for incorporation into the cuticle. An ultrastructural examination of cuticle formation in Drosophila embryos noted that cuticle deposition is visible 12–16 hr after egg-laying (Hillman and Lesnik,1970; Moussian et al.,2006). However, it is not until stage 17, the stage at which resilin expression is seen in this study, that chitin formation is seen in embryos (Moussian et al.,2005). Consequently, the resilin expression seen in the current study at 17–20 hr occurs well after the initiation of cuticle formation but at about the same time as chitin formation.
It is not clear why resilin is expressed in bilateral, segmental patches but we can speculate that this pattern may produce regions of highly flexible lateral cuticle through the incorporation of pro-resilin into the developing cuticle. The lateral patches happen to overlie the lateral penta-scolopidial organs (lch5), stretch receptors in the embryo that are slung between a proximal and distal attachment point where the cable-like ligament cell and attachment cell make contact with the hypodermis (Hartenstein,2005). Functional chordotonal organs are essential for normal peristaltic locomotion by larvae (Caldwell et al.,2003; Hughes and Thomas,2007). One possibility is that resilin elasticises the cuticle between the upper and lower attachment points of the chordotonal organs. Alternatively, differential flexibility of cuticle within segments may be related to sites of muscle attachment. Ultrastructural analysis or confocal microscopy using different subcellular markers is required to distinguish the precise cellular location of the pro-resilin in these embryonic tissues.
The possibility that the embryonic expression patterns seen in Drosophila represent cross-reaction to non-resilin proteins or related resilin-like proteins should be considered. We consider that this is unlikely for several reasons. First, the recombinant protein used to generate the antibody (Elvin et al.,2005) is composed of 17 repeats of a 15–amino acid YGAP-containing motif designed to emulate the first exon of CG15920, the only resilin-like gene in Drosophila and one that is orthologous to resilins of other insect species (Lyons et al.,2011). Second, we have demonstrated cross-species specificity to known or suspected sites of resilin in insects as diverse as fleas, dragonflies, and flies (Orders Siphonaptera, Odonata, and Diptera, respectively), without any significant cross-reaction at other cuticular sites, defined as staining that approaches or matches the staining intensity at the known sites of resilin. Further, the same anti-Rec1 body binds to well-characterised resilin-bearing sites in yet another order of insects, the jumping apparatus of froghoppers (Order Hemiptera) (Burrows et al.,2011).
The anti-Rec1 antibody described here is an important addition to the arsenal of tests available to detect resilin-bearing cuticle in insects. For the first time, resilin expression has been reported during insect embryogenesis, in this case in Drosophila and Bactrocera fly embryos.
The antibody was produced by immunizing rabbits against a recombinant protein composed of the first exon of the Drosophila melanogaster resilin gene, CG15920, the same recombinant protein that produced an efficient rubber-like hydrogel after photochemical cross-linking (Elvin et al.,2005). An alignment of CG15920 with the resilin sequence from flea, buffalo fly, and dragonfly has been published (Lyons et al.,2011). The antibody was affinity-purified by passing polyclonal serum over a sepharose affinity column to capture Rec-1-binding antibodies. The column was washed and the bound anti-Rec1 antibodies then eluted.
Immunostaining of Staged Drosophila melanogaster and Bactrocera tryoni Embryos
Wild type Oregon R Drosophila melanogaster and Bactrocera tryoni adults were induced to lay eggs on an apple juice agar plate (0.175 g agar, 6.5 ml water, 0.17 g sugar, 1.7 ml apple juice per plate) at 25°C. Drosophila egg collections were timed to provide embryos aged between 17 and 19 hr at 20-min intervals. The eggs were collected the next day and placed in 50% v/v sodium hypochlorite. After 10 min, the sodium hypochlorite was removed with two to three washes of Drosophila saline (0.65% NaCl, 0.014% KCl, 0.02% NaHCO3, 0.012% CaCl2, and 0.001% NaH2PO4.2H2O at pH 7). Embryos were then transferred into a microcentrifuge tube and the saline removed. To fix the embryos, 500 μl of 4% paraformaldehyde (1:3 40% paraformaldehyde in PBT (4 ml of 10% Triton-X, 0.25 g of BSA, 10 ml of 10× PBS in 100 ml) and 500 μl of heptane solution was then added to the microcentrifuge tube. This was subsequently incubated at room temperature for 16 min on a tabletop rotator. The fixative was then removed from the solution by removing the lower layer, and 500 μl of methanol was added to the mixture and vortexed for 1 min. Devitellinized and fixed embryos were washed two to three times in methanol.
Embryos were rehydrated in PBT, followed by three 5-min washes in PBT. The embryos were then incubated in blocking solution (horse or goat serum) at 1:100 dilution for 1 to 2 hr at room temperature on a rotator. This was then replaced with 1:100 primary Drosophila anti-Rec1 antibody and incubated overnight at 4°C. The primary antibody was removed using three quick washes and three 5-min washes in PBT. The secondary antibody (Sigma HRP conjugated anti-rabbit secondary diluted 1:25) was then added and incubated for approximately 2 hr at room temperature. The secondary antibody was then rinsed out by three quick washes and three 5-min washes in PBT. A 1:1 solution of DAB and PBT was added and the embryos incubated for 5 to 10 min. After the incubation, 1 μl of a 30% hydrogen peroxide solution was added. The color reaction was monitored under the dissecting microscope and stopped by 2 fast washes in PBT followed by two 2-min washes in PBT. Embryos were then cleared in 70% glycerol overnight and mounted in 100% glycerol the next day. Stained embryos were imaged under a Zeiss Axioskop microscope at 20×/40× and 100× (transmitted light SNT 12V 100W) with a color CCD camera MTI 3CCD camera model DC330E using Scion Image software. Filter set FS 01 was used for imaging with an excitation wavelength of 365 nm and an emission wavelength of 397 nm.
pH and Autofluorescence of Adult Structures
A pH 1.1 and a pH 12 phosphate buffered saline (8 g NaCl, 0.2 g KCl, 1.44 g Na2HPO4, 0.24 g KH2PO4) solution was prepared for this part of the study. Legs, wings, and resilin pads from Drosophila melanogaster and the flea, Ctenocephaledes felis were dissected from surrounding tissue and placed on a glass slide. Petroleum jelly was placed at the four corners of the coverslip and the coverslip placed on the specimen. The PBS was then changed using a tissue to wick the solution from one end while absorbing a drop of pH-adjusted saline from the other. Specimens were viewed using a Zeiss Axioskop (transmitted light SNT 12V 100W; fluorescence HBO100W) and imaged with a colour CCD camera (MTI 3CCD model DC330E) at identical settings for the pH range.
Preparation of Samples for Sectioning and Antibody Staining
Drosophila melanogaster pupal legs, flea Ctenocephaledes felis resilin pads, and dragonfly resilin flight tendons (unidentified species of dragonfly) were dissected and incubated in 4% paraformaldehyde in 0.1 M phosphate buffer either for 4 hr at room temperature or overnight at 4°C. The fixative was then washed out by three rinses of 10 min each in phosphate buffer at room temperature. This was then taken through another three rinses of 10 min each in distilled water at room temperature. Specimens were then taken through a series of dehydration steps in a graded ethanol series. They were dehydrated at 10 min each, starting from 50, 60, 70, 80, 90%, and, finally, twice at 100%. Specimens were then infiltrated with LR White resin at room temperature. Specimens were placed in ethanol:resin mixture at 1:1 for an hour followed by 100% resin for 1 hr and finally 100% resin overnight. Infiltration was performed on a rotator at room temperature. The resin was then polymerised in gelatin capsules at 60°C for 24 hr.
Samples were removed from the gelatin capsules and sectioned on an ultra-microtome. Semi-thin 3-μm sections were cut, removed from the water reservoir using Pasteur pipette, placed onto a drop of water on a slide, and flattened onto the slide on a heating block (80°C). Some slides were stained with 0.5% toluidine blue in 1% borax solution before immunostaining.
Sections were washed quickly twice, and then twice for 5 min in PBT on a slide in a staining chamber prepared by placing a round moistened filter paper in a plastic Petri dish. Blocking was then performed by adding 5% normal goat serum (NGS) in PBT for 1 to 2 hrs. The primary antibody was then added at 1:100 dilution with 5% NGS in PBT and incubated overnight at room temperature. Three quick washes were then performed followed by three 5-min washes in PBT. A HRP-conjugated rabbit secondary antibody was placed on the slide (1:25 dilution in PBT) and incubated for an hour at room temperature. The antibody was removed by three 5-min rinses. For detection, a 1:1 solution of DAB to PBT with 1 μl of 30% hydrogen peroxide was placed on the slide. The colour reaction was monitored under a dissecting microscope and terminated by two quick washes followed by two 2-min washes in PBT.
We thank Jason Rice for providing Bactrocera eggs and Novartis Australia for supplying Ctenocephalides.