8/11/2019 2000-Fujii The regulation of motile activity in fish chromatophores.pdf
1/20
Copyright Pigment Cell Res 2000PIGMENT CELL RES 13: 300319. 2000Printed in Irelandall rights resered ISSN 0893-5785
Review
The Regulation of Motile Activity in Fish Chromatophores
RYOZO FUJII1
Department of Biomolecular Science, Faculty of Science, Toho Uniersity, Miyama, Funabashi, Chiba 274-8510, Japan
*Address reprint requests to Dr. Ryozo Fujii, 3-22-15, Nakaizumi, Komae, Tokyo 201-0012, Japan. E-mail: [email protected]
Received 3 March 2000; in final form 31 March 2000
times, they are also useful in courtship and mutual communi-Chromatophores, including melanophores, xanthophores, ery-
throphores, leucophores and iridophores, are responsible for cations among individuals of the same species, leading to an
the revelation of integumentary coloration in fish. Recently, increased rate of species survival. Such strategies are realized
by complex mechanisms existing in the endocrine and/orblue chromatophores, also called cyanophores, were added to
nervous systems. Current studies further indicate that somethe list of chromatophores. Many of them are also known to
paracrine factors such as endothelins (ETs) are involved inpossess cellular motility, by which fish are able to change their
these processes. In this review, the elaborate mechanismsintegumentary hues and patterns, thus enabling them to exe-
cute remarkable or subtle chromatic adaptation to environ- regulating chromatophores in these lovely aquatic animals are
described.mental hues and patterns, and to cope with various ethological
encounters. Such physiological color changes are indeed cru-
cial for them to survive, either by protecting themselves from Key words: Melanophore, Erythrophore, Xanthophore, Leu-
cophore, Iridophorepredators or by increasing their chances of feeding. Some-
blue chromatophores in callionymid fish, naming them
cyanophores (5). Thus, six kinds of chromatophores are
now known in poikilothermic vertebrates. Various combina-
tions of these chromatophore species in various proportions
realize various hues in certain regions of the integument,
thus enabling animals to adapt to environmental conditions
for their survival (2).
In order to effect such chromatic strategies, poikilother-
mic animals also make good use of the cellular motile
activities of pigment cells. Namely, the rapid physiological
color changes have elaborately evolved during the long
history of evolution. The colorations and color changes,
thus obtained, constitute critically important strategies to
avoid attack by predators and to obtain prey more easily for
survival. On many occasions, furthermore, delicate and
subtle changes in hues and patterns, thus realized, are usedfor communication with conspecifics. These phenomena are
especially remarkable in bony fish. The extraordinarily so-
INTRODUCTION
We joyfully appreciate beautiful colors and patterns dis-
played by many species of animals. Such integumentary
colors are dependent on the presence of pigment cells in
the skin (13). We know that in homeothermal verte-
brates (mammals and birds), melanocytes producing
melanin are the sole pigment cells responsible for their
coloration. By contrast, various types of pigment cells, as
well as pigmentary substances, are involved in the col-
oration of lower animals that include poikilothermal verte-
brates and invertebrates. These pigment cells have
inclusively been called chromatophores. If we deal solely
with vertebrates, at least five kinds of chromatophores are
present, namely, melanophores (black or brown), xan-
thophores (ocher or yellow), erythrophores (red), leu-
cophores (whitish), and the iridophores (metallic or
iridescent). This nomenclature is now widely accepted,which the present author has also endeavored to establish
for a long time (3, 4). In addition, we recently discovered
Abbreiations ACTH, adrenocorticotropic hormone; ET, endothelin; MC, melanocortin; MCH, melanin-concentrating hormone; MC-R,MC receptor; MSH, melanophore-stimulating hormone; MT, melatonin; MT-R, MT receptor; NAT, N-acetyl transferase; NE, nor-epinephrine; PG, prostaglandin; POMC, proopiomelanocortin; PRL, prolactin; SL, somatolactin1 R. Fujii is now at 3-22-15, Nakaizumi, Komae, Tokyo 201-0012, Japan.
Pigment Cell Res. 13, 2000300
8/11/2019 2000-Fujii The regulation of motile activity in fish chromatophores.pdf
2/20
phisticated properties of their chromatic systems that we
now observe have certainly developed during evolution of
more than 400 million years (2).
The motile activities of chromatophores are dependent on
the intracellular presence of motor-proteins, namely tubulin,
dynein and kinesin. Current understanding about the cellu-
lar motility per se of vertebrate chromatophores has been
reviewed elsewhere (3, 6, 7).
It has been a fairly long time since we have published a
review relevant to the present title (8). On this occasion
therefore, the author tried to outline his views on the current
status of studies concerning the regulation of the motileresponses of fish chromatophores.
REVELATION OF COLORS
In homeothermal vertebrates, melanin-containing organelles
(melanosomes) are synthesized in melanocytes that reside in
the basal layer of the epidermis, and are transferred into
epidermal cells. The darkness of the skin is responsible for
the absorption of light by these recipient cells for the most
part. As a group of poikilotherms, teleosts possess
melanophores as homologues of melanocytes. Like
melanocytes, they are dendritic cells, but extend a number of
cellular projections almost parallel to the plane of the skin.
In teleosts, melanophores are mostly found in the dermis
and are often called dermal melanophores. Sometimes, how-
ever, melanophores are also found in the epidermis, but the
melanosomes are mostly kept confined within the cells, and
aggregate into the perikaryon or disperse throughout the
cytoplasm in response to various signals, as do dermal
melanophores (3).
In many species of fish, melanophores take principal part
in physiological color changes, but there also exist other
kinds of dendritic chromatophores in the skin, i.e. xan-
thophores, erythrophores, cyanophores and leucophores.
Pigmentary organelles contained within them are now called
xanthosomes, erythrosomes, cyanosomes and leucosomes,
respectively (3, 5), and are inclusively called chromatosomes.Excepting for the light-scattering leucosomes, they are light-
absorbing. The melanosomes effectively absorb light rays
within the entire range of visual spectrum, but other chro-
matosomes absorb rays of complementary color to that the
cells exhibit. Leucosomes, by contrast, scatter light rays of
wider wavelengths. Thus, leucophores look whitish when
illuminated by incident light (3).
Although very commonly existing in whitish or silvery
parts of the skin, iridophores are rather peculiar chroma-
tophores, because they are usually non-dendritic and do not
contain colored organelles (3). Instead, stack(s) of transpar-
ent thin crystals of guanine are present in the cytoplasm.
The thin crystals are called reflecting platelets, since they
are strongly light-reflecting owing to their very high refrac-
tive index (of no less than 1.83). Within a stack of them,
higher reflectivity can be achieved as a result of the multiple
thin-film interference phenomenon. As for detailed descrip-
tions about the optics of iridophores, our previous articles
can be referred to (2, 3). In iridophores that are responsible
for silvery glitters and whiteness of side and belly skin, the
platelets are arranged in a stack, to exhibit the multi-layer
thin-film interference phenomenon of the ideal type. Such
iridophores are immotile cells, and are not directly involved
in the physiological color changes.
By contrast, iridophores in some teleostean species have
cellular motility, which plays a predominant role in their
fascinating color changes (3). These iridophores contain
stacks of very thin platelets, and in a given stack, the
distance between platelets is very uniform. Simultaneous
changes in the distance between platelets in a stack result in
changes in the light-reflecting characteristics. Naturally, the
optical treatment of the multi-layer interference system
should be far from that of the ideal system. When thedistance increases, the motile iridophores reflect light of
longer wavelengths. When the spacing between the platelets
decreases, conversely, the spectral peak shifts towards
shorter wavelengths. The former response was designated
the LR response, being an abbreviation of the Longer-
wavelength light-Reflecting response, while the latter one is
called the SR response, an abbreviation for Shorter-wave-
length light-Reflecting response (3, 9). In later sections,
these terms will frequently be employed to describe the
reaction of motile iridophores.
Motile iridophores with dendritic processes have recently
been described in some gobiid fish, including the dark
sleeper goby Odontobutis obscura obscura (10). As with
iridophores of many amphibians, reflecting platelets aggre-
gate into the perikaryon or disperse to dendritic processes in
response to neural or hormonal stimuli (1). When the
platelets aggregate in the perikaryon, the cells appear bluish
in color. However, the same cells look yellowish when the
platelets are dispersed. The bluish tone is considered to be
due to the gradual formation of organized piles of platelets
during their aggregation (11).
Each chromatophore is a small entity, usually containing
a single kind of pigmentary material or stack(s) of light-
reflecting platelets. When differently colored chroma-
tophores are distributed in the skin, the resulting color
appears to be a mixture of different colors. By making good
use of the divisionistic effects, the fish can exhibit a numberof intermediate hues almost at will (2, 3). Although simpler
than those in anuran skin (1), dermal chromatophore units
are found in the skin of colorful specimens such as bluish
damselfish (2, 3).
Under the epidermis of fish, there are wide extracellular
spaces of rather uniform thickness, composed mainly of
collagen fibrils. The dermal chromatophores are usually
present below this compact collagenous layer, and are not in
direct contact with the bottom of the epidermis. Parallel
collagen fibrils form a thin sheet, and several sheets are
arranged as lamellae, but the fibrils within alternating sheets
run approximately perpendicular to those in adjacent ones.
Resembling plywood, the lamellar structure apparently rein-
forces the thin integument, and protects underlying fragile
chromatophores (3, 12). The laminated collagenous struc-
ture can also be assigned another important role since its
architectural features closely resemble those of the stroma
(substantia propria) of the vertebrate cornea. The latter, of
course, is extremely transparent to light, in addition to its
mechanical rigidity. The attained transparency of the struc-
ture overlying the chromatophores must be of great impor-
Pigment Cell Res. 13, 2000 301
8/11/2019 2000-Fujii The regulation of motile activity in fish chromatophores.pdf
3/20
tance for animals in executing effective chromatic responses
(2, 3, 12).
When animals are on land or in the air, the light reflectiv-
ity at the very surface of the body covering may not be less
than 2.4%, while that for animals in the water can be
calculated to be 0.022% (13). Those values were based on
the assumption that the refractive index of the body surface
material is 1.37, the value being adopted from that represen-
tative of the cytoplasm of living cells. Being normally kera-
tinized, squamous or cuticulized, the coverings of terrestrial
animals may have refractive indices higher than 1.37, and
thus, the light reflectivity should be somewhat higher thanthe value given above. The strikingly smaller value for
aquatic animals reflects the fact that the uppermost epider-
mal cells are normally unkeratinized and alive. In this way,
the reflectivity at the body surface of an aquatic animal is
practically negligible. Under such morphological situations,
the colors due to the states of chromatophores are clearly
visible from the outside.
Working on the ice goby, Leucopsarion petersii, Goda and
Fujii (13) reported a special case of the role of
melanophores in the color revelation. As the common name
signifies, even adult specimens of this fish are transparent,
but a small number of melanophores and xanthophores
were found in the skin. In addition, very large melanophores
exist deep inside the body, namely in the peritoneum and
near the vertebrae. They are clearly visible from outside, and
are responsive to various agents. Apparently these
melanophores do not belong to dermal cells, but have
definite roles in the chromatic responses.
ENVIRONMENTAL FACTORS THATDIRECTLY INFLUENCE CHROMATOPHORES
Several physical factors, and sometimes chemical ones, from
the environment affect chromatophores. Most such stimuli
are perceived by sense organs and are brought to the central
nervous system, where the information is processed to yieldappropriate chromatic reactions from the animals. Some
factors, however, directly influence chromatophores. We
have reviewed many of these in a recent article (14), and in
the present article therefore, recent results of interest are
mainly dealt with.
Direct Effects of Light on Chromatophores
Physiological color changes in animals are frequently cate-
gorized into two types (15). One type is the so-called pri-
mary color response, in which chromatophores respond
directly to incident light. The other type is the secondary
color response, in which the chromatophores are controlled
by the nervous and/or endocrine systems. The primary color
responses are mainly observable during the embryonic and
larval stages until the time when chromatophores are not yet
under the control of endocrine and/or nervous systems. It
has often been observed that when chromatophores are
denervated, or when a blinded or a blindfolded fish is
examined, even normal chromatophores respond to light
directly.
Using melanophores from embryos, larvae or young black
platyfish, Xiphophorus maculatus, Wakamatsu (16) reported
that some melanophores in culture responded to light by
aggregating melanosomes, although all the melanophores
were initially light-insensitive. The spectral sensitivity peak
stood at about 410 nm (17). By contrast, melanophores
from larvae of the rose bitterling, Rhodeus ocellatus, re-
sponded to light by dispersing melanosomes, whereas the
melanosomes aggregated in the dark (18). The effective
wavelength of the light was around 420 nm (19). Observing
the responses of melanophores on scales plucked from adult
dark chubs (Zacco temmincki), Iga and Takabatake (20)found that the light dispersed pigment by acting directly on
the cells, although the sensitivity differed among individuals.
Using the melanophores of adult medaka, Oryzias latipes,
that had been cultured for more than 1 day, Negishi (21)
confirmed the direct responsiveness of the melanophores to
light. The most effective wavelength for the induction of
melanosome dispersion in medaka was close to 415 nm,
while melanophores of dark chubs showed a maximum
spectral sensitivity at about 525 nm (22).
Chromatophores other than melanophores have also been
studied for their responsiveness to light: for example, the
leucophores of Oryzias responded to light by dispersing
their light-scattering inclusions (23). Motile iridophores in
the lateral stripes of the neon tetra, Paracheirodon innesi,
show the LR response to light (24, 25). Xanthophores in
adult specimens of medaka were also found to respond to
light by xanthosome aggregation, and the effective wave-
length was around 400 nm (26). While examining the effect
of light on adult Oryziaschromatophores, Oshima et al. (27)
recently found that both innervated and denervated xan-
thophores responded to light (9000 lx) within 30 s by
pigment aggregation, and that the response was not medi-
ated through -adrenoceptors. The maximum spectral sensi-
tivity was about 410420 nm, and the effect was reversible.
Responsiveness was higher in summer than in winter and
Ca2+ ions and calmodulin were not involved in the re-
sponse. Their conclusion was that photoreception by visualpigment that absorbs light at 410420 nm increases phos-
phodiesterase activity, resulting in a decrease in cytosolic
cyclic AMP levels, finally leading to the xanthosome
aggregation.
Using the Nile tilapia, Oreochromis niloticus, Oshima and
Yokozeki (28) recently reported that either innervated or
denervated erythrophores responded directly to light of
defined wavelengths by pigment aggregation or dispersion.
In spectral regions between 400 and 440 nm and also
between 550 and 600 nm, erythrosomes aggregated, whereas
their dispersal was accelerated around 470530 nm. These
results suggest the coexistence of three kinds of visual
pigments in tilapia erythrophores.
Other Physical Factors
Some environmental factors other than light influence chro-
matophores either indirectly or directly, but because of their
relatively low importance, such factors have only rarely been
investigated, and accordingly, data are rather scanty. We
consider, however, some of them to be menaces to fish. For
Pigment Cell Res. 13, 2000302
8/11/2019 2000-Fujii The regulation of motile activity in fish chromatophores.pdf
4/20
example, UV rays may injure pigment cells and impair the
activity of nerve fibers that control cellular responses, espe-
cially for fish living in shallow waters. Hydrostatic pressure
should have influences on deep-sea fish, especially when they
show diurnal vertical migration for feeding. Low tempera-
tures normally reduce cellular motility. Osmolarity and pH
of the water in which they dwell should be other relevant
physical factors; for example, fish that migrate between
inland and sea waters must face drastic changes in osmolar-
ity. If we consider that homeostatic mechanisms are func-
tioning in vivo, the internal milieu around chromatophores
may not be directly influenced by them. When needed, wecan experimentally examine the effects of these factors, and
the results of such studies have actually provided important
knowledge about the physiology of pigment cells (14). Since
many of these factors have recently been reviewed (14), they
will not be further discussed here.
Chemical Factors
That some environmental chemical substances directly affect
chromatophores seems to be unlikely, because, unlike other
cells constituting the body, chromatophores are rigidly pro-
tected from the invasion of chemicals. Being different from
terrestrial animals, where layers of keratinized cells cover
the body, part of the living cell membrane, that directly
faces the environmental watery phase of the outermost
epidermal cells and the occluding junctions between those
cells, functions as a diffusion barrier in fish.
Fish possess various chemosensory organs for feeding and
reproduction (30). The perceived chemical information is
integrated in the central nervous system to arouse certain
ethological responses, as in the cases of other sensations. It
is known, however, that some chemicals, as solutes in the
water surrounding the animals, can be taken up, affecting
the chromatophores directly. The most interesting instance
may be melatonin (MT). Immersing pencilfish (Nannostomus
beckfordi) in aquarium water containing MT, Reed (31) first
observed the phenomenon, and further developed a biologi-cal assay for MT. We have also been able to observe the
effects of MT by immersing fish in MT-containing water for
analysis of circadian chromatic responses, as well as for
characterization of MT-receptors (R) (3234). Owing to its
high lipid solubility, MT can affect the state of chroma-
tophores by invading the body rather easily, probably
through the gill epithelium. By selecting less polarized
molecular species, we may be able to study the effects of
various substances on chromatophores in vivo.
Considering that signaling mechanisms, both in odor
perception by the olfactory epithelium and in chroma-
tophores, are commonly G protein-coupled, Karlsson et al.
(35) recently examined the in vitro effects of odorants on
melanophores of the cuckoo wrasse, Labrus ossifagus.
Among some odorants tested, cinnaldehyde and -ionone
were found to have melanosome dispersing actions. Later,
Lundstrom and Svensson (36) actually tried to use
melanophores on a Labrus scale for odor sensing. Although
odorant molecules are relatively nonpolar, whether they can
penetrate the skin to influence chromatophores in vivo still
remains to be tested.
HORMONAL REGULATION OFCHROMATOPHORE MOVEMENT
Information perceived by lateral eyes and other sense organs
is transferred via the optic nerve to the central nervous
system, where it is integrated to yield adequate adaptive
chromatic reactions via endocrine, paracrine and neural
routes.
A number of principles are involved in the regulation of
chromatophore motility in fish. In order to facilitate the
understanding of the system for regulating chromatophores
therefore, consider the scheme shown in Fig. 1. This dia-
gram was drawn primarily to demonstrate the systems con-
trolling dendritic chromatophores of the light-absorbing
type that include melanophores, xanthophores and ery-
throphores. The diagram may also be practically applicable
to novel blue chromatophores (cyanophores) (5), although
certain modifications may be needed. On the other hand,
because of the different optical properties the regulatory
systems for light-scattering or reflecting chromatophores are
naturally somewhat different from those for light-absorbing
chromatophores. Therefore, although some parts are quite
analogous, the above diagram cannot be applied as it stands
to control systems for leucophores or motile iridophores.
Nuclear receptors have sometimes been shown to be
involved in the control of pigmentation in fish, but theireffects are always on morphological color changes (3, 4). It
may be pointed out here that cell-surface receptors are
exclusively concerned with systems controlling physiological
color changes, except in the case of nitric oxide (NO), which
will be briefly touched upon later.
Requiring complicated analyses, studies on mechanisms
regulating the production and release of pigment-motor
hormonal substances still remain to be investigated for the
most part, and therefore, the author did not try to review
those herein. With reference to outcomes from other fields
of studies, such as on mammals, amphibians, etc., the
mechanisms may hopefully be elucidated in the near future.
In this section therefore, the roles played by several hor-
monal principles that affect chromatophores areenumerated.
Melanophore-Stimulating Hormone
Among several hormonal principles known to control fish
chromatophores, melanophore-stimulating hormone (MSH)
produced by the intermediate lobe of the pituitary must be
the most widely known. Some readers, especially those who
are working in medically oriented fields, may wonder why
the term melanocyte-stimulating hormone is not employed
here. As noted previously, the term melanocyte is not
popularly employed by zoologists who are working with
poikilothermal animals, and instead, melanophore has long
been the common expression among them. Consequently,
the term has been cut in the hormones designation. In fact,
MSH induces very rapid dispersion of melanosomes within
melanophores (physiological color changes), in addition to
its other role in morphological color changes, i.e. stimulat-
ing the proliferation of melanophores and melanization
within them (1, 4). In any case, the effects of MSH are more
remarkable on melanophores than on melanocytes. Fortu-
Pigment Cell Res. 13, 2000 303
8/11/2019 2000-Fujii The regulation of motile activity in fish chromatophores.pdf
5/20
nately, the abbreviated form of the hormone, MSH, is
common, and thus we have practically no trouble in using
two different expressions in zoological and medical fields.
Even at the present time incidentally, MSH has still been
called melanotropin or intermedin rather frequently. In
general, the former has been used to indicate more inclu-
sively the peptide hormones that affect pigmentation, even
of invertebrates. It is thus desirable to rearrange the relevant
terms in order to avoid confusion.
Among molecular species of MSHs, -MSH (an acetyl-
tridecapeptide amide) is believed to have a major role both
in the regulation of chromatophores in lower vertebratesincluding fish (Fig. 2), and in melanocytes in homeotherms.
Namely, the structure of MSH may have been conserved for
a long time since the emergence of vertebrates. Among
rather primitive fish, somewhat modified peptides have been
reported, although we are still unaware that such structures
are the ancient forms of -MSH or not. To date, some
molecular species of-MSHs have also been reported (Fig.
2). As to whether the -forms are functional in color
changes in vivo, further study is needed. All MSHs are now
understood to be derived from a multi-functional precursor
called proopiomelanocortin (POMC).
A vast number of earlier studies on the action of MSH on
fish chromatophores was initially reviewed by Pickford and
Atz (37), and later, Fujii (4) and Fujii and Oshima (8)
summarized more recent work. Visconti et al. (38) recently
reported that -MSH effectively disperses pigment in
melanophores of an elasmobranch fish, using the skin of the
freshwater ray, Potamotrygon reticulatus. The actions of
MSH are not restricted to melanophores: The peptide has
frequently been reported to disperse xanthosomes and ery-
throsomes in bright-colored chromatophores in teleosts (8,
37, 3941).
Recently, studies on motile iridophores have made much
progress (3): It was shown that those of the blue damselfish
type and of the neon tetra type responded to -MSH by theSR response, but only when very strong solutions were
applied (9, 25). In the blue damselfish (Chrysiptra cyanea),
they were completely irresponsive (42). Motile iridophores
of the dendritic type, existing in some gobiid fish, responded
to MSH by aggregation of light-reflecting platelets (43, 44).
Apparently, such responses contribute to the darkening of
skin. Concurrent responses to MSH of light-absorbing chro-
matophores and iridophores function cooperatively to real-
ize effective dark-to-pale (and reverse) changes in the skin.
Usually, the direction of responses of light-absorbing
chromatophores, comprising of melanophores, xan-
thophores and erythrophores, and that of light-scattering
chromatophores, i.e. leucophores, are reciprocal (3, 8). For
Fig. 1. Diagram showing the regulatory system for motile activities of melanophores and other light-absorbing chromatophores in teleosts.Explanations for abbreviations in the figure are arranged in order from left to right. -A-R, -adrenoceptor; NE, norepinephrine; mACh-R,muscarinic acetylcholine receptor; ACh, acetylcholine; MCH-R, MCH receptor; Epi, epinephrine; ATP, adenosine 5-triphosphate;-MT-R,-MT receptor; PRL cell, prolactin-producing cell; MCH, melanin-concentrating hormone; -ET-R, -ET receptor; AL, anterior lobe ofhypophysis; cAMP, cyclic adenosine 3,5-monophosphate; cGMP, cyclic guanosine 3,5-monophosphate; IP3, inositol-1,4,5-trisphosphate;PRL, prolactin; AS-R, adenosine receptor; MSH cell, MSH-producing cell; PIH cell, PRL-release inhibiting hormone-secreting cell; IL,intermediate lobe of hypophysis; PRL-R, PRL receptor; MIH cell, MSH release-inhibiting hormone; MC-R, melanocortin receptor; -MSH,-melanophore-stimulating hormone;-A-R, -adrenoceptor; PL, posterior lobe of hypophysis; ET, endothelin; MT, melatonin; MCH cell,MCH-producing neuron in hypothalamus; -ET-R, -ET receptor; -MT-R, -melatonin receptor; NO, nitric oxide.
Pigment Cell Res. 13, 2000304
8/11/2019 2000-Fujii The regulation of motile activity in fish chromatophores.pdf
6/20
Fig. 2. Amino acid sequences of MSHs and MCH determined hitherto from fish [modified from Fujii and Oshima (8)].
example, sympathetic cues signal the aggregation of pig-
mented chromatosomes and the dispersion of light-reflectingleucosomes. Rather unexpectedly however, leucophores of
the medaka O. latipes responded to -MSH by leucosome
dispersion (40). Namely, in Oryzias the direction of leu-
cophore response is similar to that of the light-absorbing
chromatophores, the phenomenon being rather paradoxical.
Such seemingly odd processes may have evolved in order to
realize the delicate skin hues and patterns required for
adaptation to environmental conditions.
The action of MSH on fish melanophores has been shown
to be mediated by receptors that are specific to the peptide
(3, 8, 45). It was shown that MSH receptors require extra-
cellular Ca2+ ions for their action on melanophores (46).
Working on Oryzias xanthophores and leucophores and on
Xiphophorus erythrophores, Oshima and Fujii (41) further
showed that the peptide does not act to disperse chromato-
somes unless the bathing medium contains Ca2+ ions. It is
interesting that, among a number of hormonal and neural
substances signaling motile responses of fish chroma-
tophores, MSH is the only one that requires the presence of
extracellular Ca2+ ions. Those ions are probably required
for formation of the complex between the MSH molecule
and the regulatory subunit of the receptor.
Responses of motile iridophores of the dendritic type of
the dark sleeper goby to other signaling molecules, such as
NE, are analogous to Oryzias leucophores (43, 44), and the
principal second messenger is thought to be cyclic AMP. In
these iridophores therefore, MSH may signal platelet aggre-gation by decreasing adenylyl cyclase activity resulting in
the decreased levels of cAMP. This may be an unusual mode
of action for MSH.
On the basis of their functions, we are now urged to
classify MSH receptors into two large groups. To date, the
categorization of adrenoceptors into - and -forms has
already been established. Namely, the nucleotide-cyclase
inhibiting receptors are prefixed by , while those activating
the enzyme are designated . According to that principle,trials have already started to subclass receptors mediating
motile responses of chromatophores, such as those for MT
(32), and others including melanin-concentrating hormone
(MCH) and endothelins (ETs; cf. relevant sections in this
article). In the case of MSH receptors, the same yardstick
can not be applied unfortunately, because agonistic
molecules have already been endowed with the names of-
or -MSH. However, it might be possible to use -MC-R
and -MC-R for this purpose (see Fig. 1). In any case, the
above-mentioned novel MSH receptors of Odontobutis
melanophores should be treated using a different term when
we need to distinguish them from the conventional MSH
receptors.
ACTH
Although its role in physiological color change has not yet
been established, adrenocorticotropic hormone (ACTH) has
also been shown to be melanosome-dispersing (3, 4). This
fact is understood when we recall that the ACTH molecule
includes the amino acid sequence of MSH (Fig. 2). Other
brightly-colored chromatophores respond to ACTH as well.
For example, chromatosomes in xanthophores of the mud-
sucker gobyGillichthys mirabilis (47) and of the goldfish (48)
disperse in response to the peptide. Erythrophores in cul-
tures of the swordtail Xiphophorus helleri respond similarly
(39).
The receptor mediating the action of MSH has long been
called the MSH receptor. In the endocrinology of
homeotherms, the term MCn-R, where MC is the abbrevia-
tion of melanocortin accompanied by an Arabic number,
has become widely employed to express both the receptors
for MSH and ACTH. Since the cloning of the correspond-
ing receptors in poikilothermal vertebrates has not yet been
fruitful, such expressions have not yet become popular. We
Pigment Cell Res. 13, 2000 305
8/11/2019 2000-Fujii The regulation of motile activity in fish chromatophores.pdf
7/20
presume that in the near future nomenclature for related
receptors may be revised, in view of a more firm standpoint.
Prolactin
Prolactin (PRL), another peptide hormone produced by the
anterior lobe of the pituitary, was first shown to affect
chromatophores by Sage (47), who detected its ability to
disperse pigment in xanthophores of the mudsucker
Gillichthys, with resultant yellowing of the fish. Using two
highly purified molecular species of PRL from the Mozam-
bique tilapia Oreochromis mozambicus (tPRL177 andtPRL188), Oshima and her associates examined their effects
on chromatophores of the Nile tilapia, O. niloticus, and
recognized that the peptides had little, if any, melanosome-
aggregating effects on melanophores, but that tPRL177 had
the distinct action to disperse pigment in xanthophores (49).
They further investigated the chromatosome-dispersing ef-
fects of tPRL177on xanthophores of the Nile tilapia and the
rose bitterling (Rhodeus ocellatus ocellatus), and on ery-
throphores of tilapias, swordtails (X. helleri) and paradise
gobies (Rhinogobius giurinus), and were able to further
detect seasonal changes in the responsiveness of ery-
throphores to the hormones. Based on these observations,
they concluded that the enhanced PRL action on ery-
throphores in the breeding season must be deeply involvedin expressing nuptial coloration (50). Dispersion of chro-
matosomes may be linked to the synthesis of brightly-col-
ored pigments, namely, their sparse distribution within the
perikaryon may release the Golgi-endoplasmic reticulum
system to synthesize more chromatosomes, by unfastening
the product inhibition, which would result in the generation
of the conspicuous hues for courtship.
As mentioned above, PRL seems to have rather limited
effects on teleostean melanophores (50), but Visconti et al.
(38) recently reported that PRL darkens the skin of a
freshwater ray (P. reticulatus) effectively, suggesting its ac-
tive role in elasmobranch coloration, although further com-
parative examinations are needed.
Somatolactin
Somatolactin (SL) is a novel teleostean pituitary hormone
belonging to the growth hormone-prolactin (PL) family
(51). Various molecular forms have already been cloned,
which have more than 200 amino acids (52). Using the red
drum, Sciaenops ocellatus (Sciaenidae), Zhu and Thomas
(53) found that the increase of SL in the plasma is associ-
ated with the aggregation of melanophore inclusions. How-
ever, their results to date are rather confusing, necessitating
further analyses for establishing SLs participation in
pigmentation.
Melanin-Concentrating Hormone
The presence of a hormone antagonizing the action of MSH
had long been a matter of controversial opinion. Strong
suggestion of the hypothalamic origin of such a principle
was first presented by Enami (54), who named it
melanophore-concentrating hormone (MCH). As a neu-
rosecretory hormone, it is transferred from the hypothala-
mus to the posterior lobe of the pituitary from which it is
secreted (55). Baker and her colleagues tried to characterize
it (56), and finally Kawauchi et al. (57) succeeded in isolat-
ing it from the pituitary glands of the chum salmon
Oncorhynchus keta. It is a cyclic heptadecapeptide with a
disulfide bond (Fig. 2), and it is now called melanin-con-
centrating hormone, because what concentrates are not
melanophores, but melanin-carrying organelles.
Nagai et al. (58) reported that motile melanophores of all
teleostean species they tested responded to MCH by aggre-
gation of melanosomes, as the name implies. The action of
MCH is mediated by a specific receptor (5961). It shouldbe emphasized, however, that the definite action of MCH
has been shown only in teleosts: In amphibians and reptiles,
melanophores responded to that hormone by dispersing
melanosomes, and the sensitivity was much lower than that
in fish (62). The biological significance of MCH in eliciting
color changes in lower vertebrates has recently been well
documented by Baker (63) who naturally devoted much
space about its action on fish chromatophores.
Chromatophores other than melanophores responded
similarly to MCH (14, 60, 61). For example, Oshima et al.
(60) showed that swordtail erythrophores and medaka xan-
thophores responded well to MCH by chromatosome aggre-
gation. Motile iridophores of the blue damselfish,
Chrysiptera cyanea, were among the few instances of chro-
matophores that are refractory to MCH (64), those iri-
dophores being regulated solely by nerves.
In contrast, light-scattering organelles in leucophores of
medaka dispersed in response to MCH, but much higher
concentrations of the hormone were needed (60). Further, in
contrast to its pigment aggregating action, extracellular
Ca2+ ions were needed, as for the melanosome-dispersing
action on amphibian melanophores. Thus, it was once
thought that the pigment-dispersing action of MCH might
be mediated by MSH receptors.
Castrucci et al. (61) examined the action of MCH on
melanophores of the Brazilian eel (Synbranchus marmora-
tus), and reported that at lower concentrations it aggregatedmelanosomes, whereas at higher concentrations it dispersed
them. Applying higher concentrations of MCH to
melanophores of the mailed catfish Corydoras paleatus and
the Nile tilapia, O. niloticus, Oshima and her associates also
observed that the melanosome aggregation was followed by
re-dispersion, and that Ca2+ ions were necessary for the
latter process (65, 66). As mentioned above, MSH receptors
require external Ca2+ for their action, and therefore, a
dense population of MSH receptors on the cell membrane
might have been concerned with this process. An alternative
explanation was recently put forward by Oshima who as-
sumes that there are two types of receptors for MCH (65)
that exist commonly on melanophores, medaka xan-
thophores and swordtail erythrophores. The first type of
MCH receptor would mediate pigment aggregation at phys-
iological concentrations, while those of the other type of
MCH receptor on the membranes of medaka leucophores
and of amphibian melanophores would mediate dispersion
of pigment, but only when the agonist concentration is very
high, and would require extracellular Ca2+ ions. In
melanophores of the Brazilian eel, both types of receptors
Pigment Cell Res. 13, 2000306
8/11/2019 2000-Fujii The regulation of motile activity in fish chromatophores.pdf
8/20
8/11/2019 2000-Fujii The regulation of motile activity in fish chromatophores.pdf
9/20
PARACRINE FACTORS
In higher vertebrates, several paracrine factors have been
shown to regulate the physiological responses of effector
cells. Such paracrine systems may also be operating in lower
animals, because they seem be the most primitive means of
communication among cells. In fish, such factors might
include prostaglandins (PG), angiotensin II, ETs,
bradykinin, somatostatin, and other neuropeptides, includ-
ing intestinal hormones, etc. To date, however, few reports
have appeared that demonstrate such processes in the chro-
matic systems of fish.
Opioid peptides
Opioid receptors are present in the brain, as well as in the
peripheral tissues, of vertebrates. Since they have been
shown to inhibit liberation of transmitters from nerve termi-
nals, similar roles of these neuropeptides in modulating the
primary effects of endocrine or nervous cues of chroma-
tophores could be expected.
Suggesting a possible role of opioid peptides in the secre-
tion of MSH, Satake (79) demonstrated that an intracranial
injection of naloxone, a specific inhibitor of opiate recep-
tors, induced aggregation of pigment in goldfish xan-
thophores. The effect was antagonized by
methionine-enkephalin (met-E). Next, Levina and Gordon
(80) showed that melanophores and xanthophores of ze-
brafish (Brachydanio rerio) responded to MSH and to met-E
by chromatosome dispersion, and that the effect of met-E
developed later and faded more slowly. Naloxone inhibited
the action of met-E, and the involvement of a central
mechanism was suggested in the met-E-induced darkening
of the skin. Recently, Carter and Baker (81) reported that
either the pars distalis or the neurointermediate lobe of the
pituitary actually contains substantial opiate activity. To
date, however, little information is available about the role
of opioid peptides in regulating chromatophores in fish.
Eicosanoids
Among physiologically active eicosanoids, PGs are of much
interest, because they are regarded to be important factors
in modifying the regulation of hormonal, as well as neural
signaling to effector cells. In fact, they have frequently been
shown to influence activities of various autonomically regu-
lated effectors via paracrine signaling. As early as 1974,
Abramowitz and Chavin (82) noted that PGs elicited disper-
sion of pigment in melanophores of black goldfish in vitro.
Further investigations along this line, however, have been
unexpectedly meager. It is therefore, desirable to know
whether these and related fatty acid derivatives take part in
modulating chromatophore responses.
Endothelins
It has recently been shown that human keratinocytes pro-
duce ETs, which can act as strong mitogens, as well as
melanogens, for human melanocytes (83). Keratinocytes and
adjacent melanocytes may form the paracrine linkage for
ET. Working on teleostean fish, Fujii and his associates
found that ET induced motile responses of most chroma-
tophores in the teleosts examined (84), and that their actions
were dose-dependent. The pharmacological properties of ET
receptors possessed by melanophores (85), erythrophores,
xanthophores (86), and motile iridophores (unpublished ob-
servations) resemble those of ETB described in mammalian
tissues. The direction of responses to ET of these chroma-
tophores coincides with that of the responses to sympathetic
stimuli via -adrenoceptors. In addition to cyclic AMP,
inositol 1,4,5-triphosphate (IP3) has already been found to
work as another second messenger mediating the aggrega-
tion of pigment, at least in some chromatophores (87, 88).
Therefore, the process of signaling in the response to ET ofthese chromatophores might be analogous to those disclosed
in mammalian tissues, including human melanocytes (89).
ETs, by contrast, disperse leucosomes in leucophores of
the medaka, O. latipes (90). The pharmacological properties
of ET receptors of leucophores resemble mammalian ETB,
as in other chromatophore species of fish. On the other
hand, Lerner and his associates (91), while working on
melanophores of the African clawed toad Xenopus laeis,
reported that ET dispersed melanosomes mediated by ETCreceptors. The direction of the pigmentary response to ET
was identical to that in Oryziasleucophores, but opposite to
that observed in most teleostean chromatophores (3, 84
86). Lerners group (92) also reported that an increase in the
cytosolic levels of IP3 correlated with melanosome disper-
sion in Xenopus melanophores, which in terms of the direc-
tion of melanosome displacement, was quite opposite to that
reported by us in fish (87). The involvement of IP3in motile
responses of leucophores has not yet been studied. In con-
sideration of past results on the common roles of second
messengers in teleost chromatophores (3, 14), however, it is
likely that IP3 also mediates the aggregation of leucosomes.
Namely, ET receptors of leucophores might mediate the
dispersion of leucosomes via decreases in the intracellular
levels of IP3. Thus, ET receptors of leucophores are quite
different from Xenopus ETC, and also from those of other
kinds of chromatophores of teleosts examined to date. Ten-
tatively, we named the ET receptors of leucophores -ETreceptors, and those of light-absorbing cells -ET recep-
tors. The adoption of the prefixes and is based on the
terminology of some pigment-motor substances that have
reciprocal actions on chromatophores, as touched upon
previously. In Fig. 1, which exhibits the general regulatory
system for motile activities of light-absorbing chroma-
tophores in teleosts, both -ET and -ET receptors are
incorporated. ET may be secreted as a paracrine factor to
modify the actions of the known nervous or hormonal
principles.
Working on an elasmobranch species (P. reticulatus),
Visconti et al. (38) recently reported that ETs were not able
to induce either skin lightening or darkening. Thus,
melanophores of this species may be unresponsive to ET.
Since ET has definite actions on teleostean chromatophores,
further comparative studies are needed in lower fish.
Imokawa et al. (83, 89) showed that in humans, kerati-
nocytes are the source of ET. Very recently, the secretion of
ET from goldfish epidermal cells in culture has been re-
ported (93), and thus, the possible source of ET for chroma-
tophore responses might be sought there. Epidermal
Pigment Cell Res. 13, 2000308
8/11/2019 2000-Fujii The regulation of motile activity in fish chromatophores.pdf
10/20
8/11/2019 2000-Fujii The regulation of motile activity in fish chromatophores.pdf
11/20
Fig. 4. Diagram showing the chromatic nervous pathways frommelanosome-aggregating center to melanophores in fish. Originallydescribed in the minnow, P. laeis, by von Frisch (99).
variety of-blockers developed thereafter, many later work-
ers have come to the same conclusion (3, 8, 102). Employing
radiolabeled NE, Kumazawa and Fujii (103) actually
showed that NE is released from nervous elements in re-
sponse to nervous stimuli. Current investigations, further-
more, suggest that the firing rate of postganglionic
sympathetic fibers needed to maintain melanophores in an
intermediate state of pigment aggregation in vivo is rela-
tively low, being about 1 Hz. A higher firing rate results in
a more aggregated state, while a lower frequency, or a
cessation, of impulses causes dispersion of pigmentary or-
ganelles (101, 102).Several researchers have attempted to determine the sub-
type of -adrenoceptors on chromatophores. Some have
reported that 2-agonists are more effective than 1-ago-
nists, and that transmission is more easily blocked by 2-
blockers than by 1-blockers (104, 105). Those workers
naturally came to the conclusion that the pigment-aggregat-
ing adrenoceptors are of the a2 type, and that cyclic AMP is
functioning as a second messenger. Recently, Mayo and
Burton (106) stated that adrenoceptors possessed by
melanophores of the winter flounder, Pleuronectes (syn-
onym:Pseudopleuronectes) americanus, are mostly of the a2
subtype.
Working on melanophores of the cuckoo wrasse, L. os-
sifagus, Svensson et al. (107) recently succeeded in cloning
most 2-adrenoceptors for the first time among varieties of
receptors mediating chromatophore movements. The de-
duced amino acid sequence of the peptide sequence showed
4757% homology with human 2-adrenoceptors. Together
with data from forthcoming cloned receptors, the results
may afford important data for receptor mechanisms, as well
as for understanding the phylogenetic relationships among
species in the large class, Osteichthyes.
At least in some species the aggregation of pigment may
be triggered by an increase in levels of Ca2+ ions in the
cytosol (108 110). In addition, Fujii et al. (87) recently
demonstrated the involvement of inositol 1,4,5-trisphos-
phate (IP3) in the aggregation of pigment in tilapiamelanophores. In many different cell types, IP3 has been
shown to induce the release from intracellular storage com-
partments of Ca2+ ions into the cytosol. Moreover, we are
now aware that 1-adrenergic stimuli activate phospholipase
C, which catalyzes the production of IP3. These observa-
tions indicate that in addition to 2-adrenoceptors, 1-
adrenoceptors are functional at least in some cases. In fact,
a remarkable aggregation of pigment takes place in response
to l-agonistic stimuli, and 1-type adrenolytics always have
inhibitory effects on that process.
Chromatophores other than melanophores have also been
shown to be under the control of the sympathetic system.
For instance, erythrophores of the swordtail, X. helleri(40)
and those of the squirrelfish Holocentrus ascensionis (108)
have been shown to be under the influence of the nervous
system. Comparing the physiological characteristics of xan-
thophores with those of melanophores and leucophores on
scales of the medaka O. latipes, Iwata et al. (111) showed
that xanthophores responded in quite the same manner as
melanophores. Therefore, the nervous mechanisms con-
trolling xanthophores seem to be analogous with those of
their study (96) is plausible, further detailed examinations
are needed to present more precise neuronal connections.
Recently, Grove (97) wrote an interesting review relevant
to this subject, including several historical and rather little
known outcomes, to which readers can refer with interest.
Sympathetic Innervation
The peripheral nervous mechanism controlling fish chroma-
tophores has a long history of investigation. Earlier works
indicated that chromatophores of lower fish, including elas-
mobranches, are also under the control of the nervous
system, in addition to the hormonal regulation (15). Nowa-
days, however, they are regarded as predominantly under
the control of endocrine systems (4, 8, 38, 69, 98). In bony
fish, by contrast, a strong participation of the nervous
control of chromatophores has been shown repeatedly (3, 8,
98).
Several researchers have tried to follow the tracts of
chromatic fibers from the center. As an example, a diagram
based on earlier descriptions by von Frisch (99) on the
melanin-aggregating nervous pathways in the minnow,
Phoxinus laeis, is exhibited here (Fig. 4). This scheme is still
applicable to any teleostean species without major modifica-
tions. The diagram shown as Fig. 3 (96) is the modern
version of the von Frisch original.
Apparently, von Frisch himself anticipated the presenceof the antagonistic melanin-dispersing fibers that run
alongside the aggregating fibers, but could not observe them
physically. Later workers occasionally tried to depict the
pathways, such as that presented by von Gelei (100), who
also worked on the same species ofPhoxinus minnow, but
as already touched upon above, the presence of such fibers
has been disproven.
If electrical stimulation of nerve fibers to the skin gives
rise to motile responses of chromatophores existing down-
stream, we can safely believe that those cells are under the
control of the nervous system. As far as we are aware, most
melanophores of teleosts are innervated by such nerves, and
their mode of innervation has been analyzed (29, 101).
Since innervation to chromatophores has been thought to
be sympathetic postganglionic, the peripheral neurotrans-
mitter that signals chromatophores was justifiably supposed
to be adrenergic. Observing the effects of an adrenergic
antagonist, dibenamine, Fujii (67) first demonstrated the
adrenergic nature of transmission to melanophores.
Dibenamine is known to block -adrenoceptors, and thus,
the transmission could be regarded as -adrenergic. Using a
Pigment Cell Res. 13, 2000310
8/11/2019 2000-Fujii The regulation of motile activity in fish chromatophores.pdf
12/20
melanophores. In general, however, the nervous influences
on erythrophores and xanthophores seem to be weaker
compared with melanophores.
In leucophores, nervous stimulation induces the reverse
movement, namely, the dispersion of light-scattering or-
ganelles (112, 113). In their study, Iwata et al. (111) further
showed that melanophores and the leucophores existing
nearby are under the control of the same fibers. The recep-
tors concerned are of the -adrenergic type (114, 115). Later
pharmacological analyses by Yamada (116) indicated that
those subtype of the receptors is 1. Iga (117) noticed that
under the blockade of -adrenoceptors, leucosomes aggre-gated in response to catecholamines, and concluded that the
response was mediated by adrenoceptors of the type.
Later, Morishita and Yamada (118) characterized these
receptors to be of the 2 type. It remains to be determined
whether receptors of this type actually function in vivo.
Recently, Iga and Mio (119) discovered leucophores in the
skin of the dark-banded rockfish Sebastes inermis, and
reported that adrenergic mechanisms controlling leucosome
movements are fundamentally the same as those ofOryzias.
Motile iridophores of the non-dendritic type responded to
nervous stimulation by the LR response (9, 42). In dendritic
iridophores of the goby type, platelets disperse into pro-
cesses upon nervous stimulation (10, 11).
By means of autoradiography using radiolabeled NE,
Yamada et al. (120) succeeded in visualizing the pattern of
adrenergic innervation on melanophores of the medaka O.
latipes clearly. They also demonstrated the pattern of inner-
vation to erythrophores of the swordtail, X. helleri (121).
Using medaka, Sugimoto and Oshima (122) showed that
dark background adaptation resulted in increased numbers
of melanophores and xanthophores along with denser net-
works of varicose fibers around those chromatophores, and
that reverse changes occurred in white background adapted
fish. It was further shown that, after long-term adaptation
to a white background, the responsiveness of melanophores
to NE was reduced (123). For a better understanding about
the coupling of the morphological to the physiological colorchanges, further examinations are naturally needed.
Cholinergic Transmission to Melanophores
Working on two catfish species belonging to the family
Siluridae (order: Siluriformes), Fujii and his associates
found that peripheral transmission to melanophores is
cholinergic, notwithstanding the fact that postganglionic
fibers to the effector cells are sympathetic in the usual
manner. The common Japanese catfish, S. asotus (124), and
the translucent glass catfish, K.bicirrhis(78) were the species
examined. Replacing -adrenoceptors entirely, cholinocep-
tors of the muscarinic type play an exclusive role in trans-
ducing nervous signals to the melanophores. Since they
belong to two remote genera, we presume other species in
this family may also be controlled in the same way. Surveys
have been made to examine the presence of cholinoceptors
in other catfish families within the order Siluriformes. It was
found that, in families close to Siluridae, melanophores are
often endowed with adrenergic and cholinergic receptors,
both of which mediate the aggregation of melanosomes
(125). In these fish, the neurally evoked aggregation of
pigment is mediated by -adrenoceptors, as it is in many
common teleosts. Thus, the physiological roles of these extra
cholinoceptors in those fish still remain to be solved.
Recently, Hayashi and Fujii (126) discovered that some,
but not all, melanophores of two species belonging to the
genusZacco(family: Cyprinidae, order: Cypriniformes) pos-
sess muscarinic cholinoceptors that also mediate
melanosome aggregation. That was the first report to de-
scribe the presence of cholinoceptors on chromatophores in
fish species other than those which belong to the order
Siluriformes.Making use of selective antagonists for muscarinic recep-
tors, Hayashi and Fujii (127) characterized the muscarinic
cholinoceptors possessed by melanophores of the glass
catfish,K.bicirrhis, and the mailed catfish,C.paleatus, to be
of the M3 subtype.
Until the present time, no reports have appeared about
the existence of such cholinoceptors of chromatophores
other than melanophores.
True and Co-Transmitter Interactions
It was first suggested by Fujii and Miyashita (128) that
adenosine or adenine nucleotides might take part in con-
trolling pigment dispersal in fish chromatophores. They
found that non-cyclic adenylyl compounds, which were used
as control compounds, were even more effective than cyclic
adenosine 3,5-monophosphate (cAMP) in dispersing pig-
ment in melanophores of guppies. Using guppies and silurid
catfish, Miyashita et al. (129) extended this pharmacological
analysis and came to the conclusion that the pigment-dis-
persing action of these nucleotides was mediated by
adenosine receptors since those effects could easily be antag-
onized by methylxanthines, specific blockers of adenosine
receptors.
Working on melanophores of tilapias, Kumazawa et al.
(130) detected the apparent liberation of ATP from chro-
matic nerves in response to electrical stimulation. Theyconcluded that ATP is released as a co-transmitter from
postganglionic sympathetic fibers together with the true
transmitter, NE. The concurrent release of the true transmit-
ter and co-transmitter from the fibers to chromatophores
has been confirmed in experiments with radiolabeled com-
pounds (103, 131).
The peripheral nervous mechanism, as characterized to
date, is shown schematically in Fig. 5. The true transmitter,
NE, acts to induce a rapid aggregation of melanosomes via
mediation by -adrenoceptors on the membrane. Most NE
molecules are quickly removed by being taken back up into
the nervous elements. The remainder is either removed via
the general circulation or is inactivated by catecholamine
O-methyltransferase (COMT) and monoamine oxidase
(MAO). ATP released concurrently with NE is dephospho-
rylated by ATPase and then by 5 -nucleotidase in the synap-
tic cleft. The resultant nucleoside, adenosine, survives for
some time there and functions to reverse the influence of the
true transmitter, namely, to cause the re-dispersion of pig-
ment via specific receptors for adenosine on the effector
membrane. Most of the nucleoside is finally removed by
Pigment Cell Res. 13, 2000 311
8/11/2019 2000-Fujii The regulation of motile activity in fish chromatophores.pdf
13/20
being taken back up into presynaptic nervous elements, and
the remainder is carried away by the circulation.
One important aspect of the proposed dual-transmitter
theory is that there is a substantial difference, in terms of
action, between the true transmitter and the co-transmitter.
The effect of NE disappears very quickly, while that of the
co-transmitter lasts longer. After the cessation of nervous
excitation, the latter can effectively reverse the action of the
former. Rapid dispersion of pigment after nervous excita-
tion is realized in this way. The rapid changes observable in
living animals may also be controlled by the same mecha-
nism. An identical explanation has been presented for theregulation of melanophores of the blue damselfish C.cyanea
(64) and the blue-green damselfish Chromis iridis (9). The
motile responses of amelanotic melanophores of medaka are
also regulated in the same way (132).
Recent studies on medaka indicate that leucophores re-
spond to adenosine by dispersion of leucosomes (133). Spe-
cific adenosine receptors of the A2 type mediate this
response. However, the direction of the movement of leuco-
somes in response to the co-transmitter is the same as that
elicited by the true transmitter. In fact, the recovery from
the effect of NE occurs very slowly.
The involvement of the dual-transmitter system in the
control of motile iridophores may be analogous to that of
melanophores. The motile iridophores of blue damselfish
(64), blue-green damselfish (9) and neon tetras (25) respond
to adenine derivatives of adenine with the SR response,
which is the opposite of the LR response elicited by -
adrenergic stimuli.
Feedback Inhibition of Transmitter Release
Using the tilapia O. niloticus, Oshima (134) succeeded in
showing that adenylyl compounds, including adenosine and
ATP, inhibit the release of adrenergic transmitter, possibly
by decreasing the rate of entry of Ca2+ ions into presynap-
tic portions of the fibers. Since these nucleotides are thought
to be released as the co-transmitter from the sympathetic
fibers (cf. above subsection), such a feedback inhibitory
mechanism is a kind of autocrine mechanism. Strangely,
neither inhibition via 2-adrenoceptors nor acceleration via
-adrenoceptors of the outflow of the transmitter has been
proven to date.
Relationship to Chromatic Patterns
We know well that chromatic patterns of the integument arevery important for the survival of animals in their habitat
(3). Some chromatic patterns are practically stationary. Very
frequently, such patterns change under various ethological
conditions. Among such changes, slower ones, such as those
that take place during ontogeny, are brought about by
morphological color changes, but faster changes in patterns
are due to physiological color changes. For example, in-
volvement of the pineal gland secretion, MT, in circadian
pattern changes in pencilfish has already been mentioned.
More rapid changes needed for adapting to background
patterns or for intraspecific communication can only be
realized through the activities of the nervous systems. Past
studies have indicated that changes are due to differential
neural commands to chromatophores or to groups of chro-matophores. In practice, there is a limited number of preset
patterns. On the basis of the coarseness of the background
texture, the central nervous system selects an appropriate
pattern (2, 8). Naitoh et al. (135) studied the chromatic
adaptation of the common freshwater goby, Rhinogobius
brunneus, to black and white checkerboard backgrounds,
and found that numerous nerve fibers control integumentary
chromatophores differentially and in a coordinated manner.
Several species of tilapias have recently been widely em-
ployed for analyzing communicatory functions of various
Fig. 5. Diagram showingtransmission from sympatheticpostganglionic fibers tochromatophores in which both trueand co-transmitters are involved.COMT, catecholamineO-methyltransferase; MAO,monoamine oxidase; NE,norepinephrine;-A-R,-adrenoceptor; Gi, inhibitoryG-protein; ATP, adenosine5-triphosphate; AC, adenylyl cyclase;cAMP, cyclic adenosine3,5-monophosphate; AMP, adenosine5-monophosphate; AS, adenosine;AS-R, adenosine receptor; Gs,stimulatory G-protein; AC, adenylylcyclase; IS, inosine [Modified fromFujii and Oshima (8)].
Pigment Cell Res. 13, 2000312
8/11/2019 2000-Fujii The regulation of motile activity in fish chromatophores.pdf
14/20
pigmentary patterns, and many interesting results have been
obtained (136). Working with a mouth-brooder tilapia,
Haplochromis burtoni, Muske and Fernald (137) recently
showed that nervous cues control the very rapid appearance
or disappearance of the facial stripe, the eyebar, which
signals territorial ownership and aggressive intent in males.
Nervously controlled pattern changes in flatfish that show
rapid adaptability to background patterns are famous, and
are another good example to mention, and indeed, a num-
ber of investigations have been done on these interesting fish
(8, 15). In their recent series of studies, Burton and his
associates put forward integrative analyses on the patterningof the winter flounder, P. americanus. (106, 138, 139). This
species of flatfish has a dark band, general background and
white spot components with different responsiveness in vivo
to stress. For example, measurement in vitro showed that
melanophores in the white spots showed a much higher
concentration threshold to NE than that for the other two
pattern components. They also found that the spot
melanophores responded to increased K+ concentration
and to electrical nervous stimulation faster than other com-
ponents, and further, that the inhibitory influences of -
adrenergic blockers differed among pattern components.
Based on those data, they concluded that the differential
activity associated with patterning includes a peripheral
neuroeffector component, part of which is directly associ-
ated with melanophores.
That the endocrine system takes part in pattern formation
seems difficult to understand because hormonal substances
go everywhere in the body rather homogeneously via the
general circulation. However, the involvement of MT in
such processes has recently been disclosed, as mentioned
before. Thus, the situation has become a challenge to analy-
sis, but presumably, correlated management of chroma-
tophores by both the endocrine and the neural systems can
elicit elaborate changes in patterns.
Effects of Nerve CuttingSeverance of chromatic nerve-fibers en route to chroma-
tophores naturally results in the interruption of central tonic
influences on effector cells, namely the darkening of the
downstream zone. Based on their observations of such
denervated dark bands in some teleostean and elasmobranch
fish, Parker and his colleagues came to the conclusion that
the response was caused by the repetitive firing of putative
parasympathetic melanin-dispersing fibers at the cut ends of
axons (15). For a number of reports relevant to this prob-
lem, readers can refer to a monograph by Parker (15). As
already mentioned, however, this double innervation the-
ory has not been supported, but the phenomenon itself
provides various important clues for understanding various
mechanisms of the effector systems (3, 14, 140).
Among the phenomena taking place after the denervation
of chromatophores, hypersensitization to some pigment-mo-
tor substances is worth mentioning again, because some new
observations have appeared. Using the goby C. gulosus,
Fujii (67) had already described his quantitative results on
the hypersensitivity of melanophores to epinephrine, NE
and also to MT. Karlsson et al. (141), while working on the
cuckoo wrasse L. ossiphagus, noted that after putting
melanophores in culture, they became hypersensitized to
-adrenergic stimuli, and they concluded that such effects
were due to denervation. Employing scales plucked from
individual medaka that had been adapted to a dark back-
ground for 10 days, Sugimoto (123) reported that the re-
sponsiveness of melanophores to NE significantly increased.
His conclusion was that the depressed sympathetic nervous
activities during the dark-background adaptation might
have affected the cells like denervation. Fujii and Oshima
(14), however, think that the hypersensitivity may be related
to the loss of spontaneously released adenylyl co-transmitterfrom sympathetic fibers, since the co-transmitter is now
known to antagonize the action of the true transmitter
either in vivo or in vitro, as described above. The physiolog-
ical significance of denervation hypersensitization is still
unclear.
SIGNAL TRANSDUCTION ACROSSCHROMATOPHORE MEMBRANE
Signal transduction studies up to the 1990s have been
previously reviewed by Fujii (3, 14), and Nery and Castrucci
(88) have recently reviewed work dealing with signaling
mechanisms in chromatophores in poikilothermal animals.A relevant article by Oshima (65) will also appear soon, and
therefore, the author will summarize his views on this topic,
only considering some recent works.
Membrane Potential Changes
Like smooth muscle cells, chromatophores of fish are under
the control of the sympathetic nervous system. Rather
strangely, however, their motile activities seem to be inde-
pendent of the electrical activities of the surface membrane
since Tetrodotoxin did not affect the motility of
melanophores per se (29). Action potentials therefore, are
not required for triggering cellular motility. Working on
denervated skin pieces, or on those in which the liberationof neurotransmitters was blocked, Fujii and Taguchi (70)
showed that melanosome-aggregating and dispersing agents
induced motile reactions of melanophores quite normally
when the cells were in saline in which Na+ ions were totally
replaced with an equimolar amount of K+ ions. Under such
conditions, the cell membrane should have been completely
depolarized. These results indicate that ionic fluxes across
the membrane and resultant changes in the membrane po-
tential are irrelevant to the cellular responses. Meanwhile, it
has become clear that in chromatophores of many species,
even the presence in the extracellular space of Ca2+ ions is
not required for their motility (3, 14, 70, 87). Namely, Ca2+
inflow or Ca2+ potential may not be involved in the motile
responses. The sum of these observations shows that
voltage-dependent ionic channels are not involved with the
responses.
All chromatophores in vertebrates have been shown to be
of neural crest origin, and thus, they are categorized as
so-called paraneurons (3). It is rather strange therefore,
that their cell membrane is not electrically excitable. As
briefly touched upon below, results on signal transduction
Pigment Cell Res. 13, 2000 313
8/11/2019 2000-Fujii The regulation of motile activity in fish chromatophores.pdf
15/20
across the cell membrane show that receptors that mediate
pigment-motor responses are G protein-coupled. Since that
process may not require membrane potential changes, the
irrelevance of electrical activities across chromatophore
membranes can be understood smoothly.
As a rather exceptional case for the role of the cell
membrane, we should mention again the erythrophores of
holocentrid squirrelfish. Luby-Phelps and Porter (108) re-
ported that the aggregation of pigment in erythrophores of
H. ascensionis definitely depend on extracellular Ca2+ ions.
Depolarization of the cell membrane due to K+-rich
medium may open voltage-dependent Ca2+ channels, allow-ing the inflow of ions to initiate pigment aggregation. The
sequence is quite different from that disclosed in
melanophores and other chromatophores (3, 14, 142). We
have recently observed the responses of erythrophores of
squirrelfish belonging to the family Holocentridae (order
Beryciformes), including the crowned squirrelfish,Sargocen-
tron diadema, and the soldierfish species, Myripristis ran-
dalli, and we have obtained results fundamentally identical
to those described above (108) (unpublished observations).
It should also be mentioned here that before the finding
of the peculiar responses of squirrelfish erythrophores by
Luby-Phelps and Porter (108), Iga (143) had already re-
ported that xanthophores of the medaka, O. latipes, are
directly responsive in pigment aggregation to increased con-
centrations of K+ ions. Confirming Igas observations,
Oshima et al. (144) recently concluded that Ca2+ ions
penetrate the cytosol through the voltage-dependent chan-
nels, which lower the levels of cAMP by inhibiting adenylyl
cyclase. Incidentally, in melanophores and other chroma-
tophores, K+ ions have been shown not to act directly on
the cells, but rather on nervous elements to release neuro-
transmitters which in turn aggregate pigment (3, 142). These
results suggest that there are some exceptional cases among
bright-colored chromatophores, across the cell membrane of
which depolarization takes place to allow Ca2+ influx,
resulting finally in pigment aggregation. As for other kinds
of chromatophores, no reliable data have been publishedthat indicate their existence.
At the beginning of their assignment to take part in
physiological color changes, the membranes of chroma-
tophores in ancient fish might have been excitable because
they had a common origin with neurons. Being different
from many electrically excitable cells, the chromatophores
have not been required to exert such quick reactions. We
also know that a large amount of ionic flux requires much
energy to recover the ionic distribution across the cell
membrane. Our conclusion therefore, is that erythrophores
or xanthophores of the Holocentrus type may belong to a
more primitive type of chromatophore, retaining ancient
physiological properties.
Cyclic AMP
Undoubtedly, cAMP is the major second messenger in
chromatophores. As early as 1970, Novales and Fujii (145),
while working on split tail-fin pieces of Fundulus killifish,
succeeded in detecting a melanosome-dispersing effect of
extracellularly applied cAMP. In order to obtain more
direct evidence for the role of cAMP, Fujii and Miyashita
(128) injected the nucleotide iontophoretically into guppy
melanophores and could observe the dispersal of
melanosomes. Using a photolabile caged cAMP, Furuta et
al. (146) recently succeeded in detecting its melanosome-dis-
persing effects on Oryzias melanophores. Detection by
means of radioimmunoassay of an increase in the level of
cAMP in Xiphophorus melanoma cells (147) and in guppy
melanophores (148) provided results in accordance with this
concept. Thereafter, many reports on the role of AMP in
dispersing pigment, not only in melanophores, but also in
xanthophores, erythrophores and leucophores, have ap-peared, and have been reviewed several times (3, 14, 73, 88).
Those reviews also deal with current understanding of the
mechanisms involved in the process of signal transduction.
The first step in the motile response of the chromatophore
is the binding of the first messenger, i.e. a hormonal or
neuronal substance, to corresponding receptors which con-
stitute regulatory subunits of adenylyl cyclase. The informa-
tion is then signaled via a GTP-binding protein, either Gs or
Gi, to the catalytic subunit of adenylyl cyclase. The increase
in cytosolic levels of cAMP is due to the heightened activity
of this subunit, leading finally to the pigment dispersion.
The reverse process, i.e. the aggregation of pigment, is
triggered by a decrease in the level of the nucleotide, which
results from decreased activity of the catalytic subunit via
Gi.
It has generally been accepted that the mechanisms of
action of catecholamines and peptide hormones are cAMP-
dependent. For example, we have already treated adreno-
ceptor-mediated dispersion of pigment in a relevant section,
and the adrenoceptors involved are mixtures of 1 and 2(76). The latest results along this line on fish chroma-
tophores include those on MCH. Using melanophores of the
Nile tilapia,O. niloticus, Oshima and Wannitikul (66) exam-
ined the signaling mechanism for MCH. Based on their
results obtained by employing various inhibitors, they con-
cluded that cAMP is the second messenger involved.
It is interesting to know that even in nervously evokedpigment aggregation, cAMP is the major second messenger.
Recent results are in agreement with the view that the
subtype of adrenoceptors concerned is of the a2 subtype,
which work to diminish adenylyl cyclase (104, 105, 107).
Ca2+ and IP3
As mentioned before, the cell membrane of common chro-
matophores is quite resistant to changes in the external ionic
composition. Meanwhile, Luby-Phelps and Porter (108) pre-
sented their results on erythrophores of the squirrelfish H.
ascensionis in which an influx of external Ca2+ ions is
required for the aggregation of pigment. Using a Ca2+
ionophore, they further manipulated the intracellular con-
centration of Ca2+ ions, and showed that the response was
dependent on the concentration of ions. Working on ery-
throphores of the same material, and after permeabilizing
the surface membrane of the cells by treatment with a Brij
surfactant, McNiven and Ward (149) found that free Ca2+
ions at 100 mM induced the aggregation of pigment,
whereas lowering of the concentration to 10 nM caused
Pigment Cell Res. 13, 2000314
8/11/2019 2000-Fujii The regulation of motile activity in fish chromatophores.pdf
16/20
dispersion. Both processes were ATP-dependent but cAMP-
independent. Furthermore, Oshima et al. (144) recently
showed that K+-induced aggregation of pigment in xan-
thophores of medaka is accompanied by Ca2+ entry in the
cytosol. One should remember, however, that chroma-
tophores represented by squirrelfish erythrophores and
medaka xanthophores are rather peculiar.
Fujii and Taguchi (70) showed that external Ca2+ ions
were not required for movement of pigment per se. Subse-
quently, the absence of a requirement for extracellular Ca2+
ions has been demonstrated in many types of chroma-
tophores (3, 14, 42, 87, 150, 151). By electron-microscopiccytochemical localization of Ca2+ ions, Negishi and Obika
(109) showed that in melanophores of medaka, an increase
in the cytosolic level of Ca2+ ions was associated with the
aggregation of melanosomes in the perikaryon, whereas the
Ca2+ level was much lower in cells with dispersed pigment.
Employing fluorescent Ca2+ indicators, Oshima et al. (110)
found that an increase in the intracellular level of free Ca 2+
ions occurred after NE stimulation of platyfish (X. macula-
tus) melanophores, which had been dissociated and sus-
pended in saline. Since the possible dynamics of Ca2+ ions,
calmodulin, cyclic nucleotide phosphodiesterase, etc., have
been recently reviewed by us (14) and also by Nery and
Castrucci (88), further explanation is not given here.
Confirming the irrelevance of extracellular Ca2+ ions in a
study of tilapia (O. niloticus) melanophores in culture, Fujii
et al. (87) recently concluded that D-myo-inositol 1,4,5-
trisphosphate (IP3) functions as another second messenger
for the aggregation of pigment (Fig. 1). It may take part in
transducing adrenergic signals via 1-adrenoceptors at least.
IP3 is generally known to cause the liberation of Ca2+ ions
from their intracellular storage compartments. In
melanophores too, IP3may act via the release of Ca2+ ions
from such compartments within the cell. Elements of
smooth endoplasmic reticulum are potent candidates, since
such compartments exist abundantly in the cytoplasm (3,
152). Readers interested in the signaling mode of IP3should
refer to the original descriptions (87) or to later explanations(3, 14, 88).
Although cAMP may be the major second messenger in
chromatophores, both the cAMP and IP3-Ca2+ systems
probably interact cooperatively to move chromatosomes. It
is suggested that slower responses are mediated by decreases
in levels of cyclic AMP, while faster ones are realized by the
IP3-Ca2+ system.
CONCLUSION AND PERSPECTIVES
It must be astonishing for many readers to know that in
fish, so many hormonal, neural and even paracrine factors
are involved in the regulation of motile activities of chroma-
tophores in the skin. Naturally, those chromatophores are
endowed with various receptors and other devices for receiv-
ing numerous cues, either from intrinsic or from external
sources. The author believes that several additional novel
principles may be found in the near future that regulate
chromatophore motility, necessitating repeated additions of
sections in forthcoming review articles in relevant fields.
Among chromatophores, melanophores usually play the
most important part in generating the remarkable and yet
subtle changes in hues or shades, as well as in color patterns.
Thus melanophores, among several types of chroma-
tophores, usually possess more species of receptors than
other cells do. It is interesting to point out that
melanophores are the closest homologues to melanocytes of
homeotherms, to which humans belong. Without a system
for cellular motility, the activities of melanocytes are con-
trolled in a simpler manner. It may safely be said therefore,
that fish possess a much more sophisticated chromatic sys-
tem than we do. We now know that fish and mammalsstarted separate ways of evolution more than 400 million
years ago. In mammals, the pigment cell system may have
become simplified, in other words, they have devoluted. For
example, no definite role of the nervous system in regulating
melanocyte function has been proven.
Why have fish evolved to possess chromatophores with so
many kinds of receptors to sense signals? The answer can be
sought in the crucial roles of chromatophores in survival
strategies of the animals. In their habitat, they actually
employ variously defined types of colorations and patterns
such as for cryptic and aposematic purposes (2). It should
be emphasized here that such colors and patterns are very
often changeable, and function to cope with various etho-
logical stressors. In order to actualize such abilities,
exquisitely fine-tuned mechanisms for controlling chroma-
tophores have thus evolved.
The molecular structures of simple hormonal substances,
such as catecholamines, MT and even smaller peptides such
as -MSH, have been well conserved both in Pisces and in
Mammalia, although their physiological assignments are
more or less modified. As for the larger peptides, sequences
and even the number of amino acids are fairly different
between the two classes of vertebrates. It is interesting to
point out here that even MSHs show considerable molecular
diversity, especially among more anciently emerged fish, a
fact that should be due to their longer phylogenetic history
of evolution. Different from the situation in mammals whichis much more advanced, the cloning of receptor peptides on
fish pigment cells still remains mostly unexplored, the sole
result obtained hitherto being an 2-adrenoceptor on
melanophores of a teleostean species L. ossifagus (107). It is
evident, however, that, having a larger number of amino
acids, receptor molecules should have diverged beyond our
expectations.
In past studies classifying receptors for fish chroma-
tophores, pharmacological analyses have been the main
approach. In those investigations, agonists and/or antago-
nists for receptors developed for therapeutic uses in humans
have mostly been employed. For rough categorization, such
approaches are appropriate, and have certainly been fruitful.
However, we should now consider that, since they separated
long ago, receptors in fish and mammals constitute two
distinct groups, each of which have diverged independently.
Therefore, the structures, as well as the pharmac