Retinal Damage and Sensitivity Loss of A Light-Sensitive Crustacean Compound Eye (Cirolana Borealis): Electron Microscopy and Electrophysiology

The compound eyes of some animals are adapted to dim light. These scotopic eyes have lenses with large acceptance angles (Land, 1981; Nilsson & Nilsson, 1981; Stavenga, 1979) and may contain reflecting layers (tapeta) in crustaceans (for review, see Hallberg, 1978) and other invertebrates (Land, 1972, 1981; Miller, 1979). A tapetum is considered to be an adaptation to ensure an efficient photon capture (Snyder, 1977; Snyder, Laughlin & Stavenga, 1977). In some deep-sea crustaceans the area of the photoreceptor membranes is enlarged, and thus the rhabdoms appear hypertrophied (Elofsson & Hallberg, 1977; Hallberg, Nilsson & Elofsson, 1980; Meyer-Rochow, 1981; Nilsson, 1982, 1983). These morphological adaptations also include an eye sensitivity that can cope with the dim light (Donner, 1971; Lindstrom & Nilsson, 1983).

The deep-sea-adapted eyes have been shown to be vulnerable to excess light. Even moderate amounts of natural daylight or artificial light cause the photoreceptor membranes to break down (Loew, 1976; Meyer-Rochow, 1981; Nilsson, 1982; Tuurala & Lehtinen, 1971). A similar photo-induced destruction has also been observed in Limulus (Behrens & Krebs, 1976), in visual mutants of Drosophila (Cosens & Perry, 1972; Harris & Stark, 1977), and Procambarus (Kong & Goldsmith, 1977), and in a number of vertebrates (for references, see Williams & Baker, 1980).

A correlation between the morphology of the retina and the sensitivity of the eye has been established in the vertebrate eye (Kuwabara, 1970; Lawwill, 1973a,b; Noell, Walker, Kang & Berman, 1966). The present investigation was undertaken to determine the effects of measured light exposures upon morphology and sensitivity of the eye of the deep-water-living crustacean Cirolana borealis. The results show that there is a correlation between these effects and that normal eye function is abolished by too strong illumination.

Cirolana borealis Lilljeborg (Crustacea: Isopoda) were obtained at night in baited traps from approximately 100-m depth in Raunefjord, south of Bergen, Norway, and sorted in less than 10 min under red safety light. They were shipped by air to the Tvarminne Zoological Station, Finland, where the experiments were performed. The well-fed animals (by the bait) were transferred to 500ml glass beakers (Jena), filled with approximately 3 cm of water from the sampling station, and with two or three animals in each beaker. These were kept in a 3 °C cooling chamber in a wide light-tight box in a coordinate system, allowing for easy identification of the beakers in the dark. The animals were maintained without food. All handling of the animals took place in the dark using infrared (i.r.) light and i.r. image converters (Find-R-Scope, FJW Industries). The animals were dark-adapted 1-7 days prior to the experiments, and during the experiments the animals were exposed to the light intensities presented in Table 1.

After light exposure one batch of animals (N = 50) was used for histological fixation of the eyes, for light and electron microscopy; and another batch (TV =53), for electrophysiological measurements (electroretinogram, ERG).

No light exposure.

(0·47 W m⁻²). The beaker containing the animals was exposed to ordinary room light (illuminance 117 lx). The light was calibrated (μWcm⁻²) by using interference filters as for the ERG measurements (see below) and a UVM-8LX Luxmeter calibrated in absolute units by Airam laboratories for a wavelength of 564 nm (Donner & Lindström, 1980). Knowing the filter parameters, an integration of the flux between 393 and 673 nm and in steps of about 20 nm could be done (Fig. 1). Exposures took place at 10 °C.

(4·9 and 9·8 W m⁻²). The beaker containing the animals stood on a piece of white Styrox and was centred under a vertically mounted Osram 6 V 15 W (79152) microscope lamp with housing. The diameter of the white light patch was adjusted to give uniform light over the bottom of the beaker and with illuminances of 1250 lx (expt 3) and 2500 lx (expt 4) at the level of the Cirolana eyes, measured with standard Luxmeter (Dr Bruno Lange, Berlin). The exposure light was calibrated as in expt 2 (Fig. 1). These exposures took place in a cooling chamber at 3 ° C.

Three daylight exposures with different intensities were performed: (1) Sunny day with clouds, for 4h, with an energy content of 1 ·4 × 10⁶ J m⁻² (97 W m⁻²). (2) Sunny day without clouds, for 10 min, with an energy content of 4·2 × 10⁴ J m⁻² (70 W m⁻²). (3) Sunny day without clouds, for 2h with an energy content of 9·8 × 10⁵ J m⁻² (136Wm⁻²).

The beaker containing the animals was kept at less than 10 °C in an ice bath outdoors. There was no shading of the beaker. The energy content of the exposures was measured with a solarimeter (Kipp & Zonen Solarimeter integrator CCI) positioned next to the beakers. The irradiance values are calculated averages from the exposure times and the energy contents.

Dark-adapted eyes were dissected in the fixative (i.r. light) and light-exposed eyes in red safety light (expts 2–5, Table 1). Fixation for light and electron microscopy was performed in a mixture of paraformaldehyde/glutaraldehyde according to Karnovsky (1965) with Ca²⁺ omitted. Fixation time was 3 h. Post-fixation was carried out in 2% OsO4 for 2h at 4 °C, and 0·2M-cacodylate buffer, pH 7·2, was used throughout. Further details of the histological procedure are given by Nilsson (1982) and Richardson, Jarret & Finke (1960).

Following light exposure, animals were handled in the dark using i.r. image converters (Lindstrom & Nilsson, 1983). Each excised Cirolana head was pinned to a piece of cotton, immersed in sea water from the sampling station and placed in the experimental chamber. The preparations were cooled to about 10 °C and the temperature was continuously recorded with a thermocouple.

Flashes of monochromatic light (duration 600 ms; HW_max ∼ 10 nm), focused on the eye, evoked eye potentials, which were recorded through a glass micropipette (tip diameter 10 μm and filled with 1 M-NaCl) on a dual-beam oscilloscope, using a.c. coupling. The interval between the flashes was 2 min. A readily detectable and repeatable on-response of 5 pN at 495 nm was used as criterion response in the sensitivity threshold experiments (see also Lindstrom & Nilsson, 1983). The intensity of the stimulus was adjusted using neutral density filters and a neutral density wedge. The stimulus and recording system is described in detail by Donner (1971) and by M. Lindström (1983, in preparation). The preparations were allowed to accommodate for approximately 30 min before the first stimulus light was presented.

The following description of the anatomy of the dark-adapted (DA) compound eye of Cirolana borealis concentrates on some details of the structure of the dark-adapted retinula cells (nuclear region) which have a bearing on the results of different light exposures to the eye. A more detailed description of the morphology of the Cirolana eye is given by Nilsson (1982, 1983).

The rhabdom was formed by the rhabdomeres of seven retinula cells, and was fused distally, and open proximally (Fig. 2). Each rhabdomere consisted of several cytoplasmic folds bearing microvilli (microvillar lamellae). These folds originated from the retinula cells and projected into the rhabdom space. The lamellae were often multiforked (Fig. 2). The retinula cell cytoplasm outside the rhabdom contained electron-dense pigment granules (Fig. 3).

The orderly arranged microvilli of the rhabdom were aligned parallel to each other and were 2·5–3 gm long and 60–100 nm wide. Each villus originated separately from a short and thin neck (length and diameter about 25 nm) (Fig. 4). This arrangement could be observed as closed microvillar bases, due to a lack in perfect alignment of the section along the necks of the microvilli. A close packing along the circumference of the microvilli appeared as an almost ‘tight junction’-like structure (Fig. 5), which restricted the extracellular space to thin channels (less than 4–6 nm wide) in between the corners of the ‘hexagonally’ packed microvilli (Fig. 6). An electron-dense extra cellular deposit was present at each microvillar tip (Fig. 3, inset).

Large (150–250 nm diameter) and small (25–60 nm diameter) vesicles occurred in small numbers in the cytoplasm just beneath the microvilli. Depending on the plane of section, these could be seen connected to the microvilli. These vesicles had a double membrane and their abundance and formation will be considered further in the description of the light-exposed eyes (Fig. 4). Occasionally, small (approximately 20–30nm) coated vesicles were found (Fig. 4).

The Golgi complexes (Fig. 7, inset) were mostly found in the soma region and were fairly frequent. Small fragments of rough endoplasmic reticulum (RER) were dispersed in the retinula cell cytoplasm. The smooth endoplasmic reticulum (SER) was thin and consisted of a large number of membrane-delimited sacs and tubules (Fig. 7). Only a few free ribosomes were present in the cytoplasm. No lysosome-like bodies were found. A number of irregularly shaped vesicles were present in the cytoplasm and these are interpreted as distended cistemae of the smooth endoplasmic reticulum (Fig. 7). The mitochondria were well preserved.

Eyes were fixed immediately after exposures to different intensities of white light (Table 1). The anatomy of the retinula cells was altered compared with the totally dark-adapted eye (DA). The changes mostly involved the microvilli, and some of the organelles, such as the Golgi complex, endoplasmic reticulum, mitochondria and pigment granules, and are summarized in Table 2. Unless otherwise stated, the described differences are the same for different exposure times at the same light intensities. Rarely, lysosome-like bodies were observed. Further observations are presented below.

At 117 lx (expt 2) the microvilli maintained a close packing and an orderly pattern, although they were uneven in thickness (20–200 nm) and shorter (1–2 μm) than those of the DA eye (Figs 3,8).

In contrast to the DA eye, a large number of various sized double-membraned vesicles were present at the microvillar bases. Some of these were continuous with the microvilli and some were free in the cytoplasm (Fig. 8). They were also characterized by a fuzzy coat on both sides of the membrane (Fig. 10).

The Golgi complexes were increased both in size and number. Changes in structure included fragmentation and hypertrophy (Fig. 12). A number of retinula cell pigment granules migrated into the rhabdom area and spread into the cytoplasmic folds (Fig. 8).

As stimulus intensity was increased, there was increased disorder of the microvillar lamellae and the microvilli (Figs 9, 13). Daylight exposures caused the microvilli to disrupt (Fig. 15) and frequently, membrane whorls were seen at the microvillar bases (Fig. 11). Pigment granules were found throughout the rhabdom (Fig. 15).

Some animals that were exposed to light according to Table 1 (expts 2–5) were post dark-adapted to determine if the morphology could be returned to the DA condition, and results are summarized in Table 2. Additional comments are given below.

After 117 lx exposure followed by 12 h of dark adaptation, the structural organization was similar to that in the DA eyes (Fig. 16).

After 1250 and 2500 lx light exposures, dark adaptation resulted in further damage. Damage was more extensive with longer dark adaptation (Figs 14, 17). This also includes an increased number of membrane whorls and disruptions. Two and five days of dark adaptation after 4h of daylight exposure resulted in a retinula cell anatomy that was severely deranged (Figs 18, 19).

The ERG of the DA eye was similar to that observed previously (Lindstrom & Nilsson, 1983). On exposure to a light stimulus, there was a cornea negative on-response with peak amplitudes between 300 and 400 μV at the highest stimulus intensities (Fig. 20, inset A). The μV/logl-function had a sigmoid shape (Fig. 20); and the visual threshold sensitivity (= criterion response), defined as the number of photons required to evoke a 5 μV response at peak wavelength (λ = 495 nm), was 2·1 × 10⁸ quanta cm⁻²s⁻¹ (Fig. 21).

Electroretinograms were immediately recorded after light exposure according to Table 1 and yielded weak responses in all cases except in the 117 lx light exposure. These preparations were also typically hypersensitive to strong stimulus light (10¹³ quanta cm⁻² s⁻¹) and a high stimulus irradiance suppressed the sensitivity and it could no longer be recorded. The results are summarized in Table 3.

Following light exposure, some animals were dark-adapted (DA) for various periods of time prior to ERG recording (Table 1, expt 6). The irradiance necessary to evoke a 5 μV response at peak wavelength, λ = 495 nm, was compared with that for the DA eye, as is shown in Fig. 21 and Table 3. Additional comments are found below.

After a 10-min exposure to 117 lx, followed by 12 h of DA, the sensitivity of the ERG visual threshold was similar to that for the DA eye (Fig. 21). With longer exposure, recovery was incomplete.

Dark adaptations for different times (see Table 1, expt 6) after both 1250 and 2500 lx light exposures resulted in threshold sensitivities that could be correlated with the DA times (Fig. 20), and the results were dispersed around 2 and 2·5 log units, respectively (Fig. 21).

All preparations were characterized by a hypersensitivity to strong light stimuli, which radically depressed sensitivity and it could no longer be recorded during the experiment. No responses of the daylight-exposed eyes could be recorded even after 5 days of dark adaptation.

The compound eye of Cirolana borealis has recently been characterized functionally by the use of the electroretinogram, and this showed that the eyes were well adapted to their dim light environment (Lindström & Nilsson, 1983). It has also been shown thgt daylight exposures to the Cirolana eye alter the morphology of the photoreceptor membranes (Nilsson, 1982). The present results extend these investigations by cprrelating the function and the morphological appearance of the eye after defined light exposures and periods of recovery.

Exposure to light produced changes in retinula cell morphology. The cytoplasmic folds bearing the microvilli became distended, while the microvilli were shortened or sometimes disrupted. Double-membraned vesicles were present near the microvilli, especially after illumination at 117 lx. There was distension of the smooth endoplasmic reticulum and an increase in the number of Golgi complexes. Pigment moved from a position outside the rhabdom into the rhabdom space. Lysosome-like bodies were rarely seen, and are unlikely to be involved in membrane renewal or breakdown processes as is the case in a number of arthropods (Blest, 1980; Eguchi & Waterman, 1967, 1976; Hafner, Hammond-Soltis & Tokarski, 1980; Nemanic, 1975; White, Gifford & Michaud, 1980). At daylight intensities membrane whorls were seen at the microvillar bases.

These morphological changes are interpreted as a retinal dystrophy, for they are reflected in the electroretinogram recordings as a loss and, except after brief exposure to 117 lx, the changes could not be restored to the condition in the dark-adapted eye. The responses saturated at low amplitudes or were totally depressed. The morphological appearance of the retina has also been correlated with function in the light-exposed eye of rat (Dowling & Wald, 1960; Noell et al. 1966), rabbit (Lawwill, 1973a; Kuwabara, 1970) and monkey (Lawwill, 1973b); and also in crayfishes kept in prolonged darkness (Eguchi, 1965).

Disruption of photoreceptor membranes and whorl formation, caused by excess light, is a well-described phenomenon in both arthropods (Behrens & Krebs, 1976; Blest & Day, 1978; Loew, 1976; Meyer-Rochow, 1981; Nilsson, 1982; Tuurala & Lehtinen, 1971) and vertebrates (Kuwabara, 1970; Lawwill, Crockett & Currier, 1980). Similar effects are also caused by exposure of the eyes to prolonged darkness (Bloom & Atwood, 1981; Edwards, 1969; Eguchi & Waterman, 1966; 1979; Roach & Wiersma, 1974; Welsch, 1977), by a vitamin A-deficient diet (Carlson, Gemne & Robbins, 1969; Dowling & Gibbons, 1961; Yang, Hollenberg & Wyse, 1978), or by temperature (Meyer-Rochow & Tiang, 1979). Membrane changes can be caused by inadequate fixation procedures, as previously observed (Kabuta, Tominaga & Kuwabara, 1968; Williams, 1980; White & Michaud, 1981) but the technique employed in the present study has been found to be satisfactory (Nilsson, 1982, 1983).

Cell damage was indicated by the distended ER (David, 1978; Trump, Jesudason & Jones, 1978). This was found only in light-exposed eyes and was located in the nuclear region of the cell, so cannot be regarded as subrhabdomeric cisternae (insects: reviewed, by Autrum, 1981; crustaceans: Eguchi & Waterman, 1967; Nâssel & Waterman, 1979; Limulus:Behrens & Krebs, 1976).

At 117 lx, the increased number of Golgi complexes, and the associated large vesicles, may indicate a normal role in the membrane turnover process. At higher illumination, these features probably reflect disordered function (Ontell, 1975).

The degeneration produced by exposure to light above the damage threshold continued in the dark. A similar situation is also found in other crustacean eyes (Loew, 1976; Meyer-Rochow, 1981; Meyer-Rochow & Tiang, 1981).

Within the range of illumination in which the eye is not injured we found that even very short exposure times (10 min) needed 12 h of recovery to reach the values of the dark-adapted eye. This type of eye apparently has a very slow metabolism. This conclusion is supported by the finding that the regeneration rate of visual pigment is very slow in the deep-water-frequenting crustaceanNephrops (Loew, 1976). Further, the presence of light is not necessarily important (at least within a rather long period of time) to maintain normal function of the eye. This is evidenced by the fact that even after 2 months in total darkness the eye responds electrophysiologically quite normally when compared with dark-adapted eyes (Lindström & Nilsson, 1983).

A characteristic type of double-membraned vesicles was found to occur in association with, or close to, the microvilli. They were infrequent in the dark-adapted eyes, but increased appreciably in eyes exposed to light, both below and above the damage threshold levels. A thorough analysis of the appearance of the vesicles, their morphology and connections with the microvilli revealed a new way, previously undescribed by which retinula cells disintegrate microvilli. The process is thought to occur in several steps, which are illustrated in Fig. 22.

The first step involves a bulging in of the membranes of two or more contiguous microvilli (1), and a new neck of the shortened microvillus is formed (2). A similar invagination of the microvillar membrane occurs basally (3). The next step involves, hypothetically, a closure of the membranes involved (this closure seems to occur basally first) and the double-membraned residual vesicle is formed (4). The original basal extracellular space is released as a coated single-membraned vesicle (5). The cytoplasm of the original microvilli ends joins that of the retinula cell (6). Various intermediate steps in the process are seen in Figs 4, 9, 10 and 14.

This mechanism could occur also in other species. Double-membraned vesicles seem to be present in the Limulus eye (Behrens & Krebs, 1976). The coated vesicles, which are normally associated with membrane turnover and breakdown (Blest, 1980; Blest, Kao & Powell, 1978; Eguchi & Waterman, 1976; Stowe, 1980; White, 1967, 1968), seem to play a minor role, at least in the breakdown process, in the Cirolana eye, although these vesicles are still present.

Nothing is known at present of the future fate of the degraded microvillar membranes or how the synthesis of new photoreceptor membranes are accomplished.

In conclusion, our investigation has shown that the Cirolana eye is functionally adapted to low light intensities. This adaptation has evolved so far that exposure to moderate to high illumination is fatal to the eye and presumably also to the animal. The effects of a harmful initial exposure are continued in the dark and recovery is extremely slow. The described mechanism for microvillar membrane destruction is new among arthropods.

This investigation has been supported by a grant from the Swedish Natural Science Research Council (Grant No. 2760-103); by a 2-month scholarship, 1980-81, to one of the authors (Nilsson) from the Nordic Council for Marine Biology; by the Swedish Medical Research Council (Planning Group for Peripheral Visual Physiology, 14G-3000); and by the Maj and Tor Nessling Foundation to one of us (Lindström). Our thanks are due to the staffs of the Marine Biological Station in Espegrend (Norway) and the Tvärminne Zoological Station (Finland). The skilled technical assistance of Miss Lina Hansén, Miss Inger Norling and Mrs Rita Wallén is gratefully acknowledged. And finally we would like to express our special appreciation to Professors Rolf Elofsson and Kai Otto Donner for their interest and encouragement throughout the investigation.