firebrat Hox genes

Firebrats and rogue Hox genes
or, "How do genes change jobs?"

Modern molecular biology takes advantage of the broad general conservation of genes across animal species. The level of conservation is pretty amazing -- we can study development in a fruit fly, and learn about genes and mechanisms that are important in development and disease in humans.

But at the same time, genes also have undergone changes in evolution -- all the differences we see between a fly and a spider and a fish and a human are due to changed genes. Some genes help create new traits by relatively minor changes in their expression pattern or function, but other genes have evolved dramatically different expression and function to carry out a very different role. How can genes evolve to take on whole new roles in development without disrupting the process? In other words, how do genes change jobs?

To study how genes change, we need examples of genes in transition. By studying species that branched away at different timepoints, we can see snapshots of the process of change. These aren't perfect representatives of the ancestral state of course, since other confounding changes may have occurred over the millions of years. But comparing a series of species at different points on the tree can help use to recognize the pattern of change, and maybe even infer how that change was able to occur. In my work, I used arthropod species to study two genes in transition -- the genes called fushi tarazu and Hox3/zen/bicoid. The firebrat Thermobia domestica is a primitive wingless insect, and it provided an valuable data point for understanding how these genes evolved to change roles.

In the fruit fly Drosophila, there are eight Hox genes, called labial, proboscipedia, Deformed, Sex combs reduced, Antennapedia, Ultrabithorax, abdominal-A and Abdominal-B. These genes are famous for the bizarre phenotypes that result when they are mutated -- flies grow legs out of where the antennae are supposed to be, or have legs instead of mouthparts, or they have two pairs of wings instead of one.

These transformations revealed that the Hox genes play critical roles in setting up the organization of the body plan. In flies (and it turns out, in all animals) the genes form a rainbow of expression from the head to the tail, and give identity to each region of the animal. So the gene proboscipedia is on in segments of the head, and tells those cells to make mouthparts. If the gene is missing, those cells are confused, and end up making legs instead. Because Hox genes are so important for assigning identity along the anterior-posterior axis of the embryo, they seemed to be likely candidates for genes that can reshape the body plan during evolution (see my work on centipedes for more on this concept).

In general though, the Hox genes are paragons of conservation -- similar genes are doing similar jobs in the development of flies, mice, and chickens, and they've been consistently doing those jobs for 500 million years. So it came as quite a shock when people realized that there were two extra Hox genes in arthropods, and that fruit flies were using a reduced set. Centipedes, for instance, have ten Hox genes to build their body plan, but fruit flies get away with just eight genes

eight Hox genes in fruit fly embryo

ten Hox genes in centipede embryo

What happened to those other two Hox genes? They didn't disappear -- they changed roles. Each fruit fly has a copy of the Hox6 gene, called fushi tarazu ("fewer segments" in Japanese). The mutant is missing segments because it plays an important role in segmentation. And flies have two genes related to Hox3 that are called zen and bicoid. These genes specify the dorsal part of the egg and the anterior end of the embryo, respectively.

So how and when did the gene Hox3 change from playing a Hox gene role (specifying segment identity) to playing a zen-like role in dorsal specification of extraembryonic tissues?

In the firebrat embryo, the expression patterns of the Hox3/zen gene suggests it may be playing two different roles. It's main expression domain consists of strong expression in the mouthpart segments, like a typical Hox3 gene. But the gene also has faint expression in the extraembryonic membrane around the embryo something like zen.

These patterns may suggest that the Hox3/zen gene went through a phase where it performed two different roles in development, and that in more derived insects it specialized to play only the zen role. (Much later in insect evolution, the gene duplicated to form the separate genes zen and bicoid, and bicoid evolved dramatically to play a specialized anterior-determining role. See the papers by Stauber et al.)

For the gene fushi tarazu, the expression pattern in firebrats is similar to that of other segmentation genes in long-germband arthropods. In late embryos there is expression in the developing nervous system. These patterns suggest that like more derived insects, in firebrats the ftz gene is playing a role in segmentation and neurogenesis, but not a typical Hox-gene role. The Hox role may have been lost somewhere before the divergence of the first insects.

The roles of the Hox3/zen/bicoid and fushi tarazu genes are summarized here for various arthropod species. (In most cases, the role is inferred from the expression pattern of the gene.) These genes serve as a useful example of important developmental regulators that have changed roles dramatically during evolution. A greater understanding of how these changes occurred would help us to understand a major part of the puzzle of morphological evolution.