Hagfish embryos!

A hagfish egg with a 14.3-mm pharyngula-stage embryo inside (arrows). Scale bar, 5 mm.

I’ve been looking forward to seeing these little jewels in print since I saw Kuratani talk about them at the SICB meetings in January. Hagfish are wonderfully slimy jawless chordates that have been difficult to raise in the lab—although if you poke a whale corpse rotting in the cold deeps you’ll find them swarming everywhere. The Kuratani lab has managed to keep animals of the species Eptatretus burgeri alive and healthy in a lab aquarium maintained at cold temperatures (16°C), and has even had success in breeding them. That object to the right is a single hagfish egg, brown and leathery-shelled and surprisingly big—it’s an inch and a half long!

They collected 92 eggs, and then another limitation emerged: it took 5-7 months for embryos to develop in a small number of the eggs. Hagfish aren’t going to be your typical fast-developing model system, I’m afraid, but they are extraordinarily cool animals, and it’s good to see work beginning on them.

So what do the embryos look like? To no one’s surprise, much like your typical chordate embryo—note the pharyngula stage embryo in d and f, with the divisions of the brain visible, and distinct eyes and ears. It’s beautiful.

c, A 7.4-mm embryo, corresponding to the late neurula stage. d, A 14.3-mm embryo, corresponding to the early pharyngula stage. e, The head of the 7.4-mm embryo. f, The head of the 14.3-mm embryo. fb, forebrain; hb, hindbrain; mb, midbrain; ot, otic pit; ph, pharyngeal wall; som, somites; TC, trigeminal crest cells. Scale bars, 5 mm (c?d); 1 mm (e, f).

One reason to be interested in hagfish embryos is to resolve questions about the origin of specific chordate evolutionary novelties. They are more primitive than lampreys in many ways, and in particular, we’d like to know more about a specific embryonic tissue, the neural crest, in these animals. The neural crest was a key innovation that was crucial in providing the raw material for much of the vertebrate head.

Previous studies of hagfish embryos (from 1899! It’s been a while) relied on small numbers of embryos that were typically poorly fixed — note the large yolky eggs, which are always problematic. Those studies suggested some unusual mechanisms of neural crest generation in the hagfish.

a, The hagfish may be either the sister group of all the other vertebrates (1), or that of lampreys (2). b, c, Hypotheses on neural crest evolution. b, The former hypothesis (1) would agree with the scenario that neural crest evolution had an intermediate epithelial state. c, Alternatively, the crest might have already been established as a population of delaminating cells in the common ancestor to all vertebrates, and this would be coherent with both hypotheses (1 or 2).

The neural crest arises from a population of cells that are produced at the boundary of the neural tube and epidermis during neurulation. In familiar vertebrates, the neural crest cells delaminate or peel away from the top of the nervous system and migrate as a loose mass called mesenchyme to new destinations in the body. The early descriptions of the hagfish suggested something different: instead of delaminating, the sheet of cells that gives rise to the epidermis and neural tube retains coherence instead of delaminating, and folds into little pouches that run the length of the body (the dark loops in b, above). If true, that would be curious and different, and would suggest that perhaps neural crest first originated as an epithelium, rather than a mesenchymal mass.

Kuratami’s work suggests that the earlier interpretation was wrong, and was an artifact produced by poor fixation. They’ve now examined embryos that are well fixed, and have also been able to stain the tissue with modern probes specific for neural crest and other neural markers. The answer is that hagfish neural crest emerges like that of other chordates, as a loose collection of cells that migrates preferentially in the spaces between the somites.

Putative neural crest cells in a 7.4 mm hagfish embryo. b, Illustration of neural crest cell distribution in the embryo on the basis of three-dimensional reconstruction from the histological sections; seen from the anterior oblique view. Red arrows indicate the ventrally migrating crest cells occurring at the same intervals as the apices of somites shown by purple circles. c, Schematic representation of a horizontal section of the embryo to show the planes of sections d and e. d, A transverse section at the mid-somite level. No putative crest cells are found between the somite (som) and the neural tube (nt). Asterisks in c and d indicate the apex of each somite. e, A transverse section at the intersomitic level. Putative crest cells (arrowheads) are filling in a space between the somite and the neural tube. n, notochord; nt, neural tube; som, somite. Scale bars, 100 µm.

These are useful results — they push back the origin of important embryological features to the common ancestor of hagfish, lampreys, and gnathostomes, and suggests that they may have appeared over 500 million years ago. These animals are going to be important intermediates to study, although the difficulty and slow rate of development is rather daunting. Maybe we need to get some fanatical aquarists working on more efficient aquaculture methods — wouldn’t you love to have a tank full of hagfish to play with in your home?

Ota KG, Kuraku S, Kuratani S (2007) Hagfish embryology with reference to the evolution of the neural crest. Nature 446:672-675.


  1. #1 SN
    February 22, 2008

    Does someone know, where i can purchase live hagfish?

  2. #2 Lily Kirstin Frank
    June 26, 2008

    I am a live hagfish. Seriously.

  3. #3 Nick Gotts
    June 26, 2008

    Is it true they are the only animals that are known to be able to tie themselves in a knot? Or could some of the cephalopods manage this?

New comments have been temporarily disabled. Please check back soon.