Reconstructing the evolutionary history of the squishy soft parts of animals from fossil evidence is often difficult or impossible, and the vertebrate eye is a prime example of this problem. This is a pity, because the evolution of eyes, which have arisen independently in diverse animal groups, is highly contentious. In such circumstances, evolutionary biologists are forced to rely on evidence that can be gleaned from living animals: comparative anatomy, developmental biology and molecular genetics. When fossils do reveal the anatomical detail of soft parts, it is often down to the crafty use of novel techniques.
The animals most closely related to vertebrates today are invertebrates with a notochord (a stiffened rod along the head-to-tail axis) which are therefore included within our own phylum, Chordata. These include cephalochordates such as amphioxus (also known as the lancelet), and ascidians (sea squirts and their relatives). These animals have simple, light-sensitive eye-spots but nothing that we would describe as an eye.
The majority of vertebrates have the sophisticated type of eyes which we are familiar with: they have a lens, iris and eye muscles, and work like a camera to produce an image on the retina. Sitting between the invertebrate chordates, with their simple eye-spots, and most vertebrates, with complex camera-like eyes, are the few living representatives of the jawless vertebrates: the lampreys and hagfish (collectively known as cyclostomes). Modern lampreys have a complex eye, but modern hagfish do not, and, crucially, do not have pigment granules in the epithelium of their retina either. So, do the ‘eyes’ of modern hagfish represent a halfway house between eye-spots and full-blown vertebrate eyes?
A new paper by Sarah Gabbott at the University of Leicester and colleagues, describes eye pigments in both a lamprey and a hagfish from the late Carboniferous, around 300 million years old. The cyclostome specimens are preserved as compressions showing soft-part preservation, within ironstone concretions, from the exceptionally well-preserved Mazon Creek fossil deposit in Illinois.
The same deposits yielded the Tully Monster, Tullimonstrum gregarium, which was reappraised early this year and moved from its position as one of the celebrated weird kids of palaeontology to a more respectable identity as a hagfish ancestor. Studies of the eyes of Tullimonstrum have already shown that it had a retina with distinct layers of spheroidal and cylindrical melanosomes. Melanosomes are tiny organelles that make and store melanin. Lots of animals have melanin in their eyes, but this particular arrangement of the two shapes of melanosome, forming retinal pigment epithelium, is characteristic of the vertebrates. Tullimonstrum, however, is a far from typical cyclostome, not least in the bizarre handlebar arrangement of its eyes, mounted on the ends of a rigid traverse bar.
Gabbott and colleagues’ paper shows that the Mazon Creek lamprey Mayomyzon pieckoensis and hagfish Myxinikela siroka both possess structures that comparative anatomy suggests are eyes (Mayomyzon also shows evidence of a lens), and which are preserved as carbon-rich films composed of tiny cylindrical and oblate ‘microbodies’.
To confirm what the microbodies were composed of, they used a technique with a catchy little name: time of flight secondary ion mass spectrometry. This technique, which collects mass spectrometric data from the surface of specimens, has previously been applied to a variety of animal fossils to identify the presence of fossil melanosomes, including bird feathers, amphibians, mammals and squid ink. The secondary ion spectra acquired from the eye films of the Mazon Creek cyclostomes contain all of the organic secondary ion fragments characteristic of fossil melanin (and, importantly, ruled out the interpretation of the microbodies as bacteria).
Molecular genetic studies suggest that hagfish and lampreys share a more recent common ancestor with each other than either group do with the rest of the vertebrates: the cyclostomes are a monophyletic group. So, the presence of pigmented eyes in the Mazon Creek hagfish Myxinikela indicates that the common ancestor of all of the cyclostomes also had pigmented eyes. The eye of the modern hagfish has degenerated from an earlier, more sophisticated eye anatomy, and does not tell us anything about the evolutionary steps towards the vertebrate eye. This fascinating piece of work also suggests that the last common ancestor of the vertebrates had an eye that was at least as complex as that of lampreys.
