Centipedes
Hox genes, segmentation, and the evolution of the Myriapod body plan

The centipedes and millipedes are members of a group of arthropods called the Myriapods ("the many-legged"). Their long bodies with many pairs of legs make them fascinatingly creepy.

How did they evolve such an unusual body plan?

There are big differences between the body plan organization of a centipede versus a fruit fly. Since the body plan gets set up during the process of embryonic development, the morphological differences must reflect differences in what developmental genes are doing in the embryo of a centipede versus the embryo of a fruit fly.

All genes are not created equal -- some genes play major roles in early development and help set up the organization of the body, while other genes act later in development to fine tune specific details of bristles and pigmentation. To understand how the body plans of arthropods evolved, we need to focus on major developmental regulatory genes.

Some of the important developmental regulators are a set of genes called the Hox genes.

The Hox genes are known to play important roles in assigning segmental identity -- telling cells which segment they are in along the AP axis, and turning on a bunch of downstream target genes to make the appropriate appendages for that segment. The Hox genes are highly conserved and are thought to play similar roles in both vertebrates and invertebrates. Because these individual genes play such powerful roles in specifying segmental identity, they have been prime suspects for genes that could play major roles in the evolution of the body plan. Tweaking just one Hox gene could have a dramatic result.

I worked with the centipede species Lithobius atkinsoni. The embryo has a thin germband that lies on one side of the yolk in a spherical egg. With practice, the embryos can be dissected and in situ hybridization can reveal the gene expression patterns.

The in situs for the Hox genes revealed several interesting things. One surprise was that centipede appear to be using ten genes as Hox genes to pattern the body -- more than the eight genes that fruit flies use.

The gene Hox3 is expressed in the front mouthpart segments, while the gene fushi tarazu is expressed in the posteriormost mouthpart segment and in the fangs. In addition to the Hox-like domain, fushi tarazu also is expressed in the growth zone and in new segments of young embryos, suggesting a possible second role in segment formation.

The fangs are a modified structure unique to centipedes -- they are the front pair of legs that have evolved to be more like mouthparts, and are used for prey capture. In keeping with the morphological observation of mixed identity, the Hox genes showed an overlap in the fangs of "mouthpart genes" (Scr, ftz) and "trunk genes" (Antp).

The Hox genes suggested some of the basis for the organization of the centipede body plan, but a big part of what makes centipedes special is all those segments. In geophilomorph centipedes, the number of segments varies from species to species, and can be different between males and females. How is the formation of segments regulated? How does a centipede embryo count segments? And how does the sequential formationg of segments in a centipede compare to the simultaneous formation of segments in a fruit fly?

One of the most important segmentation genes in flies is the gene even-skipped. In flies, even-skipped is expressed in broad stripes, and when mutated half the segments are missing. But in centipedes, the segments form one-by-one from a growth zone at the tail. How is even-skipped expressed in a centipede?

While the pattern of even-skipped in a centipede looks superficially very different from that in a fly, the pattern here reflects a similar role in the formation of segments. eve is expressed strongly in the undifferentiated tissue of the growth zone, and in stripes between each of the newly-forming segments.

The pattern of eve and other segmentation genes in long-germband arthropods is strikingly reminiscent of gene patterns during somitogenesis in vertebrates. This suggests that in some ways at least, a centipede is a better model for vertebrate development than a fly!

The expression analysis of segmentation genes and Hox genes described here sheds a little light on development in the centipede. But there is still a lot we don't know about how segments are produced and specified in these arthropods.

For instance, many myriapods (including Lithobius) continue to add segments in post-embryonic development. Imagine -- most of the segments of the little hatchling centipede have formed and differentiated completely so that it can run around and eat, but inside it is still carrying around the tail of an embryo, with undifferentiated stem cells continuing to churn out segments one at a time. Perhaps it is a useful survival strategy to escape the unprotected egg even before it is quite "done" with development -- an impressive feat!