Jan Michels, Department of Functional Morphology and Biomechanics, Institute of Zoology, Christian-Albrechts-Universität zu Kiel, Am Botanischen Garten 1-9, D-24118 Kiel, Germany. Tel: +49-431-880-4511; fax: +49-431-880-1389; e-mail: email@example.com
Resilin is a rubber-like protein found in the exoskeleton of arthropods. It often contributes large proportions to the material of certain structures in movement systems. Accordingly, the knowledge of the presence and distribution of resilin is essential for the understanding of the functional morphology of these systems. Because of its specific autofluorescence, resilin can be effectively visualized using fluorescence microscopy. However, the respective excitation maximum is in the UV range, which is not covered by the lasers available in most of the modern commercial confocal laser scanning microscopes. The goal of this study was to test the potential of confocal laser scanning microscopy (CLSM) in combination with a 405 nm laser to visualize and analyse the presence and distribution of resilin in arthropod exoskeletons. The results clearly show that all resilin-dominated structures, which were visualized successfully using wide-field fluorescence microscopy (WFM) and a ‘classical’ UV excitation, could also be visualized efficiently with the proposed CLSM method. Furthermore, with the application of additional laser lines CLSM turned out to be very appropriate for studying differences in the material composition within arthropod exoskeletons in great detail. As CLSM has several advantages over WFM with respect to detailed morphological imaging, the application of the proposed CLSM method may reveal new information about the micromorphology and material composition of resilin-dominated exoskeleton structures leading to new insights into the functional morphology and biomechanics of arthropods.
Confocal laser scanning microscopy (CLSM) has been shown to have a great potential for the detailed visualization of the exoskeleton of arthropods using autofluorescence (Zill et al., 2000; Klaus et al., 2003; Schawaroch et al., 2005; Klaus & Schawaroch, 2006; Michels, 2007). However, the respective studies solely aimed at demonstrating that the tested method is appropriate for analyses of the general morphology, whereas the visualization of differences in the material composition of the exoskeleton was not investigated. With a single exception (Michels, 2007), only one excitation wavelength (either 543 nm or 568 nm) was used in these studies. Michels (2007) applied up to three excitation laser lines between 405 nm and 633 nm, but the resulting autofluorescences were not in all cases detected separately so that material differences were not resolved in detail. Nevertheless, the author emphasized the potential of applying CLSM and autofluorescence to analyse differences in the material composition of the exoskeleton.
Earlier studies had already successfully applied wide-field fluorescence microscopy (WFM) to investigate such material differences in insect exoskeletons based on differences in the autofluorescence composition (e.g. Haas et al., 2000; Niederegger & Gorb, 2003; Perez Goodwyn et al., 2006). As for most of the different autofluorescences observed in arthropod exoskeletons the corresponding material sources are still not exactly known, accurate interpretations of the results of such fluorescence analyses with respect to the material composition are not yet possible without applying additional microscopy techniques and histological and histochemical methods. However, the autofluorescence of resilin, which has been shown to be present in the exoskeleton of many arthropods (e.g. Weis-Fogh, 1960; Andersen & Weis-Fogh, 1964; Burrows, 2009), has been investigated in detail (Andersen, 1963). Resilin is an extremely elastic and flexible (also described as ‘rubber-like’) protein, which is found in high proportions in certain parts of the exoskeleton (Weis-Fogh, 1960, 1961; Andersen & Weis-Fogh, 1964). In most cases, such structures with high proportions of resilin are important elements of movement systems and provide high flexibility and resistance to frequent loads. Typical examples are the wing hinge and the prealar arm of locusts (Weis-Fogh, 1960), the jumping mechanism of fleas (Bennet-Clark & Lucey, 1967), the sound-producing tymbal of cicadas (Young & Bennet-Clark, 1995) and vein joints in the wings of damselflies (Gorb, 1999). As the presence of resilin strongly contributes to the particular material properties of these structures, it is essential for the understanding of the structures’ functioning to get detailed information about the presence and distribution of resilin. Because of the good knowledge of the characteristics of the resilin autofluorescence, with the excitation and emission maxima being in the range of 320 nm and 415 nm, respectively (Andersen, 1963), fluorescence microscopy with adequate filter sets provides an effective tool for the analysis of resilin. In this context, CLSM should enable to determine the presence and distribution of resilin in the exoskeleton with even greater detail. Unfortunately, most modern commercial confocal laser scanning microscopes are not equipped with UV lasers and thus do not cover the optimal excitation of the resilin autofluorescence. In studies with wide-field fluorescence microscopes, however, bandpass excitation filters transmitting light with wavelengths of 330–380 nm, 340–380 nm and 350–407 nm have proved to successfully stimulate the autofluorescence of resilin (Gorb, 1999; Haas et al., 2000; Neff et al., 2000; Niederegger & Gorb, 2003; Perez Goodwyn et al., 2006; Burrows, 2009), although these excitation bands are at the edge of the above-mentioned excitation maximum at 320 nm. Commercial confocal laser scanning microscopes are often equipped with a 405 nm laser, whose wavelength is at the upper border of the excitation spectrum of the resilin autofluorescence. Using this laser and appropriate laser and detector settings might thus make a successful visualization of resilin possible. The goal of this study was to test the potential of CLSM, in combination with such a 405 nm laser, for detailed visualization and analyses of the presence and distribution of resilin in the exoskeleton of arthropods.
Material and methods
In the following, exoskeleton structures consisting of material with a resilin proportion, which is larger than the proportions of each of the other constituents, are referred to as ‘resilin-dominated structures’.
Animals and structures used for the analyses
For testing the potential of the resilin visualization using CLSM, known resilin-dominated structures including the wing hinge and the prealar arm of the locust Locusta migratoria (Insecta, Caelifera, Acrididae; see Weis-Fogh, 1960), a vein joint in the hind wing of the dragonfly Sympetrum striolatum (Insecta, Odonata, Libellulidae; see Gorb, 1999), and the so-called resilin-bearing spring in the pretarsus of the hoverfly Eristalis tenax (Insecta, Diptera, Syrphidae; see Niederegger & Gorb, 2003) were visualized using both the WFM method, which had been used in earlier studies, and CLSM equipped with a 405 nm laser. Furthermore, the CLSM method was used to visualize additional exoskeleton structures that were very likely to contain relatively high proportions of resilin. Among those were the neck membrane of the dragonfly Libellula depressa (Insecta, Odonata, Libellulidae), the compound eye lenses of the green lacewing Chrysoperla carnea (Insecta, Neuroptera, Chrysopidae) and the ant Monomorium pharaonis (Insecta, Hymenoptera, Formicidae), tarsal joints in the legs of the beetle Coccinella septempunctata (Insecta, Coleoptera, Coccinellidae) and mechanoreceptors in the pretarsus of E. tenax and the cerci of the cricket Acheta domesticus (Insecta, Ensifera, Gryllidae).
Besides exclusively using autofluorescence, we also visualized the resilin autofluorescence in combination with the fluorescence of Congo red stained chitin (see Michels & Büntzow, 2010). For this, the calanoid copepod Temora longicornis (Maxillopoda, Copepoda, Calanoida, Temoridae) was used.
Specimen preparation and mounting
All animals used in this study were freshly frozen and stored at −70°C. After thawing, the structures of interest were removed using a stereomicroscope (Wild M8, Leica Microsystems GmbH, Wetzlar, Germany) and, according to requirements, a razor blade, small scissors and fine preparation needles. Subsequently, they were transferred to glycerine (≥99.5%, free of water, two times distilled; Carl Roth GmbH & Co. KG, Karlsruhe, Germany). In preparations where air was trapped in or on the specimen (e.g. in the veins of the dragonfly wing), the respective structures were immersed in a mixture of 25 ml distilled water and five small droplets of Triton X-100 (Sigma-Aldrich Chemie GmbH, Steinheim, Germany) for 0.5 h. Triton X-100 is a detergent whose application results in the wetting of the entire preparation (except for closed cavities). After this procedure, these specimens were also transferred to glycerine, which was replaced by fresh glycerine several times to completely remove the remaining Triton X-100. The mounting was performed as described by Michels & Büntzow (2010). The glycerine-coated specimens were transferred to the object slide inside a pipette to avoid the creation of new air bubbles. Subsequently, additional glycerine and a high-precision cover slip (thickness = 0.170 ± 0.005 mm, refractive index = 1.52550 ± 0.00015; Carl Zeiss MicroImaging GmbH, Göttingen, Germany) were placed on each specimen. For the combined visualization of resilin and Congo red stained chitin, a fresh specimen of Temora longicornis was stained and mounted as described by Michels & Büntzow (2010).
Wide-field fluorescence microscopy imaging
All relevant WFM imaging parameters described in the following are listed for each of the specimens in Table 1. The specimens were visualized on an upright microscope (Zeiss Axio Imager.M1m, Carl Zeiss MicroImaging GmbH) using, depending on the size of the respective specimens, a 10× (Zeiss Plan-Apochromat, air immersion, numerical aperture = 0.45), a 20× (Zeiss Plan-Apochromat, air immersion, numerical aperture = 0.8), a 25×[Zeiss Plan-Apochromat, multi-immersion (oil, glycerine, water) lens, used with oil immersion (immersion oil from Carl Zeiss MicroImaging GmbH, refractive index = 1.518), numerical aperture = 0.8], or a 40× lens (Zeiss Plan-Apochromat, oil immersion, numerical aperture = 1.3). A metal halide lamp (HXP 120, Carl Zeiss MicroImaging GmbH) was used as excitation light source. For the visualization of the resilin autofluorescence, we applied a DAPI filter set with a bandpass excitation filter transmitting 321–378 nm and a bandpass emission filter transmitting 420–470 nm. The excitation band of 321–378 nm is comparable to excitation bands used in earlier studies having successfully analysed the resilin autofluorescence (Gorb, 1999; Haas et al., 2000; Neff et al., 2000; Niederegger & Gorb, 2003; Perez Goodwyn et al., 2006; Burrows, 2009). However, compared to these excitation bands, the excitation band used in this study has the advantage to be located more closely to the excitation maximum of the resilin autofluorescence (around 320 nm; Andersen, 1963). Similarly, the detected emission band of 420–470 nm is in the range of an emission band (413–483 nm; Burrows, 2009) and includes emission wavelengths (420 nm and 425 nm; Gorb, 1999; Haas et al., 2000; Niederegger & Gorb, 2003; Perez Goodwyn et al., 2006) detected in these earlier studies [with the exception of the study of Neff et al. (2000), who detected wavelengths ≥420 nm]. Furthermore, it is located very closely to the emission maximum of the resilin autofluorescence (around 415 nm; Andersen, 1963). Accordingly, the DAPI filter set used is very appropriate for visualizing the resilin autofluorescence.
Table 1. Overview of the embedding media, microscope lenses, immersions, excitation and emission wavelengths and camera settings used for the WFM visualization of the different specimens.
The micrographs of the resilin autofluorescence were overlaid on bright-field micrographs to exactly show the location of the resilin within the specimen. For taking these bright-field micrographs, the lamp HAL 100 of the microscope was used. In order to visualize additional autofluorescences in the analysed specimens to get information about differences in the general material composition of the structures of interest, we also used other available filter sets (filter set 1: excitation = 450–490 nm, emission ≥515 nm; filter set 2: excitation = 532.5–557.5 nm, emission = 570–640 nm).
All micrographs were taken using the camera Zeiss AxioCam MRm and the Zeiss Efficient Navigation (ZEN) software (Carl Zeiss MicroImaging GmbH). The intensities of the lamps and the exposure times were chosen so that the saturation of the respective micrograph was optimal without getting any oversaturation. The aperture diaphragms of both the transmitted bright-field illumination and the epi-fluorescence illumination were set to relatively small values to obtain micrographs with a high depth of field and contrast while still having a sufficiently high resolution. The luminous-field diaphragms of both light paths were adjusted to allow for the best possible illumination of the visualized sections of the specimens. On the final micrographs, the resilin autofluorescence is shown in blue, and the autofluorescences detected with the filter sets 1 and 2 are shown in green and red, respectively. Overlays of the fluorescence and the bright-field micrographs were created in the ZEN software. The final micrographs were adjusted for tonal values, contrast and brightness using the software packages Nikon Capture NX 2 (Nikon Corporation, Tokyo, Japan) and Adobe Photoshop CS4 (Adobe Systems, San José, CA, U.S.A.).
The position of the excitation spectrum of the resilin autofluorescence is known to be pH dependent with a shift to shorter wavelengths at lower pH values (Andersen, 1963). As glycerine, which is a very appropriate embedding medium for fluorescence microscopy because of its refractive index (1.4735–1.475) being close to that of cover glasses and immersion oils (see above) and its extremely weak autofluorescence, has a pH value of 5.5, its application might have a negative influence on the analysis of the resilin autofluorescence. To test this, we immersed a fresh leg of E. tenax in phosphate buffered saline (PBS) buffer (0.1 M, pH = 7.4; Carl Roth GmbH & Co. KG) for 5 h, then mounted it within PBS buffer and visualized the resilin autofluorescence as described above. For all following visualizations, the intensity of the excitation light source and the settings of the aperture and luminous-field diaphragms were kept identical. Subsequently, we immersed the same specimen in glycerine for 5 h, mounted it within glycerine and visualized the autofluorescence using the same exposure time as with PBS. In addition, we took micrographs with an optimized exposure time. This procedure was repeated a second time using distilled water (pH = 7) and both the optimized exposure time for PBS and that for glycerine.
Confocal laser scanning microscopy imaging
All relevant CLSM imaging parameters described in the following are listed for each of the specimens in Table 2. The specimens were visualized with the confocal laser scanning microscope Zeiss LSM 700 (Carl Zeiss MicroImaging GmbH) equipped with an upright microscope (Zeiss Axio Imager.M1m) and four stable solid-state lasers [laser lines: 405 nm, 5 mW (at fibre end); 488 nm, 10 mW; 555 nm, 10 mW; 639 nm, 5 mW]. The resilin autofluorescence was visualized using the 405 nm laser line and a bandpass emission filter transmitting 420–480 nm, the latter being comparable to the emission wavelengths detected in the framework of the WFM. To visualize additional autofluorescences being comparable to those detected with the filter sets 1 and 2 with WFM, the 488 and 555 nm laser lines were used in combination with longpass emission filters transmitting light with wavelengths ≥490 nm and ≥560 nm, respectively. In addition, the 639 nm laser line was applied to visualize further autofluorescence. For this, we also used the longpass emission filter ≥560 nm, which is possible because the main beam splitter of the Zeiss LSM 700 has an extremely efficient laser line suppression. The Congo red fluorescence of the stained copepod was excited and detected using the 555 nm laser line and the longpass emission filter ≥560 nm. It was visualized in combination with autofluorescences of the exoskeleton excited with the 405 nm, 488 nm and 555 nm laser lines and detected as described above.
Table 2. Overview of the embedding media, microscope lenses, immersions, types of fluorescence and CLSM settings used for the CLSM visualization of the different specimens.
All different fluorescences were excited and detected sequentially. For the CLSM approach we applied the same lenses as for the WFM approach. With one exception, the laser power was set to values between 2% and 30%, depending on the intensity of the respective autofluorescence. Only for the visualization of the cercus of A. domesticus one laser had to be applied with a laser power of 40% because of the low intensity of the respective autofluorescence. The detector gain was manually adjusted prior to image stack collection as described by Michels (2007). For most of the visualizations digital gain and digital offset were set to 1 and 0, respectively, which are the default values of ZEN and yielded the best results. Only in the case of the neck membrane of L. depressa and the cercus of A. domesticus these parameters had to be adjusted because of the low intensity and the relatively pronounced bleaching of the respective autofluorescences. Except for two visualizations, the pinhole size was always set to one Airy. In the case of the prealar arm of L. migratoria and a mechanoreceptor on the dorsal side of the E. tenax pretarsus, optical sections through the specimens were collected. For this purpose, the pinhole size was chosen in a way so that for each of the different autofluorescences the thickness of the respective optical section was identical. Accordingly, the Airy values differed slightly depending on the detected wavelengths, but all values were around one Airy. For each image stack, overlapping optical slices were visualized for the entire thickness of the specimen. In this context, we used the CLSM software ZEN to automatically determine the necessary ideal distance between the centres of two focal planes and the optimal image size (limited by the CLSM system to a maximum of 2048 × 2048 pixel) according to the Nyquist theorem. All image stacks were collected with a line average of 2. Scan rates were chosen so that the best signal to noise ratio and a reasonable collecting time were obtained. Based on the image stacks, maximum intensity projections were created by means of the ZEN software. In doing so, we chose the following colours to make the results comparable to those obtained by WFM: blue (excitation = 405 nm and emission = 420–480 nm), green (excitation = 488 nm and emission ≥490 nm) and red (for both: excitation = 555 nm and emission ≥560 nm, excitation = 639 nm and emission ≥560 nm). The final micrographs were adjusted for contrast and brightness as described above.
Results and discussion
The WFM images clearly reveal the presence and the distribution of resilin in the visualized specimens (Figs. 1a–c, 2a–c, 3a–c and 4a–c). The results are comparable in quality to those obtained in earlier studies applying the same technique to visualize resilin in exoskeletons of arthropods (e.g. Gorb, 1999; Haas et al., 2000; Neff et al., 2000; Niederegger & Gorb, 2003; Perez Goodwyn et al., 2006; Burrows et al., 2008; Burrows, 2009). However, because of the relatively small depth of field and the large amount of out-of-focus light being collected, these wide-field micrographs are quite blurry and provide no detailed information on the microstructure of the analysed exoskeleton parts. By contrast, the CLSM images show the exoskeleton structure of the respective specimens in great detail, and even smallest structures such as tiny bristles and other surface textures are clearly visible (Figs. 1d, 3d and 4d). The distribution of the blue autofluorescence stimulated by the 405 nm laser line is similar to that of the autofluorescence of resilin visualized with WFM using UV excitation (Figs. 1, 3 and 4). Accordingly, it is very likely that both autofluorescences observed are identical and emitted by resilin.
Our results show that resilin can be visualized using a laser excitation wavelength of 405 nm. Furthermore, the application of additional laser lines enables the detailed CLSM analysis of differences in the material composition of the visualized exoskeleton parts. For example, the different structures and material compositions described for the locust wing hinge (Weis-Fogh, 1960) can be perspicuously differentiated (Fig. 1d): (The following colour descriptions correspond to the colours assigned to the different autofluorescences detected in this study.) (1) the sclerotized cuticle is dominated by red autofluorescence; (2) the tough flexible part, consisting of chitin and smaller amounts of resilin, autofluoresces in blue, green and red (resulting in pink, yellow and green colours within the overlay, depending on the proportions of the respective materials); (3) the lamellar rubber-like part with a higher proportion of resilin has a light blue autofluorescence, and the lamellar structure of this part is clearly revealed by CLSM; (4) the part, which consists of pure resilin, autofluoresces only in blue. Accordingly, this pure resilin part appears dark blue on the final maximum intensity projection. In general, our results indicate that the autofluorescence of strongly sclerotized exoskeleton parts is mainly stimulated by green to red laser light with the emission being mainly in the red part of the light spectrum, whereas less sclerotized structures have a green autofluorescence, which is mainly excited by blue to green laser light. When being exposed to UV or violet laser light in combination with blue laser light, the resilin-dominated structures will autofluoresce light blue (a mixture of blue and a bit of green fluorescence) if they contain some chitin, and in dark blue if they consist of pure resilin. However, our analyses have shown that the latter case is rare, and that in most cases the resilin-dominated structures contain a perceptible amount of chitin.
Two further good examples for the large amount of additional detailed morphological information provided by CLSM compared to WFM are the analyses of the vein joint in the dragonfly hind wing (Fig. 3) and the resilin-bearing spring in the hoverfly pretarsus (Fig. 4). CLSM clearly reveals the gradual transition of the material composition within the vein joint from the centre, consisting nearly of pure resilin, to the strongly sclerotized veins (Fig. 3d). The proportion of chitin within the material gets larger, and the sclerotization degree of the material becomes higher, indicated by a change of the dominant autofluorescence colours from blue via green to orange/red. The distribution of the different material compositions can be seen in great detail providing a good three-dimensional impression of how the veins are connected through the resilin pad. Even the transparency of the resilin becomes evident because of the dorsal vein, located below the resilin pad, slightly shining through the resilin in the centre of the vein joint. In the case of the resilin-bearing spring of the hoverfly pretarsus, tiny sclerotized bristles on the spring's surface are clearly visible on the CLSM images (Fig. 4d and e), whereas these structures cannot be specifically discerned on the WFM images (Fig. 4c). The function of these stiff bristles may be in preventing the flexible resilin surface from adhering to itself (in the case of folding) and to other surfaces.
CLSM offers great advantages over WFM in that it is possible to obtain reliable information about differences in the material composition and to create optical sections. In WFM images, the intensities of the different autofluorescence colours are often related to the thickness of the respective structures as the collected light comes from all layers of the specimen. Accordingly, the intensity of each colour does not always reflect the proportion of the corresponding autofluorescence source within the analysed structure. By contrast, CLSM provides individual optical sections of identical, uniform thickness yielding a clear dependency of the composition of the autofluorescence colours from the material composition. As this also holds true for maximum intensity projections created from stacks of such optical sections, differences in the material composition of the specimens can be precisely visualized covering the whole sample independent of the thickness of the single autofluorescence sources. CLSM images thus allow drawing reliable conclusions on differences in the material composition of the analysed structures. However, one has to keep in mind that only very few material sources of the natural autofluorescences of arthropod exoskeletons have been determined so far. Accordingly, in many cases the proportions of known material sources such as resilin can only be described in relation to proportions of unknown material sources.
The possibility to create optical sections enables the analysis of internal structures and may avoid misinterpretations in cases when the analysed specimens possess preparation artefacts. For example, when preparing the prealar arm it is difficult to completely uncover the resilin structures from any residuals of the adjacent tissues. In the case of our specimens, parts of the ligament (see Weis-Fogh, 1960) and parts of other tissues could not be completely removed from the resilin. The difference in the autofluorescence composition between the sclerotized base and the resilin-dominated part becomes evident on the WFM image, however, because of the green and red autofluorescences of the residuals on top of the resilin-dominated material (indicating larger proportions of chitin and higher degrees of sclerotization in the residuals), some parts of the resilin-dominated material do not appear light blue but slightly green and orange in the overlay image (Fig. 2c). This may lead to the misinterpretation that the material composition might not be dominated by resilin. The optical section obtained by CLSM, however, clearly shows that the respective material composition is strongly dominated by resilin, whereas the green and red autofluorescences are restricted to the surface where the residual adjacent tissue is located (Fig. 2d).
The test of the applicability of glycerine as embedding medium showed that this chemical, despite its pH value of 5.5, is very appropriate for the visualization of resilin. When using WFM and PBS buffer (pH = 7.4) as embedding medium, an exposure time of 5.68 s was necessary to visualize the autofluorescence with an optimal image saturation (Fig. 5b). In case of distilled water (pH = 7), the same exposure time yielded a nearly similar image (Fig. 5c). By contrast, with glycerine as embedding medium only 1.34 s of exposure time were necessary to optimally visualize the autofluorescence (Fig. 5d). Using this exposure time in combination with distilled water resulted in the autofluorescence signal being strongly underexposed on the micrograph (Fig. 5f), whereas a combination of glycerine and the exposure time necessary for the preparation embedded in PBS or distilled water caused a micrograph with a pronounced oversaturation of the autofluorescence in large parts of the resilin-dominated structures (Fig. 5e). This indicates that the unfavourable shift of the excitation maximum, described for acidic conditions (Andersen, 1963), is minor at a pH value of 5.5 and can be equalized by the optical properties of glycerine. As these optical properties are much more suitable than those of distilled water, the whole optical system is even much more effective when using glycerine.
We observed that, compared to other autofluorescences present in the exoskeleton of small arthropods, bleaching is much more pronounced in the autofluorescence of resilin. In structures with high proportions of resilin this bleaching effect was found to be particularly intensive. Accordingly, when visualizing preparations a second time, a higher laser power or detector gain is necessary to obtain the same autofluorescence signal from the resilin-dominated structures. As a result of this effect, the proportion of blue autofluorescence in the non-resilin–dominated parts of the exoskeleton may be unproportionally larger on the final images. This important aspect has to be kept in mind when analysing differences in the material composition of exoskeleton structures.
Besides the bleaching, there is another argument to perform CLSM analyses of the autofluorescence proportions within the structures of interest just once per each preparation in order to get reliable results. In fresh unfixed specimens degradation processes of the tissue may occur during the visualization procedure. This may result in an accumulation of molecular damage pigments, whose autofluorescences often have excitation and emission maxima in the same ranges as the autofluorescence of resilin (Fletcher et al., 1973). Accordingly, these autofluorescences may interfere with the results. Therefore, it is advisable to use fresh specimens and to rely on only the results from the first visualization procedure.
When analysing the presence and distribution of resilin in arthropod exoskeletons using fluorescence microscopy, one has to keep in mind that besides resilin several other proteins exhibit autofluorescences, which are comparable to that of resilin in having excitation maxima in the UV range and emission maxima in the blue range of the light spectrum (e.g. Garcia-Castineiras et al., 1978; Fujimori, 1978; Gast & Lee, 1978). Accordingly, in case of analyses of unknown structures, which are supposed to contain resilin, it is absolutely necessary to apply a combination of different test methods to assure that the observed autofluorescence in fact originates from resilin. In this context, appropriate methods are staining by toluidine blue or methylene blue and analysing the respective structures for extreme deformability, stress-birefringence, their swelling behaviour in aqueous media and their digestibility by proteases like pepsin (e.g. Weis-Fogh, 1960; Andersen & Weis-Fogh, 1964; Thurm, 1964).
To further test the proposed CLSM visualization method, we additionally analysed structures, which were very likely to contain considerable proportions of resilin. Among those was the neck membrane of dragonflies, a flexible cuticle connecting the postcervical (neck) sclerites and very likely allowing for an extremely high mobility of the head (see Gorb, 2000). Its high flexibility and foldability lead to the assumption that this cuticle contains resilin. Our results indicate that resilin is present in relatively high proportions within the material of the neck membrane of Libellula depressa, whereas the neighbouring sclerites contain only low amounts of resilin and are dominated by chitinous and sclerotized material (Fig. 6a). The CLSM images even reveal the microfolds within the pattern of membrane folds usually viewed using a scanning electron microscope.
Legs of several insect species have already been described to possess joints with resilin-dominated structures, for example the tibio-tarsal joint of cockroaches (Neff et al., 2000) and the joints between tarsal segments of cockroaches (Neff et al., 2000), hornets (Gladun et al., 2009) and flies (Niederegger & Gorb, 2003). This study shows that between tarsomere 1 and tarsomere 2 of the leg of Coccinella septempunctata a prominent resilin-dominated pad exists (Fig. 6c). However, a considerable amount of green and some red autofluorescence is also emitted from this pad indicating that the proportion of resilin is lower than that of other resilin-dominated structures.
As resilin is, by contrast to other exoskeleton components, amorphous and extremely transparent, it is imaginable that this protein might be very appropriate for the construction of compound eye lenses. In earlier studies such lenses have been shown to contain resilin (Sannasi, 1970; Jaganathan & Sundara Rajulu, 1979; Dey & Raghuvarman, 1983; Viswanathan & Varadaraj, 1985). Our CLSM analyses yielded a pronounced dominance of blue autofluorescence within the compound eye lenses of the tested insects and crustaceans, which strongly indicates that the lens material contains high proportions of resilin (Fig. 6b and d).
Typical arthropod mechanoreceptors, often found in the exoskeleton of insects, are hair plate sensilla and campaniform sensilla (e.g. Thurm, 1964; Heusslein & Gnatzy, 1987; Keil, 1997). These receptors exhibit so-called joint membranes and cap membranes, which have been described to probably contain resilin because of their chemical, mechanical, optical and staining properties and the presence of blue autofluorescence (Thurm, 1964; Keil, 1997). The CLSM results of our study clearly confirm these data: the bases of the large hairs located on the dorsal side of the pretarsus of Eristalis tenax are each surrounded by a resilin-dominated membranous structure, which may facilitate the mobility of the respective hair shaft (Fig. 7a–c). In this context, the complete distribution of resilin within each hair plate sensillum can only be revealed by optical sections (Fig. 7d and e). Furthermore, our results show that the bases and sockets of the cercal filiform hairs together with the associated campaniform sensilla, located on the cerci of crickets (see Heusslein & Gnatzy, 1987; Heußlein et al., 2009), are embedded in a structure, which mainly consists of resilin, and that the campaniform sensilla themselves possess resilin-dominated parts (Fig. 7f).
As 405 nm is at the upper edge of the excitation spectrum of the resilin autofluorescence, the question may arise whether the blue autofluorescence observed during excitation with 405 nm laser light originates only from resilin or also partially comes from chitinous material and other exoskeleton constituents. The possible origin of the autofluorescence from chitinous materials can be tested by treating the exoskeleton structures of interest, including the resilin-dominated parts, with the fluorescence dye Congo red, which selectively stains the chitin in the exoskeleton of arthropods (see Michels & Büntzow, 2010). We have observed that in many calanoid copepods the autofluorescence emitted by larger parts of the ventral side of the cephalothorax is strongly dominated by blue autofluorescence. Accordingly, these exoskeleton parts seem to contain large proportions of resilin and be rather flexible. This is verified by Congo red staining: although the lateral and dorsal parts of the cephalothorax are intensively stained indicating high proportions of chitin in the exoskeleton, the fluorescence of the ventral parts described earlier still exhibits a strong dominance of the blue autofluorescence (Fig. 8). Accordingly, this autofluorescence does not origin from chitin.
This study has revealed the great potential of the described CLSM method for detailed three-dimensional visualization of resilin in the exoskeleton of arthropods. In addition, this method has proved to be an excellent tool to accurately analyse differences in the material composition within the exoskeleton structures of interest. Compared to WFM, CLSM may provide new information about the micromorphology and material composition of resilin-dominated structures in arthropod exoskeletons leading to new insights into the functional morphology and biomechanics of arthropods.
This study was supported by the virtual institute ‘PlanktonTech’ of the Helmholtz Society. Esther Appel helped with the collecting of specimens and the species determination. Anneke Michels made the linguistic revision of the manuscript. The permit for collecting dragonflies was issued to Esther Appel by the ‘Landesamt für Landwirtschaft, Umwelt und ländliche Räume’, Schleswig-Holstein, Germany. We strongly appreciate the work of three anonymous referees and their valuable comments and suggestions for improving the manuscript.