Confocal laser scanning microscopy: using cuticular autofluorescence for high resolution morphological imaging in small crustaceans

Authors


Jan Michels Tel.: +49 471 4831-2009; Fax: +49 471 4831-1918; e-mail: Jan.Michels@awi.de

Summary

The utility of cuticular autofluorescence for the visualization of copepod morphology by means of confocal laser scanning microscopy (CLSM) was examined. Resulting maximum intensity projections give very accurate information on morphology and show even diminutive structures such as small setae in detail. Furthermore, CLSM enables recognition of internal structures and differences in material composition. Optical sections in all layers and along all axes of the specimens can be obtained by CLSM. The facile and rapid preparation method bears no risk of artefacts or damage occurring to the preparations and the visualized specimens can be used for later analyses allowing for the investigation of irreplaceable type specimens or parts of them. These features make CLSM a very effective tool for both taxonomical and ecological studies in small crustaceans; however, the maximum thickness of the specimens is limited to a few hundred micrometers. Three-dimensional models based on CLSM image stacks allow observation of the preparations from all angles and can permit, improve and speed up studies on functional morphology. The visualization method described has a strong potential to become a future standard technique in aquatic biology due to its advantages over conventional light microscopy and scanning electron microscopy.

Introduction

Confocal laser scanning microscopy has become an important technique in many scientific disciplines such as biology, medicine, material sciences and chemistry. The main reason for this development is the strong potential of CLSM for a large number of applications providing solutions for a wide variety of scientific problems and questions. In the field of biology CLSM has mainly been used in the context of cell biology, molecular biology, immunobiology and neurobiology.

To date only a few studies have shown the potential of CLSM for the visualization of arthropod morphology. The successful and efficient application of cuticular autofluorescence has been demonstrated and described for CLSM visualization of insect morphology (Klaus et al., 2003; Schawaroch et al., 2005; Klaus & Schawaroch, 2006). CLSM has been used in combination with fluorescent dyes for investigations of the innervations of copepod antennules (Bundy & Paffenhöfer, 1993), the developmental biology of marine copepods and decapod larvae (Zupo & Buttino, 2001; Buttino et al., 2003) and the morphology of nauplius stages of the copepod species Temora stylifera (Carotenuto, 1999). In three studies CLSM imaging of copepod morphology was applied to corroborate taxonomic descriptions but the method was not described (Galassi, 1997a, b; Galassi & De Laurentiis, 1997). Galassi et al. (1998) employed autofluorescence of the cuticula in an assessment of integumental morphology in copepods by means of CLSM, however, did not describe their method in detail.

The main goal of the present study was to examine the potential of CLSM for morphological imaging in small crustaceans using autofluorescence of the cuticula. For this approach copepods were chosen as model organisms as they possess all attributes required of CLSM preparations: small size, transparency and fluorescence. Furthermore, copepods play a very important role in most aquatic ecosystems and are therefore an essential subject of many scientific studies.

In recent years three-dimensional (3D) modelling of biological structures has become more and more important in order to improve the knowledge of functional morphology (e.g. Hamm et al., 2003). Consequently, the generation of accurate 3D models of the investigated structures is of high interest. Therefore, the creation of 3D models based on CLSM image stacks was tested in the present study.

Material and methods

The visualization of copepod morphology based on autofluorescence was studied using (1) mandibular gnathobases of the calanoid species Heterorhabdus sp. and the cyclopoid species Acanthocyclops mirnyi Borutsky & Vinogradov, 1957, (2) the fifth swimming legs of a male Heterorhabdus sp. and (3) one complete specimen of the recently described harpacticoid copepod species Alteutha potter (Veit-Köhler & Fuentes, 2007).

In order to evaluate the influence of embedding media on the final images three preparations were embedded in euparal (Waldeck GmbH & Co. KG, Division Chroma®, Münster, Germany) while all other preparations were embedded in glycerine jelly. One left mandibular gnathobase of a male Heterorhabdus sp. was visualized by scanning electron microscopy (SEM) using the method described by Michels & Schnack-Schiel (2005).

Preparation of mandibular gnathobases and swimming legs of copepods

Mounting in glycerine jelly.  For the preparations both formaldehyde (Borax buffered 4% formaldehyde/seawater solution) and ethanol (98%) fixed copepods were used. Before the dissections the formaldehyde fixed copepods were soaked in 98% ethanol for 2 h. The copepods were immersed in a droplet of glycerine and the mandibular gnathobases and swimming legs were dissected from the specimens using a stereomicroscope (Leica MZ 16, Leica Microsystems GmbH, Wetzlar, Germany) and fine preparation needles. The gnathobases and swimming legs were transferred into a droplet of glycerine jelly located on a microscope slide and a cover slip was carefully placed on the preparations. Small globules of chewing gum with a diameter of about one millimetre were used as spacers. They were compressed by carefully exerting pressure on the cover slip until the cover slip was located as close as possible to the preparation. In doing so, attention was paid to avoid any contact of the preparation with the cover slip and to keep the microscope slide and the cover slip parallel to each other. Subsequently, the preparations were kept at a temperature of 20°C for 12 h to harden the glycerine jelly.

Mounting in euparal.  A supporting layer of euparal was prepared by hardening a droplet of euparal on a clean microscope slide in a drying oven (60°C) for 2 h to avoid the gnathobases contacting the slide. Subsequently, a copepod was immersed in a fresh droplet of euparal placed on top of the supporting layer. The gnathobases were dissected from the specimen and arranged in an appropriate position for imaging using a stereomicroscope and fine preparation needles. The unused parts of the copepod were discarded. The slide was placed in a drying oven (60°C) overnight to avoid disturbing the position of the preparation when applying the cover slip. After hardening another small droplet of euparal was added and a cover slip was placed on the preparation. The slide was then returned to the drying oven for at least 2 days.

Preparation of a complete copepod specimen

A formaldehyde fixed specimen of Alteutha potter was immersed in lactic acid for 2 h in order to bleach and clear the relatively strongly sclerotisized and pigmented cuticula. The copepod was then soaked in 98% ethanol for 2 h prior to mounting. A thin layer of glycerine jelly was placed on the central area of a clean microscope slide and hardened at 20°C for 2 h before the copepod was mounted on top of this layer in glycerine jelly as described above.

After visualization on the CLSM the specimen was recovered by heating the glycerine jelly to 60°C (ca. 15 min), taking it out of the jelly and rinsing it in 98% ethanol. Subsequently, the same copepod was mounted in euparal using the procedure described above.

CLSM imaging

The mandibular gnathobase of Acanthocyclops mirnyi embedded in euparal was viewed on a Leica TCS NT (Leica Microsystems GmbH, Wetzlar, Germany) equipped with an inverted microscope (Leica DM IRBE) and a krypton-argon laser. An excitation wavelength of 488 nm was used and the emitted fluorescence was detected using the beam splitter RSP 510 and a 515 nm long-pass filter. All other preparations were viewed on a Leica TCS SP5 equipped with an inverted microscope (Leica DMI 6000), one UV laser (diode 50 mW 405 nm) and four visible light lasers (DPSS 10 mW 561 nm; HeNe 2 mW 594 nm; HeNe 10 mW 633 nm; Ar 100 mW 458, 476, 488, 496 and 514 nm). Chromatically corrected lenses were used. For each preparation the most appropriate lens was chosen regarding preparation size and embedding medium in order to apply the highest numerical aperture possible and to keep mismatches of the refraction indices as small as possible resulting in maximum possible quality, intensity and resolution of the detected signal. The most effective excitation wavelengths and the wavelengths of the emitted fluorescence were determined using the entire spectrum of lasers and spectral detectors. In some cases the reflection mode was used in addition to the fluorescence mode to check if any extra morphological information could be obtained. Since the reflection mode never provided any additional information the results were not included in the image processing. The lenses and settings used for the visualization of each preparation are given in Table 1.

Table 1.  Overview of the embedding media, microscope lenses and CLSM settings used for the visualization of the different preparations. ch1, ch2 and ch3 = detection channels 1, 2 and 3 (only mentioned if more than one channel was used).
PreparationEmbedding mediumLens/numerical aperture/immersionExcitation wavelength (nm)Detected emission wavelength (nm)Image format (pixel)
 405 
P5 of Heterorhabdus sp. 20×/0.7/glycerine458ch1: 466–5502048 × 2048
561ch2: 595–717 
Mandibular gnathobase ofGlycerine jelly20×/0.7/glycerine488ch1: 497–5442048 × 2048
Heterorhabdus sp. ch2: 547–682 
Entire Alteutha potter 10×/0.4/glycerine488ch1: 503–5512048 × 2048
561ch2: 574–633 
 
 458ch1: 449–467 (reflection) 
Mandibular gnathobase 20×/0.7/glycerine488ch2: 498–5491024 × 1024
of Heterorhabdus sp. 633ch3: 662–753 
Mandibular gnathobaseEuparal40×/1.25/oil488≥510 nm512 × 512
of Acanthocyclops mirnyi 
 458ch1: 445–470 (reflection) 
Entire Alteutha potter 10×/0.4/air488ch2: 494–5262048 × 2048
633ch3: 650–753 

Laser power was set to 50 percent. Amplitude offset and detector gain were manually adjusted prior to image stack collection so that the best combination of (1) black background, (2) avoidance of black pixels in the preparation and (3) prevention of oversaturation masking structures of interest was achieved. The image stacks were obtained by collecting overlapping optical slices for the entire thickness of the preparation. The necessary ideal distance between two focal planes and the optimal image resolution according to the Nyquist theorem were applied as automatically given by the used CLSM software. Only in case of relatively thick specimens leading to large image stacks was the distance between two focal planes slightly increased in order to reduce the collection time and the amount of data. The images were collected at scan rates and frame averaging settings that produced the best signal/noise ratio within a reasonable collecting time.

3D modelling

Top-down maximum intensity projections were created by means of the 3D projection features of (1) Amira 3.1.1 (Mercury Computer Systems/VSG Group, Düsseldorf, Germany) for the gnathobase of Acanthocyclops mirnyi and (2) the Leica TCS SP5 software for all other preparations. 3D surface-rendered models were created using Amira 3.1.1 on a PC equipped with a 2.86 GHz Pentium 4 processor and 2.56 GB RAM.

Results and discussion

Maximum intensity projections show the morphology of Alteutha potter (Figs. 1a and b), mandibular gnathobases (Figs. 2a and b, Fig. 4a) and swimming legs (Fig. 3) with high accuracy, similar to that obtained by SEM (Fig. 2c). Even rather small structures such as setae of copepod appendages can be visualized (Fig. 3).

Figure 1.

Maximum intensity projections showing a complete specimen of Alteutha potter (ventral view): (a) embedded in euparal, (b) embedded in glycerine jelly. Scale bars = 500 μm.

Figure 2.

Left mandibular gnathobase of a male Heterorhabdus sp.: (a) maximum intensity projection, gnathobase embedded in glycerine jelly, (b) maximum intensity projection, gnathobase embedded in euparal, (c) scanning electron micrograph. T = tubular lumen. Scale bars = 100 μm.

Figure 4.

Left mandibular gnathobase of a female Acanthocyclops mirnyi, embedded in euparal, visualized using different methods (all based on the same CLSM image stack): (a) maximum intensity projection, (b) isosurface model, (c)–(f) smoothed isosurface model shown from different angles. Scale bar = 20 μm.

Figure 3.

Maximum intensity projection showing the fifth swimming legs (P5) of a male Heterorhabdus sp., embedded in glycerine jelly. Scale bar = 100 μm.

One great advantage of CLSM over SEM is the potential to visualize internal structures such as the tubular lumen inside the ventral tooth of the left gnathobases of Heterorhabdus sp. (Fig. 2a and b) described by Nishida & Ohtsuka (1996) using SEM and transmission electron microscopy.

In some cases structures can be distinguished from one another due to the dominance of different types of fluorescence. This phenomenon is particularly pronounced in the gnathobases of Heterorhabdus sp.: the tooth-like structures in the distal part of the gnathobases show mainly green fluorescence whereas the central and proximal parts of the gnathobases are dominated by red fluorescence. Furthermore, the fluorescence of the tooth-like structures is more intense than that of the proximal parts (Fig. 2a and b). These differences in fluorescence very likely reflect different material compositions.

In the present study euparal turned out to be an unfavourable embedding medium for CLSM studies using an excitation wavelength of 488 nm or less. At these wavelengths relatively intense red autofluorescence of euparal could be observed generating an interfering background signal. Manually setting the background to black led to a loss of information and caused a shift towards higher portions of green fluorescence in maximum intensity projections (Figs. 1a and 2b). At longer wavelengths the autofluorescence of euparal decreased, which explains why euparal was successfully used for CLSM investigations of insect morphology at an excitation wavelength of 543 nm without mentioning autofluorescence of the embedding medium (Klaus et al., 2003; Schawaroch et al., 2005; Klaus & Schawaroch, 2006). Contrastingly, only negligible amounts of autofluorescence of glycerine jelly were recorded in the present study. This very weak background signal did not have any influence on the visualization of the preparation. Maximum intensity projections of specimens embedded in glycerine jelly (Figs. 1b, 2a, 3) are much clearer and provide a more accurate display of very tiny structures than those of specimens embedded in euparal. This makes glycerine jelly a very favourable embedding medium for the visualization method described above.

Although the optimal excitation wavelengths determined in the present study differed between the different species, the largest and most important part of the detected autofluorescence was stimulated by laser lines covered by the krypton-argon and argon lasers used (Table 1). These laser types are very common components of CLSM systems making most of the existing CLSMs very suitable for the visualization of the morphology of small crustaceans. For example, the gnathobase of Acanthocyclops mirnyi was visualized just using a single laser line (488 nm) from a krypton-argon laser (Fig. 4a). All the other laser types the Leica TCS SP5 was equipped with played no or only a minor role in the present study. This leads to the assumption that such in part rather expensive lasers will not be necessary in most CLSM studies using cuticular autofluorescence of small crustaceans.

3D models created on the basis of CLSM image stacks show all details of the modelled structures (Fig. 4b–f). They provide the advantage of visualising the morphology of the object from all different angles despite having only one image stack taken from a single angle (Fig. 4c–f). Furthermore, they make precise load simulations and stability calculations possible as already shown for 3D models of diatom frustules (Hamm et al., 2003).

CLSM using cuticular autofluorescence is a powerful method for the visualization of the morphology of small crustaceans. By contrast to the tedious and time-consuming preparation of specimens for SEM, which also carries the risk of artefacts or damage occurring during the drying, mounting and coating processes, the embedding for CLSM is straightforward. The specimens are not affected by the preparation method. This fact makes CLSM an attractive choice for taxonomical studies on type specimens that must not be damaged.

The only disadvantage of CLSM imaging is the maximum thickness of the visualized structures: they must not be thicker than a few hundred micrometers since aberration, scattering and absorption lead to a decrease of fluorescence signal intensity and quality with increasing penetration depth. Accordingly, in the present study while scanning the entire Alteutha potter a very slight decrease in intensity and quality of the fluorescence signal, not evident in the maximum intensity projections, was observed in a few slices furthest away from the excitation laser. This effect can be reduced or negated by scanning the specimens from two opposite sides (e.g. Klaus et al., 2003) and later combining the image stacks. In addition the maximum size of the specimens is also limited by the working distance and the field-of-view of the objective lenses used. For example A. potter (total length about 1.3 millimeters) was the largest structure which could be scanned in total by means of the Leica TCS SP5 using a 10× lens.

CLSM offers many advantages for investigations which require detailed information on morphological characters, including both taxonomical and ecological research: (1) CLSM imaging is much faster (If the cooling of the glycerine jelly is strongly shortened, e.g. by putting the preparation in a refrigerator, a 3D model can be created within 1 h including specimen preparation, CLSM imaging and image processing.) and provides much higher resolution and contrast than line drawings, (2) optical sections in all desired layers and along all necessary axes of the specimens can be obtained and subsequently processed and manipulated with image analysis software, (3) images and 3D models can be stored and distributed digitally and (4) CLSM can facilitate and speed up studies on functional morphology.

In conclusion CLSM must be considered a technique with a great potential for extensive future utilisation for morphological imaging and 3D modelling in small crustaceans. It is possible that CLSM could also be used to visualize and model the morphology of several other groups of aquatic organisms. Consequently, CLSM might become a standard device in many fields of aquatic biology.

Acknowledgements

The author would like to thank S. Liebe, U. Schwarz, W. Goltz and in particular O. Lévai from Leica Microsystems GmbH for their invaluable help concerning the appropriate CLSM application, for making the Leica TCS SP5 available and for taking their time to obtain and process CLSM image stacks. The Acanthocyclops mirnyi and Alteutha potter specimens were provided by K.M. Swadling and V. Fuentes respectively. C. Held and S.B. Schnack-Schiel improved the manuscript due to many fruitful discussions and comments. R. Alheit made the linguistic revision of the manuscript.

Ancillary