1 2 3 4 5 6 Colour in bivalve shells: using Resonance Raman spectroscopy to compare 7 pigments at different phylogenetic levels 8 9 J. Wade1, H. Pugh2,3, J. Nightingale1, J.S. Kim1, S. T. Williams2* 10 11 1 Imperial College London & Centre for Plastic Electronics, Blackett Laboratory, Prince 12 Consort Road, London SW7 2AZ, United Kingdom; 13 [email protected]; [email protected]; [email protected]; 14 2 Natural History Museum, Department of Life Sciences, Cromwell Rd, London SW7 5BD, 15 United Kingdom; 16 [email protected]; [email protected] 17 3 Imperial College London, Department of Life Sciences, London SW7 2AZ, United Kingdom. 18 19 20 21 22 23 24 * Corresponding author. Email: [email protected] 25 Short title: Raman spectroscopy of bivalve shells 26 Key words: colour, bivalve, Raman spectroscopy, pigment, shell 27 1 28 Table of Contents Studies have suggested that shell colour is phylogenetically distributed within Mollusca, but this is confounded by a lack of knowledge about the identity of most molluscan pigments. We use Raman J. Wade, H. Pugh, J. spectroscopy to examine Nightingale, J.S. Kim, S. T. * bivalve pigments and Williams compare spectra from taxa at different phylogenetic levels. We show that most Colour in bivalve shells: colours in bivalves are due using Resonance Raman to partially methylated spectroscopy to compare polyenes, possibly modified pigments at different carotenoids, but we were phylogenetic levels unable to detect Raman activity for green pigments suggesting this colour is due to structural elements or a different class of pigment. 29 2 30 Abstract 31 Several studies have suggested that shell colour may be phylogenetically distributed within 32 the phylum Mollusca, but this is confounded by our ignorance of its homology and lack of 33 understanding about the identity of most molluscan pigments. We use Resonance Raman 34 spectroscopy to address this problem by examining bivalve pigments producing a range of 35 colours and compare spectra from taxa at different phylogenetic levels. The spectra of most 36 shell pigments tested exhibited a skeletal signature typical of partially-methylated polyenes, 37 possibly modified carotenoids, with the strongest peaks occurring between 1501-1540 cm-1 -1 38 and 1117-1144 cm due to the C=C (1) and C-C (2) stretching modes respectively. Neither 39 pigment class nor mineral structure differentiated Imparidentia and Pteriomorphia, but the 40 most similar pigment spectra were found within a genus. Spectral acquisitions for purple 41 pigments for two species of Asaphis suggest that identical, or nearly identical pigments are 42 shared within this genus, and some red pigments from distantly related species have similar 43 spectra. Conversely, two species with brown shells have distinctly different pigments, 44 highlighting the difficulty in determining the homology of colour. Curiously we were unable 45 to detect any Raman activity for green pigments or pigment peaks for the yellow area of 46 Codakia paytenorum suggesting that these colours are due to structural elements or a 47 pigment that is quite different from those observed in other taxa examined to date. Our 48 results are consistent with the idea that classes of pigments are evolutionarily ancient but 49 heritable modifications may be specific to clades. 50 51 3 52 Introduction 53 Colour and pattern are often associated in the natural world with camouflage, warning, and 54 sexual selection1. Understanding how new biological colours have arisen is a fundamental 55 challenge to our understanding of evolutionary ecology and developmental biology, but 56 despite over a century of interest, the evolution of colour in molluscan shells is only just 57 beginning to be explored in detail2. Of particular interest is the observation that colour and 58 the distribution of fluorescence associated with some pigmentation both seem to be 59 distributed in a phylogenetically significant manner3-6. Grant & Williams6 mapped the 60 distribution of shell and periostracal colour onto a bivalve phylogeny and showed that the 61 phylogenetic distribution of colour is statistically significant, as are the distributions of 62 individual shell and periostracal colours. This finding, however, is complicated by the fact 63 that different pigments can produce the same colours, and as such the homology of colour is 64 unknown. In order to better understand how and why colour has evolved in shelled molluscs 65 it is essential to identify shell pigments and determine their similarity across different clades. 66 There have been a growing number of studies in the last decade investigating the pigments 67 responsible for molluscan shell colour (e.g.7-11) and the genes and pathways responsible for 68 their production (e.g.10,12-14). Techniques used to distinguish pigments in shells have included 69 high performance liquid chromatography (HPLC), mass spectrometry (MS) and UV-visible 70 spectrophotometry, but difficulties in some studies have arisen from the low concentrations 71 of pigments7,15,16. Raman spectroscopy overcomes this problem and is particularly effective 72 for detection of polyene pigments, as well as being non-destructive and non-invasive17-19. 73 The term polyene is a general chemical term for poly-unsaturated organic compounds 74 including carotenoids and tetrapyrroles, with one or more arrangements of alternating 75 double and single carbon–carbon bonds18. Such complex molecules tend to be highly 76 coloured, with strong absorption bands in the visible region of the spectrum18. Polyenes are 77 exceptionally strong Raman scatterers due to electron-phonon coupling20. Electron-phonon 78 coupling describes the interconnection of -electron density and nuclear position, which 79 means that changes in Raman spectra can provide detailed information on chemical 80 structure, polymer conformation, and also effective conjugation length. Additionally, they 81 have a large resonance Raman cross-section, which means that the Raman signature of a 82 pigment can be obtained with high selectivity from the shell itself. 83 Raman spectroscopy has previously been used in a broad range of molluscan shell pigment 84 studies (e.g.7,17,21-28.29,30) and to detect pigments in freshwater pearls (e.g.25,31,32). These 85 studies identified most shell pigments as polyenes, in some cases suggesting that molluscan 86 shell pigments are unsubstituted polyacetylenes, possibly of the carotenoid class of 87 compounds 22. However, Raman can be used to distinguish among all three of the main 88 classes of pigments known to occur in mollusc shells (melanins, tetrapyrroles and 89 carotenoids). For instance, Raman has been used to identify melanins in hair and 90 feathers33,34 and cuttlefish ink35; tetrapyrroles have been identified in molluscan shells, 91 brachiopod shells36, bird eggs37 and dinosaur eggs38; and carotenoids, carotenoproteins or 4 92 structurally modified carotenoids have been identified in molluscs, brachiopods and 93 corals22,31,36. 94 In this study we use Resonance Raman spectroscopy to compare spectra of shell pigments 95 across a range of bivalve taxa, with a range of different colours in order to explore the 96 pigments found in this group. We compare multiple spectra from fifteen specimens of 97 bivalves representing the taxonomic groups Pteriomorphia and Imparidentia at different 98 phylogenetic levels to determine whether pigments among clades are more similar than 99 those in different clades. 5 100 Experimental 101 Samples 102 Fifteen bivalve shells were selected from the Natural History Museum collections, from 103 major clades corresponding to the subclass Pteriomorphia (Alectryonella plicatula, Fig. 1M; 104 Arca zebra, Fig. 1C; Mimachlamys crassicostata (two colour morphs, Figs 1D, 1K;), Mi. 105 gloriosa, Fig. 1G; Mi. varia, Fig. 1A; Monia zelandica, Fig. 1O) and the superorder 106 Imparidentia (Acanthocardia tuberculata, Fig. 1F; Asaphis deflorata, Fig. 1I; As. violascens, 107 Fig. 1H; Chama brassica, Fig. 1N; Codakia paytenorum, Fig. 1L; Johnsonella fausta, Fig. 1J; 108 Pharaonella sieboldii, Fig. 1E; Phylloda foliacea, Fig. 1B). Shells of selected specimens were 109 relatively uniformly coloured, exhibiting either one or two vivid colours. Specimens were 110 cleaned using a sonicator with detergent and water for two minutes prior to Raman 111 measurements of the shell surface. 112 Where more than one colour occurred within a single shell these were examined separately 113 to investigate intra-individual variation: red and yellow shell colour in the lucinid Codakia 114 paytenorum (Fig. 1L) and purple and orange shell colour in Asaphis deflorata (Fig. 1I). 115 Intraspecific variation was investigated in two differently coloured individuals (red and 116 purple; Figs 1D, 1K) of scallops Mimachlamys crassicosta and interspecific variation within 117 two genera: Asaphis by comparing a purple shell from A. violascens (Fig. 1H) with a purple 118 and orange shell from A. deflorata and within Mimachlamys by comparing the shell spectra 119 of red and purple Mi. crassicosta with spectra from an orange Mi. varia (Fig. 1A) and a 120 yellow Mi. gloriosa (Fig. 1G). 121 Finally, species from different clades (Pteriomorphia and Imparidentia) with similar shell 122 colours were compared, focussing on five colours: red, orange, yellow, purple, brown and 123 green. We compared a red Mi. crassicostata shell (Fig. 1D) with the red margin on the 124 internal valve of C. paytenorum (Fig. 1L), yellow shells from Mi. gloriosa (Fig. 1G), C. 125 paytenorum and Johnsonella fausta (Fig. 1J), orange in Phylloda foliacea (Fig. 1B), Mi. varia 126 (Fig. 1A) and Asaphis deflorata (Fig. 1I), purple in As. violascens (Fig. 1H), As. deflorata and 127 Mi. crassicostata (Fig. 1K), brown in Arca zebra (Fig. 1C) and Acanthocardia tuberculata (Fig. 128 1F) and three green species Alectryonella plicatula (Fig. 1M), Monia zelandica (Fig. 1O) and 129 Chama brassica (Fig. 1N). 130 131 Raman measurement 132 Bivalve spectral acquisitions were recorded using a Renishaw inVia Raman Microscope 133 calibrated using the Silicon Raman band at 520.5 cm-1 with a spot size of 1 μm2.