News & Views | Published:


Along came a sea spider

An investigation of brain development in sea spiders provides hints about how the earliest arthropod head evolved. These observations are bound to provoke controversy in an already acrimonious field.

Obscure groups of animals have been making scientific waves lately1, and few are more obscure than the sea spiders, or pycnogonids. These marine, spider-like animals differ from other arthropods, such as the true spiders, crustaceans and insects, in many ways. Their bodies are so slender that the digestive systems and gonads are squeezed into their limbs; they possess a forward-pointing proboscis with a terminal mouth; and the males brood the eggs. Flanking their unique proboscis is a pair of pincer-bearing appendages known as chelifores, which it has long been assumed are related to the pincered fangs — chelicerae — of spiders.

Work presented by Maxmen et al. on page 1144 of this issue2, however, suggests that pycnogonid chelifores and spider chelicerae develop from different regions of the head and therefore cannot be equivalent. At first sight this is a rather esoteric finding. But if it is correct, it will shake up the field of arthropod evolution.

Maxmen et al. relied on the fact that each of an arthropod's pairs of appendages is derived from one of the repeating segments that make up the arthropod body. In addition to its appendages, each segment has a pair of nerve concentrations, or neuromeres. The authors reason that tying the appendages to a specific pair of neuromeres should reveal which segment the appendages belong to. As chelifores and chelicerae are head appendages innervated from the brain, Maxmen et al. considered which of the brain neuromeres each appendage is associated with during larval development. Arthropod brains are divided into three regions: protocerebral, deutocerebral and tritocerebral, from front to back. The anteriormost appendage of most living arthropods, including the spider chelicera, is innervated from the deutocerebrum3. What Maxmen et al. have now shown, in surprising contrast, is that the pycnogonid chelifores seem to be innervated from the protocerebrum — the most anterior part of the brain. The association of chelifores and chelicerae with different parts of the brain implies that the two types of limb are not equivalent, but are derived from different segments.

This result cuts across previous results based on adult structure4, and to see the wider implications we need some historical background. The composition of the arthropod head is one of the bitterest and longest-running problems in animal evolution. Unresolved after more than a century of debate, this sorry tale is (in)famously known as the “endless dispute”5.

Much of the attention in this dispute has been directed towards the nervous system. It is widely agreed that the deutocerebrum, tritocerebrum and the more posterior parts of the nervous system are derived from successive segmental neuromeres. The deutocerebrum innervates the antennae of insects, the anterior antennae of crustaceans and the chelicerae of spiders and scorpions3. But the real meat of the endless dispute has always concerned the nature of the appendage-less front-most part of the brain, the protocerebrum. Is it some sort of non-segmental leftover inherited from the very earliest animal ancestors of the arthropods6 (a mystical structure called the acron in the literature), or does it represent the neuromere of a once appendage-bearing anterior segment7?

Two lines of evidence have been put forward in support of the existence in ancient arthropods of a protocerebrum with an appendage. First, all extant arthropods (except pycnogonids) possess a small, appendage-like outgrowth of the body which lies just in front of the mouth and is called the labrum. Confusingly, the labrum is not innervated by the protocerebrum (fanning the flames of the dispute); however, in the embryo it starts off right at the front of the animal, and migrates backwards during development. If the labrum represents a highly modified appendage, then its anterior position in development might indicate that it is the long-sought limb of a protocerebral segment.

Second, there is the fossil evidence of the earliest arthropods from 530–490 million years ago. Many of these early arthropods possessed a pair of large, grasping or branched appendages, known as the ‘great appendage’, found at the anterior of the head. Indeed, a phylogenetic reconstruction published a few years ago suggested that the great appendage was innervated from the protocerebrum7. We cannot investigate the nervous system of a fossil, however, and this reconstruction has been hotly disputed, with many researchers preferring to see the great appendage as equivalent to the antennae of insects and crustaceans8.

However, if, for the sake of argument, we accept these two lines of evidence at face value, we could reasonably conclude that the protocerebral appendage started out as a great appendage that has subsequently shrunk to the small nub of tissue we now see in most living arthropods as the labrum.

The wider significance of the conclusions of Maxmen et al.2 now becomes clear. The presence of a bona fide appendage on the pycnogonid protocerebrum (and the absence of a labrum) gives support to the protocerebral origin of the great appendage and to the idea that the labrum is the remnant of this ancient appendage. More excitingly, it implies that the pycnogonids are extraordinary living fossils, retaining an organization of their head that all other living arthropods lost hundreds of millions of years ago.

How, then, might we test these new results2? First, we would like a way to verify the association of chelifore with protocerebrum. One way to achieve this would be to use domains of gene expression as segmental markers. Hox genes are especially useful in this regard, as they have relatively stable domains of expression along the anterior–posterior axis of arthropods. This approach has been used successfully to line up the head segments of arachnids and insects, for example3. Thus, the anteriormost expression of the Hox gene Deformed (Dfd) marks the fourth segment in both arachnids (Fig. 1a) and insects. Because the arachnid/insect first segment has no associated appendage, this Dfd expression in the fourth segment lines up with the third appendage. If the traditional interpretation of pycnogonid appendage assignment is correct, the anteriormost appendage — the chelifore — will be associated with the second segment and, as in arachnids, the fourth-segment expression of Dfd will therefore be seen in the third appendage of the pycnogonid larva (Fig. 1b). But if Maxmen et al. are correct, the chelifore comes from the first segment and, counting backwards, Dfd expression in the fourth segment will therefore be seen in the fourth appendage rather than the third (Fig. 1c).

Figure 1: Testing the results of Maxmen et al.2 with Hox gene expression.

a, Deformed (Dfd) expression in an arachnid embryo (the mite Archegozetes longisetosus3) compared (b, c) with that expected in a stage III pycnogonid larva under two different reconstructions. b, If the traditionally accepted equivalence of pycnogonid chelifores and arachnid chelicerae is correct, the anterior boundary of pycnogonid Dfd will be in the third appendage. c, If Maxmen et al. are correct in associating the chelifore with innervation from the protocerebrum, that boundary should be in the fourth appendage. (Appendage innervation: proto, protocerebral; deuto, deutocerebral; trito, tritocerebral. Larval image modified from ref. 10.)

The second avenue is phylogenetic. The evolutionary scheme we have outlined implies that the transition from great appendage to labrum happened once in the common ancestor of all living arthropods apart from the pycnogonids, which must therefore be very basal in evolutionary terms. But if the pycnogonids truly are the sister group of the spiders and scorpions (which some molecular data suggest9), then the results of Maxmen et al. will be hard to square (Fig. 2). Testing the phylogenetic position of pycnogonids is therefore crucial.

Figure 2: Possible evolutionary position of pycnogonids and interpretation of Maxmen and colleagues' results2.

a, If pycnogonids branched off before the appearance of insects, crustaceans, myriapods and arachnids, we can interpret their protocerebral chelifores (red) as equivalent to the supposedly anterior great appendage of fossil groups such as Anomalocaris. The labrum (green) would have evolved in the common ancestor indicated with a star. b, But if pycnogonids are related to arachnids, then either their protocerebral chelifores could be an atavistic re-evolution of the great appendage, or the labrum must have evolved independently in arachnids and the other three taxa. Both of these latter hypotheses are contentious, and could raise doubts about the conclusions of Maxmen and colleagues.

The conclusions of Maxmen et al. overturn entrenched ideas about the body plan of the sea spiders and, furthermore, lend support to some controversial theories of arthropod evolution. Unlike their terrestrial analogues, sea spiders lack a poisonous bite, but this paper is bound to inject venom into what is already one of the most controversial of all zoological topics.


  1. 1

    Bourlat, S. J., Nielsen, C., Lockyer, A. E., Littlewood, D. T. & Telford, M. J. Nature 424, 925–928 (2003).

  2. 2

    Maxmen, A., Browne, W. E., Martindale, M. Q. & Giribet, G. Nature 437, 1144–1148 (2005).

  3. 3

    Telford, M. J. & Thomas, R. A. Proc. Natl Acad. Sci. USA 95, 10671–10675 (1998).

  4. 4

    Winter, G. Z. Zool. Syst. Evol. 18, 27–61 (1980).

  5. 5

    Rempel, J. G. Q. Entomol. 11, 7–25 (1975).

  6. 6

    Scholtz, G. Zoology 103, 99–111 (2001).

  7. 7

    Budd, G. E. Nature 417, 271–275 (2001).

  8. 8

    Chen, J. -Y., Waloszek, D. & Maas, A. Lethaia 37, 1–17 (2004).

  9. 9

    Regier, J. C., Shultz, J. W. & Kambic, R. E. Proc. R. Soc. Lond. B 272, 395–401 (2005).

  10. 10

    Vilpoux, K. & Waloszek, D. Arthropod Struct. Dev. 32, 349–383 (2003).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.