It’s April (not anymore—it’s September as I repost this), it’s Minnesota, and it’s snowing here (not yet, but soon enough). On days like this (who am I fooling? Every day!), my thoughts turn to spicy, garlicky delicacies and warm, sunny days on a lovely tropical reef—it’s a squiddy day, in other words, and I’ve got a double-dose of squidblogging on this Friday afternoon, with one article on the vampire squid, Vampyroteuthis infernalis, and this one, on squid evolution and cephalopod Hox genes.

Hox genes are members of a family of genes with a number of common attributes:

• They all contain a common 183-basepair sequence called the homeobox, which codes for a conserved 61 amino acid section of the protein, called a homeodomain.
• They are all regulatory genes that bind DNA to switch other genes off and on.
• They are found in clusters of Hox genes, strung together in a series on a chromosome.
• They are involved in spatial patterning along the longitudinal axis. The genes are turned on to define the head end, the tail end, the middle, etc., of the animal.
• Their expression pattern is collinear, that is, the order of genes in the cluster reflects the temporal and spatial sequence of gene activation in the organism.

In vertebrates, these genes get activated at that roughly defined period, the phylotypic stage, and are key players in organizing the animal’s properties along its length. The first genes in the cluster are turned on in the head, another set farther back in the neck, the next in the thorax, etc.; they are crucial in laying down the foundation of our body plan. The Hox genes are powerful regulators of other genes’ activities, and they also get re-used for other functions. Certain Hox genes get reactivated later in development and turned on in a predictable sequence along the length of the limb, for instance, to define the proximo-distal organization of your legs and arms.

Hox gene expression is also associated with segmental organization in arthropods and vertebrates. They corral groups of adjacent segments and confer specialized identities on them, such as head, thorax, or abdomen.

Molluscs are particularly interesting subjects for study of Hox gene expression for a couple of reasons. Although they may have been primitively segmented, most modern groups show little vestige of it, so how their organization is bounded is an open question. In addition, molluscs are incredibly diverse in their organization, with some amazing evolutionary specializations. Think of chitons and clams, snails and squid…all members of this same phylum, and all looking far different from one another in their apparent organization. How have they managed such morphological flexibility? What role do these pattern-forming Hox genes have in building a squid vs. an abalone?

The diagram below illustrates how a tentacled, free-swimming cephalopod is thought to have evolved from a basal monoplacophoran—a sort of snail-like beast with a simple shell and a foot and an underlying segmental architecture—by way of a plectronocerid, the common ancestor of the cephalopods. You can see how the characteristic tentacles of the squid are derived from a branching of the structure of the muscular foot.

We do know that molluscs have Hox genes, and that in the abalone larva, they are at least partially collinearly expressed in the nervous system. The question is, how are the Hox genes used in this radically modified mollusc, the cephalopod? Here is what the developing embryo of the squid, Euprymna scolopes, looks like, and you can see the problem. They acquire some novel specializations, like the ring of 10 tentacles (I-V on each side) and the funnel tube that are more radially than longitudinally organized, and it’s not clear how Hox genes would be recruited to confer specialized identities on each, if they are used at all.

Lee et al. (2003) examined the spatial and temporal pattern of 8 Hox genes in Euprymna and found a few things that were expected, and a few that were interestingly different. Like the abalone, some of the Hox genes are expressed collinearly in the nervous system; the ganglia that are at one end of the animal express the genes found at one end of the cluster, and the ganglia at the opposite end express genes found at the opposite end of the cluster. These are primitive structures, and demonstrate the ancestral pattern of anterior-posterior gene expression. Highly derived structures, like those nifty squiddy arms, use the Hox genes in a novel way. Expression is not collinear, but each of the five tentacles on each side have their own unique combinatorial pattern of Hox expression.

This diagram illustrates the level of expression of each of the 8 Hox genes (Esc-lab through Esc-Post-2 on the left side of the table) at 3 stages of development, early, middle, and late. The expression in each arm and the funnel tube are shown in the six columns; for instance, you can see that none of the Hox genes are active in Arm I early, Esc-lab, Esc-Lox5, and Esc-Post-1 are weakly active in the middle stages, and those same 3 are active late, but now Esc-Post-1 is strongly expressed.

Lee et al.‘s conclusion is that there is something interesting going on—at the same time we see morphological innovation, we see a dramatic relaxation of the rules constraining the expression of patterning genes. What may be going on here is that cephalopods have found a way to break the fairly strict coupling of the expression of adjacent Hox genes that we see in arthropods and vertebrates to allow them to be used more independently.

Hox genes have been correlated with key events in body plan evolution in ecdysozoans and deuterostomes. Here we have shown that specific Hox expression is developmentally correlated with a number of morphological innovations in a lophotrochozoan. Expression of multiple Hox genes in the brachial crown, funnel tube and stellate ganglia in cephalopods is the first reported nonvertebrate example of coordinated Hox expression in evolutionary novelties, although it does not reflect the expected collinear pattern. The Hox expression patterns suggest that in this cephalopod, the cooption of subsets of the Hox complex is not necessarily constrained by the regulatory mechanisms of collinearity seen with multiple Hox recruitment in vertebrate innovations (for example limbs, genitalia). In fact, our expression data indicate that Hox genes can potentially be co-opted to function in diverse developmental contexts, not all of which are related to patterning the anteroposterior axis. Although this is one of the first molecular investigations of body plan evolution in the Mollusca, our data, and previously reported Hox and engrailed expression patterns in other molluscs, suggest that the morphological plasticity that is so striking in this group of lophotrochozoans may be due to relaxation of regulatory constraints on the recruitment of various regulatory genes involved in morphological patterning of bilaterian animals.

---

Lee PN, Callaerts P, de Couet HG, Martindale MQ (2003) Cephalopod Hox genes and the origin of morphological novelties. Nature 424:1061-1065.