Physiological color change is important for background matching, thermoregulation as well as signaling and is in vertebrates mediated by synchronous intracellular transport of pigmented organelles in chromatophores. We describe functions of and animal situations where color change occurs. A summary of endogenous and external factors that regulate this color change in fish and amphibians is provided, with special emphasis on extracellular stimuli. We describe not only color change in skin, but also highlight studies on color change that occurs using chromatophores in other areas such as iris and on the inside of the body. In addition, we discuss the growing field that applies melanophores and skin color in toxicology and as biosensors, and point out research areas with future potential.
Research on a variety of animal species, ranging from insects to different vertebrates, has shown that body colors and patterns are traits under strong selection pressures. A classic and significant experiment on guppies indicated that color patterns often evolve as adaptations to environmental surroundings, in which it frequently resulted in cryptic coloration. Moreover, sexual selection acting on color and patterns can instead lead to colorful displays and also more diverse coloration patterns (Endler, 1980). In later studies using cichlid fish, color dimorphism has been found to be an important cue for non-random mating and is also related to establishment of reproductive isolation resulting in speciation (Wagner et al., 2012), and color polymorphisms in frogs have been linked to reduced extinction risks at species level (Forsman and Hagman, 2009). Examples across the animal kingdom show the importance of cryptic colors and patterns as an anti-predator strategy (Ryer et al., 2008; Cook et al., 2012). In lower vertebrates such as fish, different color morphs may however not be very fixed because body coloration can be manipulated at the individual level both by morphological (Sugimoto, 2002; Leclercq et al., 2010) and physiological color change (Aspengren et al., 2009a,b). In many aspects, there is still a lack of information concerning the effects of color change for animal behavior and evolution in particular. The power of phenotypic plasticity in terms of color change rather than genetic adaptations was in this context a surprising explanation behind the diversity of colors among the largely cryptic sculpins in different areas of Alaska (Whiteley et al., 2009). The ability to change color may thus have a suppressive effect on selection by allowing a larger niche. A comparative study on dwarf chameleons indicate that that social signaling rather than camouflage can be a driving force behind the evolution of color change (Stuart-Fox and Moussalli, 2008), but seemingly social colors can also be cryptic depending on the distance, light, and vision of the observer (Marshall and Johnsen, 2011). These and many more studies show the pivotal but also complex roles of animal coloration and patterning in animal fitness and evolution.
Body coloration in many poikilothermic animals is plastic and can be adjusted at the individual level. Particularly in fish, but also to some extent in frogs and reptiles, this can occur rapidly (Sumner, 1940). This rapid change is called physiological color change and refers to synchronous movement of pigment organelles within pigmented cells in the skin called chromatophores (black melanophores containing melanin, yellow xantophores containing pteridine, red erythrophores containing carotenoids, and the more rare blue cyanophores containing an unknown cyan biochrome), as well as in changes in angles of light reflecting crystals in iridophores and leucophores (Fujii and Oshima, 1994; Fujii, 2000; Aspengren et al., 2009a). When aggregating the dark melanosomes of the melanophores, the skin not only becomes pale but also more transparent. Increased body transparency can also aid background matching. The advantage of rapid color change is obvious, because it allows rapid adjustments and flexibility at the individual level depending of the situation. It is used for background matching (Sumner, 1911; Ramachandran et al., 1996; Kelman et al., 2006) as well as for communication and sexual display (Morris et al., 1994; Fuller and Berglund, 1996; Kodric-Brown, 1998; Monteiro et al., 2002; Sköld et al., 2008). Studies on pipefish have shown that the color ornaments of the females are shut off if a predator enters the mating area, clearly indicating the advantage of an adjustable body appearance and the risk of a colorful display (Fuller and Berglund, 1996). In cichlids of the species Astronotus ocellatus, males defeated in combat become black with white barring, which signals suppression (Beeching, 1995). Similarly, juveniles of Arctic charr with low social status in the group also become dark, but only when kept in pale tanks (Höglund et al., 2002). Individuals kept in aquaculture or captivity, have been observed to change color when stressed. For some species such fish become pale (Van der Salm et al., 2006), whereas others stand out as unusually dark and with reduced fitness (Ruane et al., 2005). A study on cultured sole showed that adjustment to sand by developing a light skin took several days and adjustment of color intensity and hue took several weeks (Ellis et al., 1997). A potentially useful application from this study is to train or preadjust reared fish that will be released into the wild for restoration attempts.
While frogs and reptiles usually change colors in the time frame of an hour or more, some fish can adjust in minutes (Kelman et al., 2006; Rhodes and Schlupp, 2012). This difference in speed is partly attributed to neurohumoral regulation of fish chromatophores, whereas frogs use hormones for this purpose (Sumner, 1940; Fujii, 2000), but there are also differences at the intracellular level where fish chromatophores show smaller, better coordinated, and higher speed of the pigment organelles compared with frogs (Aspengren et al., 2009a,b). If the reflective iridophores or leucophores are involved, the color change in some fish can occur in or even within seconds (Oshima and Fujii, 1987; Nagaishi and Oshima, 1989; Mäthger et al., 2003). These cells may also be regulated by light directly (Kasai and Oshima, 2006). Very rapid color change has been observed for several fishes (Ramachandran et al., 1996; Kodric-Brown, 1998; Mäthger et al., 2003), but actual time measurements or data on mechanisms are often missing. Reflective changes in iridophores or leucophores may however not be as energy demanding as the more long-distance translocations of the other pigment organelles, being potentially both efficient and cost effective. In the majority of fishes with blue-green hue, there are no pigment cells with these colors. Instead, green or blue is the result of iridophores or leucophores overlaying xanthophores and melanphores in the skin (Fujii et al., 1989). Frequent short-term oscillations in dark and yellow body coloration were reported for the blackspotted rockskipper, but this phenomenon is rather unusual for most species (Heflin et al., 2009). A potential benefit of such oscillation may be priming of the chromotophores (Burton and O'Driscoll, 1992; Hatamoto and Shingyoji, 2008). In general, although the social context of color change in some fish species, as well as in reptiles (Stuart-Fox and Moussalli, 2011), has received some scientific attention, there are still remarkably few such studies on amphibians. However, in a recent study on the tropical toad Bufo luetkenii, males were shown to be bright lemon yellow before mating and then turned cryptic brownish similar to the female during and post-pairing (Doucet and Mennill, 2010). The authors also observed that the males lost their yellow color soon after capture, possibly due to stress. In the monography by Cott (1940), which describes a range of color-changing species, their appearances, and situations where color change occurs, special attention is given to the tree-frog Phyllomedusa hypochondrialis. This little frog can apparently change body coloration to green, gray, and brown or mottled to fit the surroundings, or with the legs turning black and purple at night when it is active. As discussed by Stuart-Fox and Moussalli (2011), while the ability to change color allows plasticity in behavior and mobility between environments, there may also be costs and trade-offs related to colors and color change. This is a relevant question for physiological color change that depends on either energy driven pigment organelle translocations as in vertebrates and insects or muscle contractions as in cephalopods (Aspengren et al., 2009a; Stevens and Merilaita, 2011). A difference between terrestrial animals and fish is that color change is involved also in thermoregulation in the formers (Stuart-Fox and Moussalli, 2011). For frogs, there may therefore be more complicated compromises in the choice of body color than for fish.
While body coloration has received enormous scientific interest, chromatophores are also located elsewhere in the body (Moresco and De Oliveira, 2009; Figure 1), where they have been reported to sometimes change color. The eyes of fish are, for example, often spectacularly colored, either similar to the colors and patterns of the body but sometimes also in contrasting colors (Figure 1A). A study on guppy juveniles describes that certain eye color patterns correlated with aggressive behaviors and that the color also changed rather rapidly during the observations of the fish (Martin and Hengstebeck, 1981). Guppies also show a blackening of their iris when they are stressed by presence of a predator (Magurran and Seghers, 1991). Later studies on Nile tilapia and juvenile salmons showed that eye color was associated with social status and used in communication (Suter and Huntingford, 2002; Volpato et al., 2003). Also in these studies, the authors reported that the eye color sometimes changed. In other studies, a striking dark bar across the eyes was sometimes observed in males of Astatotilapia burtoni, and this appearance depended on social rang and male–male conflicts (Muske and Fernald, 1987a,b). Another peculiar phenomenon that has not been thoroughly studied to date is that also internal fish chromatophores sometimes respond to stimuli. To our knowledge, this was first described by Levina and Gordon (1983) when they investigated effects of MSH (melanocyte-stimulating hormone) and enkephalin on various zebrafish transplants, such as the peritoneum (the epithelia layer that covers the abdominal cavity). Using abdominal biopsies of female 2-spotted gobies that contained the peritoneum, the colorful inside of this species was shown to respond to a variety of hormones and transmitter substances very similar to the reaction on the dermal side of the same biopsies (Sköld et al., 2008). In a living goby, Goda and Fujii (1996) reported background adaptation of internal melanophores present on the peritoneum and around the vertebrae. Also in gobies, the meninx that covers the brain contains melanophores that were shown to change color as a response to stress, and particularly so in females (Gibson et al., 2009). In most, if not all, of these investigated species, the fish of the study is rather transparent, and the internal color is to a degree visible also from the outside. Using eight different species with different levels of body transparency, it was clearly shown that the capacity for internal color change is indeed related to how transparent the fish is (Nilsson Sköld et al., 2010). In the frog Eupemphix nattereri, melanin pigmentation on testes increased rapidly (approximately 2 h after inoculation) in response to the exposure to bactericidal lipopolysaccharides (LPS) and then decreased to control levels 24 h post-inoculation (Franco-Belussi and de Oliveira, 2011). This result indicates that rapid internal color change can occur also in frogs. It also further confirms a role of internal melanin and melanocytes in innate immunity, as reviewed by Aspengren et al. (2009a,b).
Factors mediating color change in fish and amphibians
The regulatory mechanisms and genetics behind vertebrate coloration and pattern formation are starting to become clear in many cases from the work of mutational screens and investigations of domestic species. For information on these mechanisms, we refer to excellent review and studies on this subject (Kelsh et al., 1996, 2009; Mills and Patterson, 2009). A master regulator of skin darkness is MSH. The melanocortin-1 receptor (MC1R), which binds MSH and ACTH, has been found to be responsible for many adaptive color variants in mammals and higher vertebrates (Rosenblum et al., 2004). A similar pivotal role of MCIR in adaptive evolution of fish pigmentation seems however not to be the case (but see, Gross et al., 2009), and the genetics behind natural color morph variants in fish remain largely unknown (Henning et al., 2010). Nevertheless, α-MSH is commonly involved in morphological color change as well as in the rapid color change of fish that result in skin darkening (Höglund et al., 2002; Sugimoto, 2002). Several fish studies point out the nuptial red/orange coloration effects of prolactin, and estrogen which has long-term coloration effects may be responsible for its release in vivo (see Sköld et al., 2008 for a discussion). Prolactin also induced bright yellow-green coloration in adults of Rana pipiens, but that seemed to be mediated by morphological color change effects by increase in pigments (Brown, 1976). Important regulators for skin paling in fish are noradrenaline and MCH (melanocyte concentrating hormone). Melatonin is not effective in all investigated fish species, but interesting because it is not only a circadian hormone but also a strong antioxidant, which as such probably deserves more attention in research on regulation of condition-dependent color ornaments (Tan et al., 2010). The regulation of short-term oscillation of body coloration that was described for the rock skipper remains unknown (Heflin et al., 2009). Spontaneous reversal after hyperdispersion is described for erythrophores on isolated fish scales (Nilsson, 2000) and in isolated melanophores where the motor protein dynein which is responsible for melanosome aggregation was inhibited (Sköld et al., 2002). It is therefore possible that these oscillations can be regulated by intracellular feedback mechanisms and/or at the receptor level through multiple ligands (Fujii, 2000). The priming effect of repeated color change responses is also interesting as it appears to be coupled to memory in the fish and may thus have potential as an assay for such functions (Hatamoto and Shingyoji, 2008).
In contrast to fish, physiological color change in amphibians is considered to rely mostly on hormonal control, and this is probably a main reason for the longer response times in these animals compared with fish (Aspengren et al., 2009a,b). The most conspicuous vertebrate melanophore-dispersing agent, α-MSH, plays a pivotal role in amphibian skin color adaptation. Interestingly, amphibian skin contains a number of regulatory peptides that stimulate α-MSH release from the pars intermedia of the pituitary, suggesting a regulatory loop between the pars intermedia and the skin (Vaudry et al., 1999). Multiple external stimuli such as light, temperature, and starvation/feeding converge on neuroendocrine melanotropic cells, and a change in plasticity results in activation or inhibition of α-MSH secretion (Roubos et al., 2010), resulting in color change. The key player in pigment aggregation in amphibian melanophores is melatonin. In addition to controlling skin coloration, melatonin regulates various aspects of physiology and behavior, including immunological functions and circadian responses (Sugden et al., 2004). Noradrenaline, a strong pigment aggregating hormone and neurotransmitter in most fish species, can induce both aggregation and dispersion in amphibian melanophores, depending on species and receptor type. α-adrenoceptors dominate in the skin of Rana pipiens, accounting for a paling reaction, while activation of β-adrenoceptors in the skin of Xenopus laevis results in pigment dispersion (Salim and Ali, 2011). Perhaps particularly for amphibians, there is a great deal of variation in the response of vertebrate melanophores to different stimuli, not only between species but also within individuals. Concerning morphological color change, which is not covered in this review, steroids have a pivotal role in nuptial and adult coloration patterns. Noteworthy is that frog melanophores in skin biopsies can respond directly by pigment dispersion to some steroids (Himes and Hadley, 1971). This seems not to be the case for fish (Fujii, 2000; Sköld et al., 2008). It shall also be noted that much more is known about melanophore regulation in amphibians than for the other chromatophore types. For lists of factors regulating physiological color change in fish and amphibians, see Tables 1 and 2, respectively.
Table 1. Factors mediating color change in fish
Color and/or pigment cell response
Dark. Dispersion of melanophore erythrophore and xanthophores
However, while most in vitro studies have investigated one factor at a time, this is unlikely to be what happens in the animal. It is much more probable that multiple factors act together, but there are few studies on the effects of mixtures. In our work on the 2-spotted goby, we found that combining noradrenaline with MSH gave a pale yellowish skin, whereas the combination melatonin and MSH or prolactin gave a pale but red skin (Sköld et al., 2008). Such different effect was a surprise because treatment with only noradrenalin or melatonin as single factors gives very similar paling effects.
Regulation of internal versus external chromatophores appears similar in fish when it comes to which hormones that mediate the color change (Sköld et al., 2008). However, genetic studies on mice indicate that the external and internal pigment cells may not be exactly the same because different factors are involved in their development and maintenance (Aoki et al., 2009). This difference is also the case for chromatophores in iris, as evident from genetic screens where some mutations affect cutaneous melanophores but not those in the iris (Kelsh et al., 1996). The lack of a black bar across the eyes of male Astatotilapia burtoni was correlated to differences in the level of responsiveness to noradrenaline (Muske and Fernald, 1987b). Remarkably, little is otherwise known about what other possible factors, such as hormones and transmitter substances that can mediate physiological color change in eyes. Concerning the internal chromatophores of the 2-spotted goby, the cells of the peritoneum respond to melatonin, MSH, prolactin, and noradrenaline. A study on peritoneum in zebrafish shows that also enkephalin can regulate internal melanophores (Levina and Gordon, 1983).
Color change in toxicology and bio-detection
Melanin is not only a pigment but known to be able to act as an antioxidant as well as to produce free radicals upon UV irradiation (Aspengren et al., 2009a,b). It can serve as a metal ion sink as well as a reservoir for the homeostasis of metal ions such as calcium and zinc (Hong and Simon, 2007). Many of the intermediates formed during melanogenesis are toxic and failure to enclose reactive intermediates can result in cytotoxicity, but also autoimmunological diseases in mammals. More than 120 pigment genes have been identified, and dysfunction or loss of melanosome proteins are the causes of many hereditary pigment diseases that affect the color of the skin, hair or eyes either directly or indirectly. Much of this knowledge comes from studies on mammals, which is not covered in this review, but the increased focus on the function, genetics and also social consequences of especially human pigmentation has relevance for lower vertebrates in many ways. Pharmaceuticals and different toxins can affect pigmentation, and many substances are offered on the market to either increase or decrease pigmentation. One example of the latter is hydroquinone, a skin-lightening agent, which we recently have shown not only to affect intracellular transport of melanosomes in X. laevis melanophores (Aspengren et al., 2012) but also to induce bundling of microtubules and disassembly of actin filaments at lower concentrations than what previously has been stated as cytotoxic. This means that we are increasing the amount of substances in the environment that may pose threats to water supplies and that might affect fish and frog coloration, thereby affecting their ability to avoid predators or to communicate. There is therefore a great need for bio-detectors based on function to provide early warnings regarding the presence of toxic substances in the environment (e.g., Iuga et al., 2008), but also for studies of potential pharmacological and pharmaceutical agents from for example plants (e.g., Ali and Meitei, 2011).
The ease of achieving a combination of morphological and functional transport studies of pigment organelles in fish and frog chromatophores has made it possible to use cultured chromatophore cells as a bio-detector. Melanophores from X. laevis can be cultivated in microtiter plates and the change in optical density of the color change reaction easily monitored using a microplate reader (Figure 2). Image analysis by light microscopy reveals important morphological data. Microtubules and actin filaments are evolutionary well conserved proteins (due to their important roles in eukaryotic cells) and using immunocytochemistry together with optical detection of pigment transport (Aspengren et al., 2006, 2012; Hedberg and Wallin, 2010) predictions can be drawn about possible toxicological effects even in other cells and species. Such studies are preferentially made on aggregated cells because melanosomes quench the fluorescence signals from stained microtubules and actin filaments. Isolated fish chromatophores have been used as detector for many different substances and several environmental toxins, bacterial pathogens, neurotransmitters, and cell membrane effectors caused changes in pigment granule distribution (Chaplen et al., 2002; Dierksen et al., 2004; Mojovic et al., 2004; Sharma et al., 2005; Dukovcic et al., 2010). They have also been used to study the effects of chemicals such as odorants (Karlsson et al., 1994), the estrogenic endocrine disruptor nonylphenol (Park et al., 2010), ethanol (Peng et al., 2009), and resveratrol (Galgut and Ali, 2011). Studies on chromatophores can furthermore be performed in a more complex situation in situ. Colanesi et al. (2012) used cultures of zebrafish embryos to perform a screen for small molecules that alter chromatophore development, which both confirmed earlier studies but also provided new insight into developmental processes that also could be of importance for testing of drugs. Fish scales with chromatophores can be maintained in vitro and were recently used for toxicological testing of a novel antifouling boat-paint (Lennquist et al., 2010). Further development is ongoing with a combination of cell science, bioengineering, and information technologies. Karlsson et al. (2002) took advantage of the presence of G-protein-coupled receptors in frog melanophores that are coupled to melanosome transport by expression of a human opioid receptor and could thereby use the melanophores as a specific biosensor for the detection of opioids in body fluids. Bioterrorism is increasing and detection platforms for rapid detection and confirmation of biothreat agents that are portable, user-friendly and testing multiple agents simultaneously are needed to protect the public health (Lim et al., 2005). The need for such platforms will also be useful from the perspective of environmental protection of non-mammals such as fish and frogs.
Chromatophores are in the interesting interphase between cell physiology, visual perception, and animal behavior. Given the visible color of these cells, they have provided classic and most useful model systems for fundamental understanding of intracellular transport mechanisms. This information has been of importance for the development of the melanophores and skin pigmentation in ecotoxicology and as biosensors. Further fundamental research is now needed to reveal the regulation and functions of the largely neglected extracutaneous pigmentary systems, such as the chromatophores of the eyes. The very rapid color and pattern changes of some teleost fishes deserve more attention too. Such color change involves direct or neural control of various types of chromatophores, and the mechanisms at different levels are not fully understood. There is also potential for ethologists and sensory ecologists to determine the behavioral functions of rapid color change, including costs and conflicting appearances, and such information would shed light also on general functions of coloration and patterns. Lessons from the factors presented in this review that mediate rapid color change have potential for investigations on animal behavior, appearance and condition-dependent colorations. This also includes combining factors, mimicking the real situation. Research that includes physiological color change with genetics in species with multiple color morphs would also provide important insights into effects of individual plasticity on adaptation and evolution. We conclude that the future is bright for research on animal pigmentation and chromatophores.