Pigments through time

The aim of palaeontological research is to decipher the evolution, biology and function of life on Earth by studying the fossil remains of organisms preserved in the geological record. Most often the material that survives over tens of millions of years is typically hard, biomineralized tissue such as bone and shell. This is mostly due to the inherent robust physical and chemical properties of such hard biomaterials. ‘Soft tissues’ including hair, feathers, skin, internal organs, cuticles and plant matter rapidly degrade in most environments, but in exceptional and isolated circumstances, these less durable biomaterials can be preserved over millions of years. The morphological details revealed by such preservation have not only allowed palaeontologists to interpret and reconstruct how many ancient and now long-extinct organisms may have appeared, but have also provided insight to their biology, ecology, behaviour, physiology and possible interactions with their environment. However, one extremely important characteristic of ancient organisms has until recently been conspicuously elusive: colour. Colour is known to play a vital role in many facets of the behaviour of living organisms, including adaptations to specific environments such as camouflage, and also has direct impacts upon ecology, such as communication and sexual selection. Indications of colour in fossilized organisms may help to resolve these characters in fossil remains and how they may have influenced the evolution of a species. In some very rare cases, fossils may show a visible indication of pigment patterns such as spots or stripes, but positive identification of specific colour or pigment mechanisms has so far remained extremely difficult. This is in part due to the limitations, such as detection limits, of analytical techniques that may be used to search for pigments, and partly due to the entrenched belief that the chemistry and micromorphology of a living organism is altered or replaced by a number of fossilization processes to such a degree that all useful information has been lost. In recent years, a growing number of studies have investigated the potential preservation of pigments within fossil tissues, with particular emphasis falling upon melanin. The focus on melanin is a result of the fact that melanin is almost ubiquitous in life on Earth, having been identified in a wide variety of living organisms including most chordates (Figure 1), many groups of invertebrates including echinoderms, sponges, molluscs and arthropods, and is also observed in plants, fungi and bacteria. Its presence in both highly derived and more primitive species indicates that melanogenesis represents a very ancient biosynthetic pathway utilized by the earliest life on the planet. Indeed, the tyrosinase enzyme, a key copper-bearing participant in melanogenesis, may have evolved from an extremely primitive biochemical function going back over one billion years (Van Holde et al., 2001). Additionally, the focus on melanin is also a result of the fact that unlike some of the other forms of pigmentation, melanin is stored in organelles – melanosomes – that form diagnostic shapes indicative of the type of melanin pigment contained within. This structural component, albeit microscopic in scale, has the potential to be preserved over millions of years and may be detectable by standard imaging techniques such as scanning electron microscopy (SEM). Zhang et al. (2010) was one of the first of several studies by different groups to identify melanosomes preserved in fossil tissues. Using SEM imaging, they provided evidence of melanosome like structures, rod-shaped eumelanosomes (~2 microns in length) and spherical phaeomelanosomes (~1 micron in diameter) preserved in the feathers of derived theropod (predatory) dinosaurs and birds from the Cretaceous Period (~120 million years old, China). This and several subsequent studies proceeded to present whole-organism pigment reconstructions based on the presence of melanosomes in select sampled areas. Others have taken a different approach in an attempt to identify melanin pigment in fossils. By employing a suite of chemical techniques along with structural analysis, Wogelius et al. (2011) investigated the metal patterns known to be affiliated with melanogenesis in a range of exceptionally preserved fossil tissues. They performed elemental mapping and spectroscopy of copper in the feathers of a 120-million-year-old primitive bird and concluded that the residual organometallic copper complexes were derived from the original copper associated with melanogenesis. Additionally, melanosome-like structures were identified in areas of increased copper but not in areas lacking copper. In their recent report, Lindgren et al., have built upon these previous studies by again combining structural imaging with chemical analysis [in this case, time of flight-secondary ion mass spectrometry (TOF-SIMS)] of fossil tissues. Here, they examined fossils of extinct marine reptiles (a 55-million-year-old leatherback turtle, a 86-million-year-old mosasaur and a ~190-million-year-old ichthyosaur) and showed evidence for structural preservation of eumelanosomes (via SEM imaging) and molecular evidence for the presence of the eumelanin molecule or its breakdown products via TOF-SIMS. The TOF-SIMS fragmentation patterns of the preserved fossil melanosomes bore a distinct similarity to comparable extant melanized tissues as well as to eumelanin standards. Lindgren et al., go on to suggest that the presence of eumelanin in relatively distantly related marine organisms provides evolutionary evidence of convergent melanism. The main advantage that Lindgren et al., has over previous studies of fossil pigments is that melanosome structures and/or trace metals are in effect proxies for melanin pigment, but are not direct evidence for the preservation of the melanin molecule itself. Lindgren et al., show evidence for remnants of the eumelanin macromolecular structure being preserved in situ with structural preservation. However, there are many complicating factors with this research field which should be carefully considered and discussed. Most importantly, melanosomes bear a remarkable similarity to quite common and pervasive bacteria both in size and shape. Indeed, previous research identified such structures in fossils as microbial. The reinterpretation of these fossil structures as melanosomes is still under active debate with a study by Moyer et al. (2014), published soon after Lindgren et al., convincingly arguing a bacterial origin for structures which had been interpreted by others as melanosomes. Despite the melanosome/bacteria debate continuing (Moyer et al., 2014), newly published studies are starting to assign these structures as melanosomes by default (Li et al., 2014) without presenting corroborating data for the SEM imaging from each specimen studied. For these reasons, including chemical analysis as Lindgren et al. have done is tremendously important, as relying on structure alone is fraught with peril. Another important limitation of studies that utilize microstructural imaging is that only a small fraction of a specimen can be analysed. Every study published so far using SEM images, including Lindgren et al., has examined <1% of the remnant soft tissue and then extrapolated pigment patterns over large areas. By analogy, examination of <1% of a zebra could lead one to seriously misinterpret pigment distribution in that species. Additionally, it is likely that in many fossil samples, structural preservation will not be uniform, thus creating further uncertainty in reconstructions of pigment patterns derived from melanosomes. While structural information is important, given the technical limitations of the SEM (sample size, high vacuum, slow data acquisition rates), the interpretations that can be derived will therefore be equally limited. The technology exists to be able to perform elemental analysis of 100% of a fossil specimen (Wogelius et al., 2011), which can help avoid the problems of positively discriminating microbial structures from melanosomes by correlating structure with the remnant chemistry of an entire specimen. However, chemical approaches are also not without their obstacles. Differentiating between geologically and biologically sourced chemical signals is challenging, but so far the data show that original chemical retention may be more common than previously believed. Overall, the various studies and approaches strongly indicate that remnants of pigmentation, such as eumelanosomes, phaeomelanosomes and the chemical residue of eumelanin, can indeed be preserved for over a 100 million years in exceptionally preserved fossils. But additionally, the current body of work suggests that pigment studies must make use of a suite of techniques to draw supportable conclusions and that colour reconstructions of entire organisms based on extrapolation from limited sampling may not be reliable. “colour reconstructions of entire organisms based on extrapolation from limited sampling may not be reliable.”