INTRODUCTION

As an extremely potent vasoconstrictive principle, endothelin (ET) was first found in medium in which porcine vascular endothelial cells had been cultured (Yanagisawa et al., 1988). Since then, various action of ETs on mammalian tissues have been demonstrated. Following the detection by Uemura et al. (1991) of immunoreactive ET in the plasma of several poikilothermic vertebrates, a considerable number of studies have also reported the presence of ETs as well as their functional roles in poikilothermal vertebrates. For example, Poder et al. (1991) described that ET-1 induced constriction of blood vessels of a few poikilothermal vertebrates, including mesenteric arteries in a species of catfish Ictalurus (=Ameiurus) melas. Working on the rainbow trout Salmo gairdneri (=Oncorhynchus mykiss), Olson et al. (1991) reported that ET-1 induces contraction of cardiovascular tissues. Arad et al. (1992) showed that ET-1 acts on the heart and brain of two fish species (tilapia and torpedo) and a reptile (agama) via specific binding sites on the cells. By means of immunohistochemical tests, Goniakousca-Witalinska et al. (1995) have recently detected the presence of ETs in neuroendocrine cells of gills of the bowfin, Amia calva. It is therefore quite possible that ETs play diverse physiological roles in poikilothermal vertebrates as well as higher vertebrates.

Meanwhile, ETs have been found to influence integumentary chromatophores of various species of teleostean fish: they aggregated melanosomes in dermal melanophores (Fujii et al., 1993; Hayashi et al., 1996), but dispersed light-scattering organelles in leucophores of the medaka, Oryzias latipes (Fujita and Fujii, 1997). Having different optical properties, such reciprocal responses to various stimuli between melanophores and leucophores exist rather normally (Fujii and Oshima, 1986; Fujii, 1993a).

In addition to melanophores and leucophores, chromatophores exist in the skin of teleosts that contain yellowish, reddish and sometimes bluish chromatosomes. These are called xanthophores, erythrophores, and cyanophores, respectively, and they are of course responsible for the yellowish, reddish and bluish components of the integumentary tones (Fujii, 1993a, 2000). Incidentally, cyanophores are novel blue chromatophores that have recently been found in the skin of callionymid fish (Goda and Fujii, 1995). Among them, xanthophores of an oryziatid fish, the medaka Oryzias latipes (Iga, 1969), and erythrophores of the holocentrid squirrelfish Holocentrus ascensionis (Luby-Phelps and Porter, 1982) are known to be rather peculiar chromatophores, in that entry of extracellular Ca²⁺ ions into the cytosol is required for triggering pigment aggregation. Namely, their membranes may be more excitable to depolarizing stimuli, and the physiological characteristics of these brightly-colored chromatophores are probably considerably different from those of other chromatophores elucidated to date. It was therefore interesting to know the influence of ETs on these cells. Since holocentrid squirrelfish are rather difficult to obtain here in the central part of Japan, we decided to examine the effects of ETs on xanthophores of an easily available species.

MATERIALS AND METHODS

RESULTS

Effects of ETs on xanthophores

At the beginning of each series of measurements, we usually confirmed whether the xanthophores to be examined possessed normal responsiveness by applying a K⁺-rich saline (K⁺: 50 mM) solution for a short period of time. When rapid aggregation of xanthosomes took place, the cells were regarded as normal. After ascertaining the redispersion of chromatosomes in primary physiological saline, we examined the effects of ETs.

As a typical series of photomicrographic examinations on the effects of ETs, Figure 1 illustrates the responses of xanthophores on a scale to 10 nM ET-1. The peptide effectively induced the aggregation of xanthosomes (Panels B-E), with a response that was comparable to the maximal one induced by NE (Panel H). In this particular series of photomicrographs incidentally, we can also recognize the responses of an amelanotic melanophore to ET-1 and also to NE: Although the action of ET-1 seemed to be relatively weak, NE aggregated its organelles (amelanotic melanosomes) effectively (Sugimoto et al., 1985). Microscopic observations also showed that both ET-2 and ET-3 were similarly effective in inducing xanthosome aggregation in Oryzias xanthophores (not shown).

Fig. 1

Serial photomicrographs showing the effects of ET-1 on xanthophores in a scale of the medaka (Oryzias latipes, orange-red variety). (A) Equilibrated in physiological saline; xanthosomes are completely dispersed throughout the cytoplasm. (B–G) 1, 2, 5, 10, 30 and 60 min after the application of 10 nM ET-1, respectively; ET-1 effectively aggregated pigment within the xanthophores. It may be noted however, that once the maximal level was reached (E), the effect was gradually reversed (F, G). (H) 10 min after the application of 2.5 μM NE. A large brownish grey figure observable slightly above the middle of each panel is an amelanotic melanophore, which responded to ET-1 and to NE by the aggregation of amelanotic melanosomes. The bar in panel H indicates 50 μm.

We then recorded the motile responses of xanthophores photoelectrically. Panels A, B and C of Figure 2 exhibit typical recordings showing the responses to ET-1, ET-2 and ET-3, respectively. The rate of the xanthophore response to ETs was generally low, when compared with that of the melanophores of several teleosts (Fujii et al., 1993; Hayashi et al., 1996) or with leucophores of the medaka (Fujita and Fujii, 1997). It should also be noted that, even during application of the ET solution, redispersion of xanthosomes took place. It is further noteworthy that once ETs induced the aggregation of pigment, the cells became almost completely irresponsive to the peptides. A slight aggregation of xanthosomes, however, was sometimes detectable in xanthophores, as exemplified in Figure 2C, where ET-3 was applied again after thorough washing of the initially applied peptide. Such a phenomenon may be due to the fact that the first stimulus was rather weak in bringing about the down regulation of receptor of the homologous desensitization type (Hayashi et al., 1996; Fujita and Fujii, 1997). Actually, no such responses were observable when stronger solutions were applied.

Fig. 2

Photoelectric recordings showing the typical effects of ETs on Oryzias xanthophores. In all measurements, K⁺-rich saline (50 mM K⁺) was applied first to confirm that the cells to be examined possessed normal responsiveness. Finally, 2.5 μM NE was applied to arouse the maximal response for reference. (A) 5 nM ET-1 was applied. (B) 5 nM ET-2 was applied. (C) 5 nM ET-3 was applied. In all recordings, remarkable aggregation of pigment, took place in response to ETs, but the responses were rather quickly reversed even in the presence of the peptides. It should also be noted that only slight aggregation of xanthosomes was induced by each peptide when it was applied a second time.

Quantitative studies on the actions of ET-1 and ET-3 were then made, and the relationships between the concentration of the agonists and the magnitude of xanthosome aggregation are exhibited in Figure 3. Before the application of ETs and the agonist to skin specimens, we always checked the motile responsiveness of chromatophores by treating them briefly with the K⁺-rich saline (K⁺: 50 mM) as detailed above. After subsequent washing of the skin specimens with normal physiological saline for more than 30 min, ETs at concentrations from 10 pM to 1 μM (in 10-fold increments) were applied for at least 30 min. The specimen was then washed again with the normal saline to disperse the xanthosomes. Finally, 2.5 μM NE was applied for 30 min to bring about the maximal aggregation of pigment for reference. The degree of xan-thosome aggregation attained during the application of an ET agonist at a certain concentration was compared with that in-duced by 2.5 μM NE. Six measurements were done for each concentration of an ET agonist, and the data were processed to display the concentration-response relationships.

Fig. 3

Relationships of the concentration of ET-1 (solid circles), ET-3 (open circles) and IRL 1620 (open squares) to the magnitude of the pigment-aggregating response of Oryzias xanthophores. Abscissa, negative logarithm of the molar concentration. Ordinate, magnitude of the response as a percentage of the full response induced by 30 min application of 2.5 μM NE. Vertical lines indicate SE.

Both ET-1 and ET-3 induced the aggregation of xan-thosomes in a concentration-dependent manner, and their concentration-response curves resemble each other. Discernible aggregation of Oryzias xanthosomes by these peptides could be observed at concentrations as low as 1 nM, while the maximal level of response was reached at about 10 nM. The concentrations of agonists that induced 50% of the maximal aggregation of pigment (EC₅₀) were 2.90 and 4.31 nM for ET-1 and ET-3, respectively, but there was no statistically significant difference between them. It was also shown that the maximal level attained became lower when ETs at higher concentrations were applied.

As shown in Figure 4, the xanthosome-aggregating response typically proceeded in two steps. An initial rapid aggregation of xanthosomes was quickly followed by their gradual dispersion, and then, a more slower aggregation took place. The pigment dispersing phase may be due to the down regulation of ET receptors mentioned above. Such a phenomenon may be at least partly responsible for lowering the maximal level of response.

Fig. 4

Recording showing a peculiar but rather frequently occurring pattern of xanthophore response to ETs. In this particular recording, 100 nM ET-3 was used. ET-3 induced a rapid but transient aggregation of xanthosomes. Following the brief phase of xanthosome dispersion, they slowly aggregated again, and this was again reversed gradually. Similar phenomena were also observed rather frequently when ET-1 or ET-2 was applied at higher concentrations. As a result of such phenomena, the maximal level of pigment aggregation to ETs became smaller.

Effects of ET_B receptor-specific agonists

The effects on xanthophores of two selective agonists for mammalian ET_B receptors, i.e. IRL 1620 and sarafotoxin S6c, were then examined. Both peptides potently induced the aggregation of xanthosomes, which was followed by their dispersion even in their continued presence (Fig. 5A, B). After removal of the peptides by thorough washing with normal saline, the same peptides were applied again. They sometimes induced a slight aggregation of pigment, but aggregation always failed to occur when it was applied a third time (data not shown). We then examined the concentration-response relationship with IRL 1620 which had an EC₅₀ of 32.7 nM. As shown in Figure 3, its effect seems to be weaker than that of ETs, but the differences among the effects of IRL 1620, ET-1 and ET-3 were not significant statistically.

Fig. 5

Typical recordings showing the effects on xanthophores of two ET_B-selective agonists, IRL 1620 and sarafotoxin S6c (SRTX S6c). (A) 50 nM IRL 1620 was applied. (B) 5 nM SRTX S6c was applied. In each recording, the peptide induced aggregation of xanthosomes, but the response was gradually reversed even in the presence of the peptide. When it was applied again, only slight aggregation was induced.

Effects of ETs on denervated chromatophores

Responses of denervated xanthophores to ET-1 were examined. By observing that leucophores on the same scale were not responsive to K⁺-rich saline (50 mM K⁺), we first confirmed that the xanthophores had been successfully denervated. Figure 6 shows a typical recording in which 10 nM ET-1 was applied, and a profound aggregation of xanthosomes ensued.

Fig. 6

Typical recordings showing the responses to ET-1 of the denervated xanthophores. Before examining the effects of ETs on xanthophores, we confirmed that leucosomes in leucophores on the same scale did not respond to the elevation of the K⁺ concentration in the medium (cf. Materials and Methods section). In this recording, 10 nM ET-1 was then applied and it effectively aggregated xanthosomes in the denervated xanthophores. Analogous recordings were obtained when ET-2 and ET-3 were employed.

Effects of some receptor antagonists

Blockers for some known pigment-aggregating stimuli were then tested for their possible effects. These included an α-adrenolytic, phentolamine, a β-adrenolytic, propranolol, and a cholinolytic agent, atropine. All of them were shown to be ineffective in inhibiting the xanthosome-aggregating action of ETs. As a typical example, Figure 7 exhibits a recording where phentolamine was tested for its effects on the action of ET-1. Phentolamine did not interfere with the action of ET-1, although it effectively blocked the action of NE.

Fig. 7

Typical recording showing the effects of phentolamine, an α-adrenolytic agent, on the responses of xanthophores to ET-1. Phentolamine (Phe) effectively blocked the xanthosome aggregating action of NE, but it had no effect on the response to ET-1.

Effects of ET_A receptor-selective antagonists

The effects of two antagonists specific for mammalian ET_A receptors, i.e. BQ-123 and TTA-386, on the xanthosome-aggregation action of ETs were examined. Panels A and B of Figure 8 illustrate typical recordings showing the effects of BQ-123 and TTA-386 on the action of ET-1, respectively. After treating a scale with either of those blockers for 30 min, ET-1 was applied during the continued presence of the blocker at the same concentration. Even in the presence of sufficiently high concentrations of the blockers, the pigment-aggregating action of ET-1 was not inhibited. The actions of ET-2, ET-3, IRL 1620 and SRTX S6c were also unaffected.

Fig. 8

Typical recordings showing the effects of two mammalian ET_A receptor specific antagonists, BQ-123 and TTA-386, on the responses of xanthophores to ET-1. (A) Effect of 50 nM BQ-123; pretreatment with BQ-123 for 30 min was followed by the application of ET-1 in the presence of BQ-123. (B) Effect of 500 nM TTA-386; after 30 min pretreatment with TTA-386, ET-1 was applied in the presence of TTA-386. In either case, the inhibition was not recognizable despite the fact that a very strong solution of the antagonist was applied.

Effects of an ET_B receptor-selective antagonist

The effects of BQ-788, an antagonist specific for mammalian ET_B receptors, on the action of ETs were finally examined. Figure 9 illustrates a typical recording obtained when ET-1 was used as the stimulant. After treating a scale with 500 nM BQ-788 for 30 min, 5 nM ET-1 was applied in the presence of BQ-788 at the same concentration, but no inhibition was evident.

Fig. 9

Typical recording showing the effects of BQ-788, an antagonist specific for the mammalian ET_B receptor, on the responses of xanthophores to ET-1. Following treatment with 500 nM BQ-788 for 30 min, 5 nM ET-1 was applied in the presence of BQ-788. BQ-788 effectively interfered with the pigment-aggregating action of ET-1.

After removal of the blocker by washing with the primary saline, ET-1 was again applied. Distinct from the case of blockade by ET_A antagonists, ET-1 aroused a fairly large response. It has generally been observed that once a response took place, either in the presence or the absence of an antagonist, ETs or agonists usually failed to induce the response again when they were applied for the second time (Figs. 2, 5 and 8). Such desensitization has already been observed for melanophores (Fujii et al., 1993, Hayashi et al, 1996) and also for leucophores (Fujita and Fujii, 1997), and has been thought to be due to the down regulation of the receptors. Presumably, BQ-788 binds to ET receptors more weakly, and has some inhibitory influence on that down regulation. Fundamentally similar results were obtained on ET-2 and ET-3.

DISCUSSION

When colored organelles (chromatosomes) disperse throughout the cytoplasm of dendritic chromatophores, the skin naturally tends to strengthen the hue of these organelles, whereas the hue fades when they aggregate into the perikarya of the cells. When a fish is adapted to a white background, for example, chromatosomes in the light-absorbing chromatophores (melanophores, erythrophores and xanthophores) aggregate, and those in the light-scattering chromatophores (leucophores) disperse, hence resulting in an effective blanching of the skin. The direction of movement of those chromatosomes is reversed when the animal is adapted to a dark background which causes an effective darkening of the skin. Such cooperative responses observed among various types of chromatophores elicit effective dark-to-pale, or reverse changes of the integumental coloration.

Such reciprocal responses existing between light-absorbing chromatophores and leucophores are also commonly observable when examinations are made in vitro. For example, light-absorbing chromatophores respond by aggregating of chromatosomes in response to sympathetic nervous stimuli, norepinephrine (NE) and to melanin-concentrating hormone (MCH), while leucophores respond to the same stimuli by dispersing their light-scattering organelles, the leucosomes (Oshima et al., 1986; Fujii, 1993a, 2000). When we take account of their reciprocal optical properties, such a reciprocity may rather naturally be understood since such dark-to-pale or reverse changes of color may be of extraordinary importance for survival (Fujii, 1993b).

However, there are several exceptions in the nature of such responses that exist between chromatophores and these belong to two categories. For example, melatonin generally aggregates pigmentary organelles in both kinds of chromatophores, while α-melanophore-stimulating hormone (α-MSH) and adenylyl compounds always disperse them (Fujii and Oshima, 1986; Fujii, 1993a, 2000). Furthermore, some exceptions exist even when we restrict the subject to the lightabsorbing chromatophores. Using a few species of teleosts for example, Oshima and her associates reported recently that prolactin dispersed chromatosomes in erythrophores and xanthophores of some teleostean fish, while it had practically no influence on chromatophores of other sorts (Kitta et al., 1993; Oshima et al., 1996). Nishi and Fujii (1992) and Masagaki and Fujii (1999) described that melanophores in certain regions of the integument of Nannostomus pencilfish respond to melatonin by dispersing their melanosomes which had been thought to aggregate pigment. Such complexity in the chromatic system of fish may not be a strange phenomenon, but must rather be an important feature for generating delicate hues and patterns that fish need to adapt to various environmental conditions and ethological encounters (Fujii, 1993b). Although categorized into light-absorbing chromatophores, xanthophores and erythrophores possess optical properties that stand between those of melanophores and leucophores in a certain sense. Until the present, furthermore, no results have been reported on the effects of ETs on these brightly-colored chromatophores. It was therefore very challenging for us to examine whether ETs aggregate or disperse pigmentary organelles in these chromatophores. In the present study, we have chosen xanthophores of medakas as representative of brightly colored chromatophores.

We found that ETs possess remarkable xanthosome-aggregating action, and the direction of xanthosome displacement was similar to that observed in melanophores (Fujii et al., 1993; Hayashi et al., 1996). Thus, the present observations are in conformity with many past reports that melanophores, xanthophores and erythrophores generally respond to neural or hormonal signals in the same manner in terms of the direction of the pigmentary translocation (Fujii, 1993a, 2000; Fujii and Oshima, 1986, 1994; Matsumoto et al., 1978; Obika and Meyer-Rochow, 1990). On the other hand, we have already shown that ETs disperse light-scattering organelles in leucophores of medakas (Fujita and Fujii, 1997). Namely, ETs may function to enhance the contrast of chromatic patterns on the skin. As mentioned above, such chromatic responses must be helpful for animals, at least under ethological conditions where an increased contrast of chromatic patterns is advantageous.

ETs showed their effects even at very low concentrations, which may indicate that the action of ETs on the cells is not merely a pharmacological one. Namely, at least some molecular form(s) of ET isopeptides may function to augment the whiteness of the skin, modifying the action of sympathetic postganglionic fibers and/or of hormonal factors (Fujii, 1993a, 2000; Fujii and Oshima, 1986, 1994). Very recently however, Visconti et al. (1999) reported that ETs were ineffective in inducing skin lightening or darkening of a freshwater ray, Potamotrygon reticulatus. That is, the melanophores of that species may be unresponsive to ETs. Since these peptides are thought to have definite roles in regulating chromatophores in teleosts, further comparative studies are needed in lower fishes, including elasmobranchs.

Denervated xanthophores responded to ETs quite normally. It was thus concluded that ET isopeptide(s) act directly on xanthophores, not via the stimulation of presynaptic nervous elements to liberate sympathetic neurotransmitter. IRL 1620 and sarafotoxin S6c, both being ET agonists, potently aggregated xanthosomes. Adrenergic blocking agents of the α- and β-types (phentolamine and propranolol), and a cholinolytic type (atropine) were all ineffective. Furthermore, specific blockers of the mammalian ET_A receptor (BQ-123, TTA-386) or for the ET_B receptor (BQ-788) were not effective in blocking the action of ETs on xanthophores. These results indicate that the ET action on Oryzias xanthophores is mediated by ET receptors that do not resemble mammalian ET_A or ET_B.

We now know that the classes Osteichthyes and Mammalia branched off more than 400 million years ago. Since then, receptor proteins in these two groups should have diverged independently. Therefore, even though their roles to bind to ET agonists still resemble each other, molecular structures as well as pharmacological properties of receptors should have altered considerably. It is therefore inappropriate to categorize fish ET receptors into ET_A or ET_B types on the basis of effects of drugs developed for human use, and fish ET receptors will have to be classified more precisely in the future.

Like melanophores (Fujii et al., 1993; Hayashi et al., 1996), Oryzias xanthophores responded to ETs by pigment aggregation, but there are some differences between their characteristics. The first is the slow rate of xanthosome displacement, but that can be ascribed to the intrinsic nature of the cells themselves. In xanthophores, secondly, the level of maximal aggregation was decreased at very high concentrations of agonists (Fig. 3). It is conceivable that xanthophores possess β-ET receptors in addition to α-ET receptors. At higher concentrations of ETs, the xanthosome-dispersing effect via β-ET receptors might become apparent, counteracting the xanthosome aggregating action mediated by α-ET receptors.

We suggest that ET may function as a paracrine factor or a local hormone to modify the actions of nervous or hormonal principles (Fujii et al, 1993; Hayashi et al., 1996; Fujita and Fujii, 1997; cf. also Fujii, 2000). Recently, Yada et al. (1991) showed that in humans, keratinocytes are a source of ETs which can be strong mitogens as well as melanogens for human melanocytes. In fish, tissues or organs that secrete ETs to control chromatophores have not yet been identified, but very recently, the secretion of ET from goldfish epidermal cells in culture was demonstrated (Akimoto et al., 2000). Thus, the possible source of ET regulating chromatophore movement might be sought there. In fact, epidermal melanophores are thickly surrounded by epidermal cells, and these two cell types may easily constitute the paracrine coupling. In fish however, chromatophores are rather infrequently found in the epidermis. Having been thought to have only minor roles in pigmentation, these cells have not yet been studied for the effects of ETs. By contrast, dermal chromatophores are separated from the epidermis by the basal lamina that is thought to be relatively impermeable, and also by rather thick, subepidermal collagenous lamella that belongs to the compact connective tissue (Fujii, 1968). It is therefore unlikely that epidermal cell-dermal chromatophore coupling forms the paracrine link. At this moment therefore, we presume that endothelial cells of blood capillaries that run close to the chromatophores are the source of ET. Capillaries are richly distributed in the dermis, and ETs would be liberated not from the luminal side but rather from the outer side of endothelial cells to reach the chromatophores (Fujita and Fujii, 1997; Fujii, 2000).

It should be noted here that, as in the case of melanophores (Fujii et al., 1993; Hayashi et al., 1996) or leucophores (Fujita and Fujii, 1997), the xanthosome-aggregating effect of ETs was transient, although the time course was longer than that observed on other types of chromatophores examined thus far. It was also found that once a scale had been treated with ETs, the xanthophores became refractory to them. Such refractoriness has also been observed on melanophores and leucophores. As described before, these phenomena might be ascribed to desensitization of receptors involved, or to the so-called “down regulation” (Hayashi et al., 1996; Fujita and Fujii, 1997).

Recent studies on chromatophores showed that the increase or the decrease in cytosolic level of a certain second messenger species always induces reciprocal pigmentary movements, and that the direction of the movement is common disregarding the species of chromatophores (Fujii, 1993a; 2000; Fujii and Oshima, 1994; Nery and Castrucci, 1997). That is, the increase in cyclic AMP levels always mediates dispersion of chromatosomes, while, in cases where Ca²⁺ ions (Luby-Phelps and Porter, 1982; Negishi and Obika, 1985; Oshima et al., 1988; Oshima et al., 1998) and inositol 1,4,5-trisphosphate (Fujii et al., 1991) are working as second messengers, they always induce aggregation of chromatosomes.

On the other hand, we know that adrenergic stimuli, for example, aggregate or disperse chromatosomes via α-adrenoceptors or β-adrenoceptors, respectively. Melatonin (MT) generally aggregates melanosomes, but in some cases (e.g. pencilfish) it disperses pigment in melanophores by novel MT receptors. We named them β-MT receptors, leaving the conventional and common ones to be called α-MT receptors. Similarly, ET receptors may also be categorized tentatively into two groups, i.e. α-ET and β-ET receptors. In accordance with the above cases, the α-ET receptors are assigned for those mediating chromatosome aggregation, while β-ET receptors mediate their dispersion. To date, only leucophores are known to possess β-ET receptors.

In relation to the designation of ET-receptors, the signal transduction mechanism involved in the ET action on chromatophores should be mentioned. Until recently, many researchers working on fish chromatophores have been inclined to consider that cyclic adenosine 3′,5′-monophosphate (cAMP) is the sole second messenger for motile activities of chromatophores. As for hormonal regulation, cAMP may play the major role. Meanwhile, an increase in intracellular levels of Ca²⁺ ions has sometimes been shown to be associated with the aggregation of chromatosomes. Among chromatophores employed in those studies, erythrophores of Holocentrus squirrelfish and xanthophores of medakas are rather peculiar cells, in that they are distinct from many other chromatophores reported hitherto, in that the aggregation of pigment in these brightly-colored chromatophores is triggered by the penetration of extracellular Ca²⁺ into cytosol (Luby-Phelps and Porter, 1982; McNiven and Ward, 1988; Oshima et al., 1998). Thus, these cells may better be treated separately from common chromatophores (Fujii, 1993a, 2000). Working on medaka melanophores however, Negishi and Obika (1985) reported that an increase in cytosolic Ca²⁺ levels is associated with the aggregation of melanosomes. Oshima et al. (1988) then showed that catecholamine stimulation is accompanied by an increase in intracellular Ca²⁺, resulting in the aggregation of pigment in melanophores and erythrophores of the platyfish (Xiphophorus maculatus). Based on their results on xanthophores of medakas, Oshima et al. (1998) suggest that Ca2⁺ depletion in the cytosol may lead to the inhibition of adenylyl cyclase, and may finally result in the aggregation of xanthosomes.

Working on melanophores of the Nile tilapia, Oreochromis niloticus, Fujii et al. (1991) further showed that inositol 1,4,5- trisphosphate (IP₃) may function as another second messenger for signaling adrenergic stimulation in fish chromatophores, indicating the involvement of α₁-adrenoceptors. Namely, both α₁- and α₂-adrenoceptors are present in melanophores being controlled adrenergically. In many tissues, IP₃ has been shown to induce the liberation of Ca²⁺ ions from intracellular storage compartments. In common chromatophores as well, IP₃ may act via the release of Ca²⁺ ions from such compartments within the cells, because vesicular elements of smooth endoplasmic reticulum are abundantly found in the cytoplasm (Fujii, 1966, 1993a).

It has generally been accepted that signal transduction through ET receptors mainly proceeds by activating phospholipase C, which produces IP₃ and diacylglycerol (DG). As for the role played by DG in the motile responses of chromatophores, we still have no information, although those effects, if any, may be related to morphological color changes. As in the case of adrenergic stimulation of melanophores (Fujii et al., 1991), IP₃ produced via the action of ETs may act to aggregate pigment also in xanthophores. It seems likely therefore that a mechanism similar to that involved in the action of ETs on mammalian tissues is also operating in xanthophores of fish. Adenylyl cyclase and phosphatidyl inositol systems are probably operating in concert in the signal transduction of chromatophores to realize their motile responses (Fujii et al., 1991; cf. Fujii, 1993a).