Hagfish embryos show developmental features that contradict the idea that these jawless fish are the most primitive living vertebrates. The findings also help to trace the evolution of vertebrate cranial structure. See Article p.175
Hagfish are bizarre creatures. These eel-shaped marine fish have a skull but no vertebrae, and teeth but no jaw. These and other odd features have long fascinated evolutionary biologists, who regard the 'primitive' characteristics of hagfish as a treasure trove of clues to understanding the early evolution of vertebrate anatomy and physiology. But hagfish embryos are reputedly scarce, so their development has remained an enigma. Six years ago, however, a team of Japanese researchers realized1 that hagfish embryos are not particularly rare: they simply take more time to mature than other piscine vertebrates. This discovery triggered extensive investigations of hagfish embryos, some of which are published by Oisi et al.2 on page 175 of this issueFootnote 1. The authors provide a spectacular overview of hagfish craniofacial development, with particular reference to the origins of the olfactory organs and pituitary gland, two key features in the shaping of the vertebrate skull3.
Hagfish often feed on sunken carcasses, tearing off food using a peculiar tongue-like feeding device that is armed with sharp teeth made of keratin. They were first classified as cyclostomes, alongside lampreys, because both groups of jawless fish have similar feeding devices and pouch-shaped gills, and both lack a calcified skeleton. Cyclostomes were regarded as an early offshoot of the vertebrate evolutionary tree. Then, in the 1970s, hagfish were considered to be anatomically and physiologically more primitive than either lampreys or jawed vertebrates (gnathostomes), and to have diverged earlier than all other vertebrates. However, this hypothesis has since been dismissed on the basis of DNA sequence data that show a close phylogenetic relationship between lampreys and hagfish4,5. But until recently, lamprey development had been studied much more than that of hagfish.
Hagfish and lampreys have a single median (located in the midline of the body) 'nostril', which opens anteriorly in hagfish but dorsally in lampreys. This nostril is the entrance of a tube into which two organs open: a median olfactory organ and the adenohypophysis (also known as the anterior pituitary gland, which is part of the endocrine system). This monorhinal (single nostril) condition differs from the diplorhinal (two nostrils) structure of the gnathostomes, in which two well-separated olfactory organs open to the exterior environment through independent nostrils, as do human noses. The separation of the olfactory organs in gnathostomes is now regarded as a prerequisite to the development of jaws3, because it allowed the mandibular arch to extend forward.
However, this difference in the structures poses a conundrum. The cyclostomes and gnathostomes are each monophyletic4,5, meaning that they form a group comprising an ancestral species and all its descendants. Non-vertebrate members of the phylum Chordata (to which vertebrates also belong), such as amphioxus or tunicates (sea squirts), lack an indisputable homologue of the olfactory organs and adenohypophysis. This raises the question: what was the arrangement of these organs in the most recent common ancestor to all vertebrates? To answer this, Oisi et al. studied the embryonic tissues that give rise to the olfactory organs and adenohypophysis in hagfish, lampreys and gnathostomes. These tissues, called placodes, are thickenings of the embryonic ectoderm (the outer germ layer) that develop into various sensory organs, such as the olfactory organs, eye lens, otic capsules and the adenohypophysis.
In the lamprey embryo, the olfactory and adenohypophyseal placodes are united into a single nasohypophyseal plate (NHP) that forms beneath the upper lip and gives rise to a nasohypophyseal complex, which migrates dorsally later in development (Fig. 1). Oisi and colleagues found that hagfish embryos also have an NHP, and that the gene-expression pattern in both the lamprey and hagfish NHP is characteristic of ectoderm-associated tissues. These findings clearly refute previous speculation that the hagfish NHP derives from the endoderm (the inner germ layer that later forms the pharynx and digestive tract), which, if true, would have been the only case of this among vertebrates.
But this reconciliation of the structure and position of the NHP in hagfish and lampreys still leaves two distinct conditions among vertebrates: cyclostomes have a median NHP, and gnathostomes have independent nasal and adenohypophyseal placodes (Fig. 1). Could one of these conditions be general — or 'ancestral' — for vertebrates, and the other have subsequently derived from it? Oisi et al. suggest that the cyclostome condition (a median NHP) is likely to be the earlier structure, because the monorhinal condition is found in the extinct jawless vertebrates commonly known as ostracoderms, which are between 480 million and 370 million years old and are regarded as more closely related to gnathostomes than to cyclostomes (Fig. 1).
Placodes are embryonic structures that are not preserved in ostracoderm fossils, but the skeletons of these animals provide information about their differing cranial anatomy6,7. Ostracoderms display diverse arrangements of the cranial spaces that housed the olfactory organs and the adenohypophysis, but in all of them these spaces open towards a median nostril, as in cyclostomes. For example, anaspids and osteostracans have a median dorsal nasohypophyseal complex similar to that of lampreys, which suggests some kind of cyclostome-like NHP, although other features show that osteostracans are closer to gnathostomes than to any other ostracoderm. By contrast, heterostracans and galeaspids had paired olfactory organs, still closely associated with the adenohypophysis in the latter, but opening anteriorly by a large median nostril and suggesting a broader NHP (Fig. 1).
This anatomical diversity is puzzling, but Oisi and colleagues' findings on hagfish development help to clarify the picture. The authors propose that the presence of the NHP in both hagfish and lamprey embryos supports the idea of a general, 'pan-cyclostome' pattern for craniofacial development, and that the anatomical characteristics that make present-day hagfish seem more primitive than lampreys are the result of a subsequent loss of features in hagfish during evolution. Such a pattern, combined with the fact that no NHP is seen during the development of any living gnathostomes, and the apparent progression from the monorhinal to diplorhinal condition in ostracoderms, suggests that a crucial innovation in the development of the vertebrate snout was the loss of a common nasohypophyseal duct and the rise of separate, paired nostrils.
*This article and the paper under discussion2 were published online on 19 December 2012.
Ota, K. G. & Kuratani, S. Zool. Sci. 23, 403–418 (2006).
Oisi, Y., Ota, K. G., Kuraku, S., Fujimoto, S. & Kuratani, S. Nature 493, 175–180 (2013).
Kuratani, S. J. Anat. 205, 335–347 (2004).
Kuraku, S. Zool. Sci. 25, 960–968 (2008).
Heimberg, A. M., Sempere, L. F., Moy, V. N., Donoghue, P. C. & Peterson, K. J. Proc. Natl Acad. Sci. USA 105, 2946–2950 (2008).
Janvier, P. Zool. Sci. 25, 1045–1056 (2008).
Gai, Z., Donoghue, P. C. J., Zhu, M., Janvier, P. & Stampanoni, M. Nature 476, 324–327 (2011).
About this article
Australian Dental Journal (2014)
Proceedings of the National Academy of Sciences (2014)
Using information in taxonomists’ heads to resolve hagfish and lamprey relationships and recapitulate craniate–vertebrate phylogenetic history
Historical Biology (2014)
Insight from the lamprey genome: Glimpsing early vertebrate development via neuroendocrine-associated genes and shared synteny of gonadotropin-releasing hormone (GnRH)
General and Comparative Endocrinology (2013)