Jérémie TeyssierpersName, Suzanne V. SaenkopersName, Dirk van der MarelpersName, Michel C. MilinkovitchpersName (2015date). “Photonic crystals cause active colour change in chameleons.” Nature Communicationstitle·j.
Object of study — the mechanism of rapid active colour changeterm·key in panther chameleonterm·key skin
Research question — whether modulating the spacing of the guanine nanocrystalterm lattice explains chameleon colour changeterm·key
Contribution — two superposed iridophoreterm·key layers: a tunable photonic-crystal layer for colour changeterm·key and a deep broad-band near-infraredterm·key reflector
phenomenon · claim-type · cross-domain-transfer · mechanism · claim-relation · replication-status · limitation — grounded in the full text| phenomenon | rapid active colour change in chameleons — “Many chameleons, and panther chameleons in particular, have the remarkable ability to exhibit complex and rapid colour changes during social interactions such as male contests or courtship.” | |
| claim-type | mechanism — “Hence, expansion/contraction of the photonic crystal lattice in S -iridophores is sufficient to explain the reversible shifts of colours observed in vivo .” | |
| method — evidence | — “Here, combining microscopy, photometric videography and photonic band-gap modelling, we show that chameleons shift colour through active tuning of a lattice of guanine nanocrystals within a superficial thick layer of dermal iridophores.” | |
| object of study — evidence | — “Here, combining microscopy, photometric videography and photonic band-gap modelling, we show that chameleons shift colour through active tuning of a lattice of guanine nanocrystals within a superficial thick layer of dermal iridophores.” | |
| research question — evidence | — “To test whether indeed period modulation of the lattice of guanine nanocrystals explains colour change in chameleons, we subjected samples of excited skin (white/yellow) to hypertonic solutions to generate osmotic pressure likely to cause the guanine crystal lattice to shrink to its resting state.” | |
| contribution — evidence | — “The organization of iridophores into two superposed layers constitutes an evolutionary novelty for chameleons, which allows some species to combine efficient camouflage with spectacular display, while potentially providing passive thermal protection.” | |
| mechanism | lattice spacing tunes reflectivity Wider spacing between nanocrystals in excited males shifts reflectivity to longer wavelengths, turning green skin yellow or orange. — “Hence, it is likely to be that the increase in mean distance among nanocrystals in excited male panther chameleons causes S -iridophores to shift their selective reflectivity from short (blue) to long (red or infrared) wavelengths, causing the corresponding shift from green to yellow/orange skin.” | mechanisms[0] |
| mechanism | passive thermal protection The deep iridophore layer may protect against heat by reflecting near-infrared sunlight. — “Hence, the thick layer of D -iridophores has the potential to play in some species, such as the panther chameleon, a substantial role in thermal protection.” | mechanisms[1] |
| sense | structural colour - colour made by interference in multilayer nano-reflectors, as opposed to pigments; here the tunable guanine-crystal lattice — “On the other hand, rapid active tuning of skin hue has been described in only a handful of species and generally involves structural, rather than pigmentary, components, that is, multilayer nano-reflectors with alternating high and low refractive indices that generate interference of light waves.” | senses[0] |
| sense | photonic crystal - used for the triangular lattice of guanine nanocrystals in the superficial iridophores — “Hence, expansion/contraction of the photonic crystal lattice in S -iridophores is sufficient to explain the reversible shifts of colours observed in vivo .” | senses[1] |
| relation | guanine nanocrystals —part-of→ dermal iridophores — within — “Here, combining microscopy, photometric videography and photonic band-gap modelling, we show that chameleons shift colour through active tuning of a lattice of guanine nanocrystals within a superficial thick layer of dermal iridophores.” — “a lattice of guanine nanocrystals within a superficial thick layer of dermal iridophores” | concept_edges[0] |
| weight | guanine nanocrystals —part-of→ dermal iridophores — 0.85 — “Here, combining microscopy, photometric videography and photonic band-gap modelling, we show that chameleons shift colour through active tuning of a lattice of guanine nanocrystals within a superficial thick layer of dermal iridophores.” | concept_edges[0] |
| relation | TEM —same-as→ transmission electron microscopy — (TEM) — “Our histological and transmission electron microscopy (TEM) analyses in five adult males, four adult females and four juveniles revealed that, unlike other lizards 7 14 18 the skin of panther chameleons consists of two superposed thick layers of iridophore cells ( Fig. 1c ) containing guanine crystals of different sizes, shapes and organizations ( Fig. 1d,e ).” — “transmission electron microscopy (TEM)” | concept_edges[1] |
| weight | TEM —same-as→ transmission electron microscopy — 0.95 — “Our histological and transmission electron microscopy (TEM) analyses in five adult males, four adult females and four juveniles revealed that, unlike other lizards 7 14 18 the skin of panther chameleons consists of two superposed thick layers of iridophore cells ( Fig. 1c ) containing guanine crystals of different sizes, shapes and organizations ( Fig. 1d,e ).” | concept_edges[1] |
| relation | xanthophores —same-as→ chromatophores with yellow pigments — (xanthophores — “As in other lizards 14 , the green background skin of panther chameleons contains chromatophores with yellow pigments (xanthophores, Supplementary Fig. 1a ).” — “chromatophores with yellow pigments (xanthophores, Supplementary Fig. 1a )” | concept_edges[2] |
| weight | xanthophores —same-as→ chromatophores with yellow pigments — 0.9 — “As in other lizards 14 , the green background skin of panther chameleons contains chromatophores with yellow pigments (xanthophores, Supplementary Fig. 1a ).” | concept_edges[2] |
| relation | erythrophores —same-as→ red-pigment chromatophores — (that is, erythrophores — “Note that in red skin, the upper layer of S -iridophores remains well developed but a large proportion of iridophores is replaced by red-pigment chromatophores (that is, erythrophores; Supplementary Fig. 1a ), explaining that red skin hue does not change dramatically during excitation, but its brightness increases.” — “red-pigment chromatophores (that is, erythrophores; Supplementary Fig. 1a )” | concept_edges[3] |
| weight | erythrophores —same-as→ red-pigment chromatophores — 0.9 — “Note that in red skin, the upper layer of S -iridophores remains well developed but a large proportion of iridophores is replaced by red-pigment chromatophores (that is, erythrophores; Supplementary Fig. 1a ), explaining that red skin hue does not change dramatically during excitation, but its brightness increases.” | concept_edges[3] |
| relation | RGB —same-as→ red–green–blue — (RGB) — “We used red–green–blue (RGB) photometry on high-resolution videos ( Supplementary Fig. 2 ) to analyse in vivo the optical response of the skin of male F. pardalis individuals during male–male contests.” — “red–green–blue (RGB)” | concept_edges[4] |
| weight | RGB —same-as→ red–green–blue — 0.95 — “We used red–green–blue (RGB) photometry on high-resolution videos ( Supplementary Fig. 2 ) to analyse in vivo the optical response of the skin of male F. pardalis individuals during male–male contests.” | concept_edges[4] |
| relation | resting state —contrasted-with→ excited state — smaller ... than — “Comparing multiple TEM images from F. pardalis samples of blue or green skin (resting state) with those obtained on yellow or white skin (excited state) of the same individuals, we find that crystal size in S -iridophores does not vary, but the distance among guanine crystals is on average 30% smaller in the resting than in the excited skin ( Fig. 2a and Supplementary Table 1 ).” — “30% smaller in the resting than in the excited skin” | concept_edges[5] |
| weight | resting state —contrasted-with→ excited state — 0.85 — “Comparing multiple TEM images from F. pardalis samples of blue or green skin (resting state) with those obtained on yellow or white skin (excited state) of the same individuals, we find that crystal size in S -iridophores does not vary, but the distance among guanine crystals is on average 30% smaller in the resting than in the excited skin ( Fig. 2a and Supplementary Table 1 ).” | concept_edges[5] |
| effect direction | reflectivity —causes→ absorption of sunlight — reduces — “Our measurements indicate that the reflectivity ( R ) in the near-infrared region (700–1,400 nm) is particularly high ( Fig. 3b ), causing a substantial decrease in the absorption of sunlight.” — “causing a substantial decrease in the absorption of sunlight” | concept_edges[6] |
| relation | reflectivity —causes→ absorption of sunlight — causing — “Our measurements indicate that the reflectivity ( R ) in the near-infrared region (700–1,400 nm) is particularly high ( Fig. 3b ), causing a substantial decrease in the absorption of sunlight.” — “is particularly high ( Fig. 3b ), causing a substantial decrease in the absorption of sunlight” | concept_edges[6] |
| weight | reflectivity —causes→ absorption of sunlight — 0.8 — “Our measurements indicate that the reflectivity ( R ) in the near-infrared region (700–1,400 nm) is particularly high ( Fig. 3b ), causing a substantial decrease in the absorption of sunlight.” | concept_edges[6] |
| relation | D -iridophore layer —is-a→ broad-band reflector — is a — “This shows that the D -iridophore layer is a broad-band reflector in the near infrared region, as the power spectrum is essentially a step function going from 0 below 400 nm to a plateau above 900 nm.” — “the D -iridophore layer is a broad-band reflector” | concept_edges[7] |
| weight | D -iridophore layer —is-a→ broad-band reflector — 0.8 — “This shows that the D -iridophore layer is a broad-band reflector in the near infrared region, as the power spectrum is essentially a step function going from 0 below 400 nm to a plateau above 900 nm.” | concept_edges[7] |
| evidential status | mechanism — it is likely to be — “Hence, it is likely to be that the increase in mean distance among nanocrystals in excited male panther chameleons causes S -iridophores to shift their selective reflectivity from short (blue) to long (red or infrared) wavelengths, causing the corresponding shift from green to yellow/orange skin.” | mechanisms[0] |
| evidential status | mechanism — has the potential — “Hence, the thick layer of D -iridophores has the potential to play in some species, such as the panther chameleon, a substantial role in thermal protection.” | mechanisms[1] |
| replication | original | |
| limitation | The molecular mechanism of the active tuning is not yet known. — “The molecular mechanisms involved in this process remain to be determined” | limitations[0] |
| limitation | RGB photometry does not record ultraviolet reflectivity. — “Skin reflectivity in the ultraviolet range, highly relevant for colour perception in reptiles 37 38 , is not recorded by RGB photometry.” | limitations[1] |
| limitation | Disordered deep crystals prevent more rigorous modelling. — “Note that the disorder of guanine crystals inside D -iridophores prevents the use of more rigorous modelling.” | limitations[2] |
| limitation | Whether chameleons reflect more near-infrared than other lizards needs comparative study. — “Comparative analyses with similar measurements in chameleonid and non-chameleonid species (for example, see Supplementary Fig. 4 and refs 31 , 32 ) is warranted to identify whether reflectivity in the near-infrared range is substantially and systematically higher in chameleons than in other lizards.” | limitations[3] |
| limitation | The ancestral function of the deep layer may not be thermal protection. — “However, the ancestral function of D -iridophores might not be associated with passive thermal protection, because extant species of the basal lineages in the phylogeny of chameleons 34 are dense-forest dwellers” | limitations[4] |
finding · quantity · variable · definition · key-term — grounded in the full text| finding | Chameleon skin has two superposed thick iridophore layers with different guanine crystals. — “Our histological and transmission electron microscopy (TEM) analyses in five adult males, four adult females and four juveniles revealed that, unlike other lizards 7 14 18 the skin of panther chameleons consists of two superposed thick layers of iridophore cells ( Fig. 1c ) containing guanine crystals of different sizes, shapes and organizations ( Fig. 1d,e ).” | content_tags.findings[0] |
| finding | Crystal size does not change, but spacing is 30% smaller at rest than when excited. — “Comparing multiple TEM images from F. pardalis samples of blue or green skin (resting state) with those obtained on yellow or white skin (excited state) of the same individuals, we find that crystal size in S -iridophores does not vary, but the distance among guanine crystals is on average 30% smaller in the resting than in the excited skin ( Fig. 2a and Supplementary Table 1 ).” | content_tags.findings[1] |
| finding | Hypertonic treatment blue-shifts the reflectivity of superficial iridophores. — “This treatment indeed results in a blue shift in the reflectivity of S -iridophores ( Fig. 2b ).” | content_tags.findings[2] |
| finding | Modelled colours closely match those observed in vivo and in osmotic experiments. — “Reflectivity was set to unity in the gapped region and convolution with standard X , Y , Z spectral functions returned simulated colours ( Supplementary Movie 5 ) that closely match those observed in vivo ( Fig. 1b ) and during osmotic pressure experiments ( Fig. 2c ).” | content_tags.findings[3] |
| finding | The deep iridophore layer occurs in all panther chameleons and three other chameleon species. — “This population of D -iridophores is present in all panther chameleons (regardless of sex or age) and in the three distantly related chameleon species we investigated ( Figs 1e and 3a ), and is particularly thick in comparison with the layer of iridophores observed in other (non-chameleonid) lizards.” | content_tags.findings[4] |
| finding | The deep layer is a broad-band near-infrared reflector. — “This shows that the D -iridophore layer is a broad-band reflector in the near infrared region, as the power spectrum is essentially a step function going from 0 below 400 nm to a plateau above 900 nm.” | content_tags.findings[5] |
| quantity | 30% — — “Comparing multiple TEM images from F. pardalis samples of blue or green skin (resting state) with those obtained on yellow or white skin (excited state) of the same individuals, we find that crystal size in S -iridophores does not vary, but the distance among guanine crystals is on average 30% smaller in the resting than in the excited skin ( Fig. 2a and Supplementary Table 1 ).” | content_tags.quantities[0] |
| quantity | 127.4±17.8 nm — — “The upper layer is fully developed only in the skin of adult males (and reduced in the skin of female and juveniles; Supplementary Fig. 1b ) and contains iridophores (hereafter called superficial (S-) iridophores) with small close-packed guanine crystals 19 (diameter 127.4±17.8 nm; Supplementary Table 1 ) organized in a triangular lattice ( Fig. 1d ).” | content_tags.quantities[1] |
| quantity | ~45% — — “Multiplying the sun radiance 27 (blue curve in Fig. 3b ) by 1− R , to yield the amount of light transmitted by the dermis (hence absorbed by the peritoneum or deeper tissues; red curve in Fig. 3b ), indicates that ~45% of the radiation energy in that spectral range is screened in panther chameleons by reflection on the dermis.” | content_tags.quantities[2] |
| quantity | 90 million years — — “Chameleons form a highly derived monophyletic group of iguanian lizards that originated in post-Gondwanan Africa around 90 million years ago 34 35 .” | content_tags.quantities[3] |
| quantity | about 1,200 — — “Crystal diameter ( d ) and distance between centres of nearest crystals ( s ) were measured (on about 1,200 individual crystals, Supplementary Table 1 ) on TEM images as described above and the lattice parameter a was computed as s √2.” | content_tags.quantities[4] |
| variable | lattice parameter — “Crystal diameter ( d ) and distance between centres of nearest crystals ( s ) were measured (on about 1,200 individual crystals, Supplementary Table 1 ) on TEM images as described above and the lattice parameter a was computed as s √2.” | content_tags.variables[0] |
| definition | xanthophores — “As in other lizards 14 , the green background skin of panther chameleons contains chromatophores with yellow pigments (xanthophores, Supplementary Fig. 1a ).” | content_tags.definitions[0] |
| definition | erythrophores — “Note that in red skin, the upper layer of S -iridophores remains well developed but a large proportion of iridophores is replaced by red-pigment chromatophores (that is, erythrophores; Supplementary Fig. 1a ), explaining that red skin hue does not change dramatically during excitation, but its brightness increases.” | content_tags.definitions[1] |
| key-terms | photonic crystal, iridophores, guanine, panther chameleon, structural colour, colour change, near-infrared, band-gap, osmotic pressure, chromatophores |
| concept | dermal iridophores — not defined in this paper's own words; skin cells holding guanine nanocrystals — “Here, combining microscopy, photometric videography and photonic band-gap modelling, we show that chameleons shift colour through active tuning of a lattice of guanine nanocrystals within a superficial thick layer of dermal iridophores.” | concepts[0] |
| concept | guanine nanocrystals — not defined in this paper's own words — “Here, combining microscopy, photometric videography and photonic band-gap modelling, we show that chameleons shift colour through active tuning of a lattice of guanine nanocrystals within a superficial thick layer of dermal iridophores.” | concepts[1] |
| concept | physiological colour changes — generally mediated by modifications of skin brightness — “Many vertebrates can rapidly change colour for camouflage, communication and thermoregulation 2 4 5 6 7 , but these so-called physiological colour changes are generally mediated by modifications of skin brightness (that is, diffuse and/or specular reflectivity) through dispersion/aggregation of pigment-containing organelles, especially melanosomes, within dermal chromatophores 6 7 .” | concepts[2] |
| concept | skin brightness — diffuse and/or specular reflectivity — “Many vertebrates can rapidly change colour for camouflage, communication and thermoregulation 2 4 5 6 7 , but these so-called physiological colour changes are generally mediated by modifications of skin brightness (that is, diffuse and/or specular reflectivity) through dispersion/aggregation of pigment-containing organelles, especially melanosomes, within dermal chromatophores 6 7 .” | concepts[3] |
| concept | transmission electron microscopy — not defined in this paper's own words — “Our histological and transmission electron microscopy (TEM) analyses in five adult males, four adult females and four juveniles revealed that, unlike other lizards 7 14 18 the skin of panther chameleons consists of two superposed thick layers of iridophore cells ( Fig. 1c ) containing guanine crystals of different sizes, shapes and organizations ( Fig. 1d,e ).” | concepts[4] |
| concept | TEM — transmission electron microscopy — “Our histological and transmission electron microscopy (TEM) analyses in five adult males, four adult females and four juveniles revealed that, unlike other lizards 7 14 18 the skin of panther chameleons consists of two superposed thick layers of iridophore cells ( Fig. 1c ) containing guanine crystals of different sizes, shapes and organizations ( Fig. 1d,e ).” | concepts[5] |
| concept | xanthophores — chromatophores with yellow pigments — “As in other lizards 14 , the green background skin of panther chameleons contains chromatophores with yellow pigments (xanthophores, Supplementary Fig. 1a ).” | concepts[6] |
| concept | chromatophores with yellow pigments — not defined in this paper's own words — “As in other lizards 14 , the green background skin of panther chameleons contains chromatophores with yellow pigments (xanthophores, Supplementary Fig. 1a ).” | concepts[7] |
| concept | erythrophores — red-pigment chromatophores — “Note that in red skin, the upper layer of S -iridophores remains well developed but a large proportion of iridophores is replaced by red-pigment chromatophores (that is, erythrophores; Supplementary Fig. 1a ), explaining that red skin hue does not change dramatically during excitation, but its brightness increases.” | concepts[8] |
| concept | red-pigment chromatophores — not defined in this paper's own words — “Note that in red skin, the upper layer of S -iridophores remains well developed but a large proportion of iridophores is replaced by red-pigment chromatophores (that is, erythrophores; Supplementary Fig. 1a ), explaining that red skin hue does not change dramatically during excitation, but its brightness increases.” | concepts[9] |
| concept | resting state — blue or green skin — “Comparing multiple TEM images from F. pardalis samples of blue or green skin (resting state) with those obtained on yellow or white skin (excited state) of the same individuals, we find that crystal size in S -iridophores does not vary, but the distance among guanine crystals is on average 30% smaller in the resting than in the excited skin ( Fig. 2a and Supplementary Table 1 ).” | concepts[10] |
| concept | excited state — yellow or white skin — “Comparing multiple TEM images from F. pardalis samples of blue or green skin (resting state) with those obtained on yellow or white skin (excited state) of the same individuals, we find that crystal size in S -iridophores does not vary, but the distance among guanine crystals is on average 30% smaller in the resting than in the excited skin ( Fig. 2a and Supplementary Table 1 ).” | concepts[11] |
| concept | reflectivity — not defined in this paper's own words — “Our measurements indicate that the reflectivity ( R ) in the near-infrared region (700–1,400 nm) is particularly high ( Fig. 3b ), causing a substantial decrease in the absorption of sunlight.” | concepts[12] |
| concept | absorption of sunlight — not defined in this paper's own words — “Our measurements indicate that the reflectivity ( R ) in the near-infrared region (700–1,400 nm) is particularly high ( Fig. 3b ), causing a substantial decrease in the absorption of sunlight.” | concepts[13] |
| concept | D -iridophore layer — the second thick layer of iridophores — “This shows that the D -iridophore layer is a broad-band reflector in the near infrared region, as the power spectrum is essentially a step function going from 0 below 400 nm to a plateau above 900 nm.” | concepts[14] |
| concept | broad-band reflector — not defined in this paper's own words — “This shows that the D -iridophore layer is a broad-band reflector in the near infrared region, as the power spectrum is essentially a step function going from 0 below 400 nm to a plateau above 900 nm.” | concepts[15] |
| concept | RGB — red–green–blue — “We used red–green–blue (RGB) photometry on high-resolution videos ( Supplementary Fig. 2 ) to analyse in vivo the optical response of the skin of male F. pardalis individuals during male–male contests.” | concepts[16] |
| concept | red–green–blue — not defined in this paper's own words — “We used red–green–blue (RGB) photometry on high-resolution videos ( Supplementary Fig. 2 ) to analyse in vivo the optical response of the skin of male F. pardalis individuals during male–male contests.” | concepts[17] |
| concept | lattice parameter — computed as s √2 — “Crystal diameter ( d ) and distance between centres of nearest crystals ( s ) were measured (on about 1,200 individual crystals, Supplementary Table 1 ) on TEM images as described above and the lattice parameter a was computed as s √2.” | concepts[18] |
Full text — tagged
The reference list is not part of the text we hold for this paper. Our copy was extracted from the publisher's page and the bibliography did not survive that extraction, so the article below ends before its references. Nothing has been removed by us.
Photonic crystalsterm·key cause active colour changeterm·key in chameleons Abstract Many chameleons, and panther chameleonsterm·key in particular, have the remarkable ability to exhibit complex and rapid colour changesterm·key during social interactions such as male contests or courtship. It is generally interpreted that these changes are due to dispersion/aggregation of pigment-containing organelles within dermal chromatophoresterm·key. Here, combining microscopy, photometric videography and photonic band-gapterm·key modelling, we show that chameleons shift colour through active tuning of a lattice of guanine nanocrystalsterm within a superficial thick layer of dermal iridophoresterm. In addition, we show that a deeper population of iridophoresterm·key with larger crystals reflects a substantial proportion of sunlight especially in the near-infraredterm·key range. The organization of iridophoresterm·key into two superposed layers constitutes an evolutionary novelty for chameleons, which allows some species to combine efficient camouflage with spectacular display, while potentially providing passive thermal protection.
Ever since their description by Aristotle, chameleons have populated myths and legends because of a number of features such as a long projectile tongue, independently movable eyes, zygodactylous feet, a very slow pace and the striking capacity of some species to rapidly shift from one vivid colour to another 1 2 3 . Many vertebrates can rapidly change colour for camouflage, communication and thermoregulation 2 4 5 6 7 , but these so-called physiological colour changesterm·dom are generally mediated by modifications of skin brightnessterm·dom (that is, diffuse and/or specular reflectivityterm·dom) through dispersion/aggregation of pigment-containing organelles, especially melanosomes, within dermal chromatophoresterm·key 6 7 . On the other hand, rapid active tuning of skin hue has been described in only a handful of species and generally involves structural, rather than pigmentary, components, that is, multilayer nano-reflectors with alternating high and low refractive indices that generate interference of light waves. For example, some species of squid can rapidly tune skin iridescence through periodical invaginations of plasma membrane deep into specialized cells called iridophoresterm·key, generating arrays of alternating cytoplasmic protein-rich lamellae and extracellular channels 8 9 . In fish, amphibians and reptiles, iridophoresterm·key containing transparent guanine nanocrystalsterm generate a large variety of structural coloursterm·key, and modifications of the multilayer reflector geometry has been suggested to generate colour changeterm·key in a few species 10 11 12 13 14 . Finally, it must be emphasized that the colour of a reptile skin patch is often the result of interactions among pigmentary and structural elements 14 15 .
Combining histology, electron microscopy and photometric videography techniques with numerical band-gapterm·key modelling, here we show that chameleons have evolved two superimposed populations of iridophoresterm·key with different morphologies and functions: the upper multilayer is responsible for rapid structural colourterm·key change through active tuning of guanine nanocrystalterm spacing in a triangular lattice, whereas the deeper population of cells broadly reflects light, especially in the near-infraredterm·key range. This combination of two functionally different layers of iridophoresterm·key constitutes an evolutionary novelty that allows some species of chameleons to combine efficient camouflage and dramatic display, while potentially moderating the thermal consequences of intense solar radiations.
Results Colour-change abilities of panther chameleonsterm·key We study the panther chameleonterm·key ( Furcifer pardalis ), a lizard from Madagascar, capable of dynamic colour changeterm·key. Our Raman spectroscopy analyses indicate the presence of two types of dark chromatophoresterm·key containing respectively melanin 16 and an unidentified dark-blue pigment ( Supplementary Fig. 1 ). Although panther chameleonsterm·key of both sexes and all ages can strongly modulate the brightness of the skin through dispersion of these pigments (for example, in response to stress), adult males are additionally characterized both by exceptionally large intraspecific colour variation (with various combinations of white, red, green and blue skin) among geographic locales within Madagascar and their ability to rapidly change colour (hue). Indeed, when encountering a male competitor or a potentially receptive female, a mature male panther chameleonterm·key can shift the background colour of its skin ( Fig. 1a ) from green to yellow or orange, whereas blue patches turn whitish and red becomes brighter with less conspicuous hue modifications. This process occurs within a couple of minutes and is fully reversible ( Supplementary Movies 1 , 2 , 3 ).
In-vivo photometry and skin structure We used red–green–blueterm·mth (RGBterm·mth) photometry on high-resolution videos ( Supplementary Fig. 2 ) to analyse in vivo the optical response of the skin of male F. pardalis individuals during male–male contests. The time evolution of the skin colour in the CIE (International Commission on Illumination) chromaticity chart indicates a gradual spectral weight transfer from the blue to the green to the red portions of the visible electromagnetic spectrum ( Fig. 1b ). This phenomenon is difficult to explain by dispersion/aggregation of pigments within chromatophoresterm·key alone and is likely to additionally involve tuning of a structural colourterm·key mechanism such as, for example, multilayer interference 17 . Our histological and transmission electron microscopyterm·mth (TEMterm·mth) analyses in five adult males, four adult females and four juveniles revealed that, unlike other lizards 7 14 18 the skin of panther chameleonsterm·key consists of two superposed thick layers of iridophoreterm·key cells ( Fig. 1c ) containing guanineterm·key crystals of different sizes, shapes and organizations ( Fig. 1d,e ). The upper layer is fully developed only in the skin of adult males (and reduced in the skin of female and juveniles; Supplementary Fig. 1b ) and contains iridophoresterm·key (hereafter called superficial (S-) iridophoresterm·key) with small close-packed guanineterm·key crystals 19 (diameter 127.4±17.8 nm; Supplementary Table 1 ) organized in a triangular lattice ( Fig. 1d ). This arrangement of high and low refractive index materials ( n guanineterm·key =1.83, n cytoplasm =1.33; ref. 13 ) has the potential of behaving as a so-called photonic crystalterm·key 20 21 , such as those that generate bright colours in some birds and insects 22 23 24 25 . Comparing multiple TEMterm·mth images from F. pardalis samples of blue or green skin (resting stateterm·dom) with those obtained on yellow or white skin (excited stateterm·dom) of the same individuals, we find that crystal size in S -iridophores does not vary, but the distance among guanineterm·key crystals is on average 30% smaller in the resting than in the excited skin ( Fig. 2a and Supplementary Table 1 ). Given that even slight alterations of geometry in a photonic crystalterm·key can generate dramatic changes in colour 17 , we hypothesized that panther chameleonsterm·key shift from one vibrant colour to another by modifying guanineterm·key crystal spacing in their S -iridophores. As in other lizards 14 , the green background skin of panther chameleonsterm·key contains chromatophores with yellow pigmentsterm (xanthophoresterm, Supplementary Fig. 1a ). Hence, it is likely to be that the increase in mean distance among nanocrystals in excited male panther chameleonsterm·key causes S -iridophores to shift their selective reflectivityterm·dom from short (blue) to long (red or infrared) wavelengths, causing the corresponding shift from green to yellow/orange skin. Note that in red skin, the upper layer of S -iridophores remains well developed but a large proportion of iridophoresterm·key is replaced by red-pigment chromatophoresterm (that is, erythrophoresterm; Supplementary Fig. 1a ), explaining that red skin hue does not change dramatically during excitation, but its brightness increases.
Osmotic pressureterm·key experimentsterm·mth and optical modelling To test whether indeed period modulation of the lattice of guanine nanocrystalsterm explains colour changeterm·key in chameleons, we subjected samples of excited skin (white/yellow) to hypertonic solutions to generate osmotic pressureterm·key likely to cause the guanineterm·key crystal lattice to shrink to its resting stateterm·dom. This treatment indeed results in a blue shift in the reflectivityterm·dom of S -iridophores ( Fig. 2b ). Furthermore, cell tracking during increase of extracellular osmolarity indicates that individual cells experience a gradual shift in colour across the whole visible spectrum ( Fig. 2c and Supplementary Movie 4 ). Hence, expansion/contraction of the photonic crystalterm·key lattice in S -iridophores is sufficient to explain the reversible shifts of colours observed in vivo .
Using band-gapterm·key modelling of the photonic crystalterm·key optical response 25 26 , we simulated the colours generated by a face-centred cubic lattice of close-packed guanineterm·key crystals for a fixed crystal size and a range of lattice parameterterm·mth (distance) values measured on TEMterm·mth images of various excited and unexcited male panther chameleonterm·key skin samples of different colours ( Supplementary Table 1 ). The irreducible Brillouin zone was meshed ( Fig. 2d ) and the photonic band structure was computed for each vertex using block-iterative frequency-domain methods 26 ( Supplementary Fig. 3 ). As no preferential orientation of photonic crystalsterm·key relative to skin surface was observed in S -iridophores, we also computed the average among all directions. Reflectivityterm·dom was set to unity in the gapped region and convolution with standard X , Y , Z spectral functions returned simulated colours ( Supplementary Movie 5 ) that closely match those observed in vivo ( Fig. 1b ) and during osmotic pressureterm·key experimentsterm·mth ( Fig. 2c ).
Function of D -iridophores In addition, we investigated the second thick layer of iridophoresterm·key ( Fig. 1e ), hereafter called deep (D-) iridophoresterm·key, which contain larger brick-shaped and somewhat disorganized guanineterm·key crystals (length 200–600 nm, height 90–150 nm). This population of D -iridophores is present in all panther chameleonsterm·key (regardless of sex or age) and in the three distantly related chameleon species we investigated ( Figs 1e and 3a ), and is particularly thick in comparison with the layer of iridophoresterm·key observed in other (non-chameleonid) lizards. In chameleons, we never found this layer to change colour (in the visible range) during osmotic pressureterm·key experimentsterm·mth, suggesting that the main function of D -iridophores is not associated to shifts in hue. Our measurements indicate that the reflectivityterm·dom ( R ) in the near-infraredterm·key region (700–1,400 nm) is particularly high ( Fig. 3b ), causing a substantial decrease in the absorption of sunlightterm·dom. Multiplying the sun radiance 27 (blue curve in Fig. 3b ) by 1− R , to yield the amount of light transmitted by the dermis (hence absorbed by the peritoneum or deeper tissues; red curve in Fig. 3b ), indicates that ~45% of the radiation energy in that spectral range is screened in panther chameleonsterm·key by reflection on the dermis. To test whether this infrared reflectivityterm·dom is probably due to coherent scattering on guanineterm·key crystals in D -iridophores, we generated two-dimensional Fourier spectra 28 on extensive TEMterm·mth image assemblies of panther chameleonterm·key D -iridophores (see online Methods). Note that the disorder of guanineterm·key crystals inside D -iridophores prevents the use of more rigorous modelling. We then used the computed Fourier power spectrum as an estimate of the spectral shape (red curve in Fig. 3c ) of the light back-scattered by deep iridophoresterm·key. This shows that the D -iridophore layerterm is a broad-band reflectorterm·dom in the near infraredterm·key region, as the power spectrum is essentially a step function going from 0 below 400 nm to a plateau above 900 nm. Multiplying the power spectrum by the transmittance of a 150-μm-thick layer of skin 29 (identical, in this spectral range, to that of water 30 ), we produce a reflectance spectrum (green curve in Fig. 3c ) that matches the shape of the measured reflectivityterm·dom spectrum (black dashed curve in Fig. 3c ) in the range 900–2,500 nm. The match below wavelengths of 900 nm is substantially less good, as we exclusively consider the D -iridophore crystals in our Fourier power spectrum analysis, that is, we ignore pigments and S -iridophores, which both strongly influence the measured reflectivityterm·dom in the visible range. Hence, the thick layer of D -iridophores has the potential to play in some species, such as the panther chameleonterm·key, a substantial role in thermal protection. Comparative analyses with similar measurements in chameleonid and non-chameleonid species (for example, see Supplementary Fig. 4 and refs 31 , 32 ) is warranted to identify whether reflectivityterm·dom in the near-infraredterm·key range is substantially and systematically higher in chameleons than in other lizards. It is noteworthy that the iridophoresterm·key found in non-chameleonid lizards can exhibit guanineterm·key crystals with diverse sizes, shapes and organizations (some of which generate structural coloursterm·key 14 ) but are not organized into two superposed layers of functionally different iridophoresterm·key ( Fig. 3a ).
Discussion Combining experimental methods from biology and physics, as well as optical modelling, we have shown that panther chameleonsterm·key rapidly change colour (hue) by actively tuning the photonic response of a lattice of small guanine nanocrystalsterm in S -iridophores. The molecular mechanisms involved in this process remain to be determined; however, given that iridophoresterm·key share the same neural-crest origin as pigmented chromatophoresterm·key, the active tuning of guanineterm·key crystal spacing we describe here could be considered analogous to movements of pigment-containing organelles in other types of chromatophoresterm·key, possibly through similar neural or hormonal mechanisms 33 . In chameleons, these S -iridophores are positioned on the top of a second thick layer of D -iridophores, with larger, flatter and somewhat disorganized guanineterm·key crystals, which reflects a substantial proportion of direct and indirect sun radiations, especially in the near-infraredterm·key range.
Chameleons form a highly derived monophyletic group of iguanian lizards that originated in post-Gondwanan Africa around 90 million years ago 34 35 . Undoubtedly, some species of chameleons occupy quite open environments where they are exposed to high levels of sunlight. In particular, panther chameleonsterm·key and veiled chameleons (studied here) occur in dry, hot environments (Northern Madagascar and Yemen, respectively) and are highly exposed to sunlight such that the 45% decrease in sunlight absorption caused by D -iridophores ( Fig. 3b,c ) is likely to be advantageous for survival. However, the ancestral function of D -iridophores might not be associated with passive thermal protection, because extant species of the basal lineages in the phylogeny of chameleons 34 are dense-forest dwellers (that is, not exposed to a dry and sunny environment), suggesting that the common ancestor of chameleons might have exhibited a similar ecology (but see alternative evolutionary scenarios in Supplementary Discussion ).
The organization of iridophoresterm·key into two superposed layers constitutes an evolutionary novelty for chameleons that allows some species to combine efficient camouflage with spectacular display. Additional analyses are warranted to identify whether the deep layer of iridophoresterm·key in chameleons further provide them with improved resistance to variable sunlight exposure.
Methods Animals Maintenance of and experimentsterm·mth on animals were approved by the Geneva Canton ethical regulation authority (authorization 1008/3421/1R) and performed according to the Swiss law.
Skin structure and Raman spectroscopy We examined the skin (ultra)structure by histology and TEMterm·mth, using standard procedures, for example, as described in ref. 14 . Samples were taken with biopsy pinches (diameter 2 mm) from male skin patches: when comparing skin before and after excitation, biopsies were separated by a maximum distance of 1 cm. For the relaxed state, the biopsy was taken within a few seconds after taking the animal from its cage and immediately placed in fixative. For the excited stateterm·dom, the animal was engaged in a male–male combat and a biopsy was taken again. The colour of skin biopsies was checked after fixation, to ensure that only those samples with well-preserved colours were used for TEMterm·mth.
Semi-thin (2 μm) and ultra-thin (80–90 nm) cross-sections were cut with a diamond knife on a Leica UCT microtome. Ultra-thin sections were viewed with a Tecnai G2 Sphera (FEI) TEMterm·mth at 120 kV before and after staining with uranyl acetate and lead citrate. Raman spectroscopy 16 was performed on melanophores directly on 2 μm cross-sections of skin samples with a home-made micro-Raman system composed of a 50-cm focal length spectrometer coupled to a nitrogen-cooled Princeton charge-coupled device detector and an argon laser (wavelength 514.5 nm) as the excitation source.
Nanocrystal measurements For S -iridophores, crystal height, length and spacing between nearest crystals were measured on unstained and stained sections (original magnification × 19,000) of the same skin samples. Distances between nearest crystals were similar for unstained (179.9±30.2, N =85) and stained (180.3±26.9, N =94) sections. A first set of experimentsterm·mth indicated that the average diameter of the more or less spherical ‘holes’ remaining after staining (124.2 nm, N =103) reasonably approximates the average size of intact crystals (length=149.7±15.3 nm, height=94.7±11.5 nm, N =145). Hence, all Supplementary Data were collected on multiple TEMterm·mth images of stained sections (performed on samples obtained from skin of various colours; Supplementary Table 1 ).
For D -iridophores, white ‘rectangular holes’ ( Fig. 1e ; corresponding to guanineterm·key crystals dissolved during post staining) on × 800 magnification TEMterm·mth images were fitted (in JMicroVision 36 ) with ellipses and geometric parameters (length, height and orientation) were computed subsequently. We performed Fourier transform analyses (as described in ref. 28 ), for each skin sample, on large assemblies of 20–30 high-resolution TEMterm·mth images (1 pixel=15 nm, typical size of a guanineterm·key crystal=200 nm) spanning over 100 × 100 μm, that is, about 50 times the length of the longest wavelength investigated inside the material (corresponding to 2.5 μm in vacuum). Each assembly included more than 100,000 crystals.
Photometry High-quality photographs and movies were obtained with a high-resolution digital single-lens reflex camera (Nikon D800) and a Panasonic HDC-HS700 video camera, respectively. To analyse each video frame ( Supplementary Fig. 2a ), RGBterm·mth band-pass filters were applied ( Supplementary Fig. 2b ) to select a colour window through which the variations of RGBterm·mth channels were monitored ( Supplementary Fig. 2c ). The corresponding RGBterm·mth numbers of each frame were averaged across the picture ( Supplementary Fig. 2d ) and normalized over the sum (R+G+B) ( Supplementary Fig. 2e ), to remove fluctuations of illumination as well as potential global variation in skin reflectivityterm·dom caused by movements of melanosomes. Next, each channel was normalized from 0 to 1 ( Supplementary Fig. 2f ), to exclusively measure the variation occurring (in S -iridophores) over a relatively constant background colour (generated by D -iridophores and/or pigments). After transformation of RGBterm·mth numbers from video into CIE XYZ tristimulus values, we derived the final x and y values that define colour irrespective of its luminance. Experimental traces (dots) are plotted and compared with the model (dashed line) on the CIE chromaticity chart ( Figs 1b and 2c ).
Skin reflectivityterm·dom in the ultraviolet range, highly relevant for colour perception in reptiles 37 38 , is not recorded by RGBterm·mth photometry. This does not have an impact on our conclusions, as the excellent matching between the modelled photonic response and photometry analyses validates our conclusions. In addition, photometric measurements have been validated with accurate spectroscopic measurements ex vivo ( Fig. 2b ) and show that XYZ /RGBterm·mth photometric videography is sufficient to detect the wavelength shift in the reflectivityterm·dom spectrum. In-vivo measurements of skin reflectivityterm·dom (including in the ultraviolet range) with spectrometers directly on the animal skin are difficult, mainly because the animals move and because chameleon skin darkens very rapidly when it comes in contact with the optical probe.
Optical modelling The symmetry of the close-packed photonic crystalsterm·key present in the top layer ( S -iridophores) of the skin was deduced from direct observations, under TEMterm·mth, of crystals sectioned in different planes ( Fig. 1d ). The structural element common to all close-packed structures is a triangular two-dimensional arrangement. Studies of the effect of symmetry variations (pp. 307–308 in ref. 39 ) have indicated that, for a/ λ <1 the optical response is mostly sensitive to the first Fourier component of the dielectric modulation. All close-packed structures have the same first-order Fourier component. To model the photonic structure effect that corresponds to our samples, it is therefore sufficient to choose face-centred cubic crystal symmetry. Crystal diameter ( d ) and distance between centres of nearest crystals ( s ) were measured (on about 1,200 individual crystals, Supplementary Table 1 ) on TEMterm·mth images as described above and the lattice parameterterm·mth a was computed as s √2. The first irreducible Brillouin zone (red lined contour in Supplementary Fig. 3b ) was meshed and the band energies were computed for each vertex centre using block-iterative frequency-domain methods as implemented in the Massachusetts Institute of Technology photonic band package 26 . For each direction of light propagation in the structure (that is, each mesh triangle centre at the surface of the Brillouin zone), the reflectivityterm·dom was set to unity in gapped frequency regions. As no preferential orientation of photonic crystalsterm·key relative to the skin surface was observed in S -iridophores, the weighted average among all directions was computed using the dimensionless a / λ parameter, where λ refers to the corresponding light wavelength in air and a represents the lattice parameterterm·mth. The refractive indices of guanineterm·key and cytoplasm were set to 1.83 and 1.33, respectively 13 . Convolution of local and average reflectivityterm·dom with standard spectral functions returns X , Y and Z colour numbers. Colour of each vertex and the average colour are plotted inside and outside of the irreducible Brillouin zone, respectively. Supplementary Movie 5 shows the evolution of local and average colours as the lattice parameterterm·mth varies.
Reflectivityterm·dom measurements Reflectivityterm·dom of skin samples was measured using a monochromatic source and the locked-in detection of a spectroscopic Woollam ellipsometer from 300 to 2,500 nm and with a resolution of 5 nm. Monochromatic light of the source was injected in one channel of a reflection fibre probe (QR400-7-VIS-NIR). The light reflected on the sample (near-normal incidence) was collected by the second optical fibre channel and guided to a silicon photodiode (for visible range) and constrained InGaAs detectors (for the near-infraredterm·key range). A white diffuse standard (Ocean Optics WS-1-SL) was used as reference. This set up allowed measurements in air or in a solution with adjustable osmolarity (from Ringer’s 1 × to 6 × , that is, from 236 to 1,416 mOsm).
Single-cell videography Photometric videography on individual S -iridophores, were measured with adjustable osmolarity. Osmotic pressureterm·key was applied to fresh skin samples taken with biopsy pinches (diameter 2 mm) from white stripes of male panther chameleonsterm·key. Samples were placed in a Ludin chamber where Ringer’s 1 × solution (236 mOsm) was slowly replaced with Ringer’s 4 × solution (944 mOsm). Full high-definition videos (1,920 × 1,080 pixels) were recorded with a Nikon D800 camera attached to a Leica MZ16 stereoscope. XYZ /RGBterm·mth evolution of individual S -iridophores were obtained following the normalization procedure described in the ‘Photometry’ section.
Author contributions M.C.M. conceived the study. M.C.M. and D.v.d.M. supervised the study. J.T. and S.V.S. performed TEMterm·mth analyses and osmotic pressureterm·key experimentsterm·mth. S.V.S. performed histology. J.T. performed Raman spectroscopy and optical modelling. J.T., M.C.M. and S.V.S. performed RGBterm·mth photometry. M.C.M., S.V.S., J.T. and D.v.d.M. wrote the manuscript. All authors approved the final version of the manuscript.
Additional information How to cite this article: TeyssierpersName, J. et al . Photonic crystalsterm·key cause active colour changeterm·key in chameleons. Nat. Commun. 6:6368 doi: 10.1038/ncomms7368 (2015).
Supplementary Material Supplementary Figures 1-4, Supplementary Table 1, Supplementary Discussion and Supplementary References In vivo colour changeterm·key in a F. pardalis adult male under excitation upon presentation of another adult male in its vision field. The original video is stabilised and accelerated 8 times. The first frame of the movie is shown in the lower-right for a better visualisation of the extend of colour changeterm·key.
Colour changeterm·key is fully reversible: in vivo colour changeterm·key in a F. pardalis adult male under relaxation after a male-male combat. Males are much less mobile after than during a combat such that the movie is much more stable. The original video is accelerated 8 times. The first frame of the movie is shown in the lower-left for a better visualisation of the extend of colour changeterm·key.
In vivo colour changeterm·key in a F. pardalis adult male under excitation upon presentation of another adult male in its vision field. The original video is stabilised and accelerated 3 times. The first frame of the movie is shown in the upper-right for a better visualisation of the extend of colour changeterm·key. Note that this male is dominated by the male in its vision field, hence, colour changeterm·key is mild.
Ex vivo colour changeterm·key of an adult male F. pardalis white skin sample induced by increasing osmolarity of the medium from 236 mOsm to 944 mOsm. The original video is accelerated 30×. The inset shows a 10× magnification of a single cell. This experimentterm·mth indicates that individual Siridophores experience a gradual shift in colour across the whole visible spectrum.
Simulation of RGBterm·mth colour shift at the surface of the first Brillouin zone of the FCC lattice of Siridophores when gradually reducing the crystal lattice parameterterm·mth (a) from 480 to 233 nm. The simulated colours closely match those observed in vivo (Figure 1b in main text, Supplementary Movies 1 , 2 , 3 ) and during osmotic shock experimentsterm·mth (Figure 2c in main text, Supplementary Movie 4 ).