Fins are not fingers

The evolution of arms and fingers from fish fins, a story of genetic redeployment.

There is still a great deal to learn about how our bodies and minds rise out of our genetic code. Despite a growing flood of genomic data- and we are right on the verge of the $1000 genome, meaning that everyone in the developed world will shortly have their genome sequenced as a matter of medical routine- a vast gulf remains between the concrete knowlege we now have about the two ends of the process: genotype and phenotype.

One of the great American scientists of the 20th century was Edward Lewis of Cal Tech, who studied the developmental genetics of fruit flies, focusing on mutations that affected their body plan. In one example, he developed mutants whose third thoracic segment, instead of growing tiny winglets called halteres, grew full wings, just like their second thoracic segment. They were a little like dragonflies. This led Lewis on a long path to characterize such "homeotic" mutations, (which transform body parts), and to a Nobel prize.

It is now known that the main gene Lewis studied, "Ultrabithorax" encodes a transcription regulator that sits in the middle of a large developmental network or cascade of transcription regulators. The process starts from the glimmerings of polarity determination in the mother's egg, and proceeds through successively finer divisions of space and identity within the growing embryo until we get to the ultimate effector genes that direct neuron production and migration, or muscle development, or one of a thousand other cell types that generate our tissues.

The genes that Lewis studied are collectively termed "hox" genes, short for homeobox, which itself is short for a DNA-binding motif that is found in all these genes whose mutations cause homeotic transformation, which has a characteristic DNA and protein sequence, only subtly altered in each one. They are all related because they are all evolutionary copies of a single ancestor.

These genes sit in the middle of the developmental cascade, and have themselves vast upstream regulatory regions, to gather regulatory information from earlier stages in the process. Segmentation has happened by the point they come into action, and the homeotic genes integrate the data about which segment their cell is in, and, if conditions are right, turn on expression of their encoded regulatory protein, thereby providing input to all the downstream genes that actually prompt the development of that segment's proper parts, be they wings, legs, antennae, arms, etc.

Hox genes occur in tandem clusters, and the clusters themselves have been duplicated during evolution. In the diagram above, (from wikipedia), sea urchins, at the top, have something like the original cluster of eleven hox genes, color coded by their position in the cluster, which also relates to the position along the body axis where they are expressed (at right). Fruit flies, at the bottom, lost a few copies, and gained a few others, but retain basically the same system. Fish and tetrapods, in the middle, duplicated the entire set, copying whole clusters to various chromosomes, and lost individual hox gene units along the way. This elaboration allowed more complicated body plans to develop, with the example of fingers being a new use of the hox code, added onto the basic body trunk segment-by-segment code. The head and brain are another place where the hox system has been re-used in tetrapods.

One confusing element of the field is that in tetrapods, the hox A and D clusters are partly redundant. Each can, on its own, direct formation of arm and fingers, and both need to be deleted to eliminate the arm. So the researchers in today's paper mix and match from both clusters to make their various points.

"During mammalian limb development, the activity of both HoxA and HoxD gene clusters is essential and the absence of these two loci leads to rudimentary and truncated appendages."

In the embryonic hand, expression of many Hox D genes, from d9 to d13, are required to specify tissues during development, as are a few of the Hox A genes. They have overlapping patterns rather than some neat, digital(!) code, this being messy biology, but through mutation and other studies, researchers have pieced together some information about which gene of the tandem arrays does what. The genes have some individual characteristics, but much of their regulation is collective, directed from enormous regions on both sides of the cluster, comprising over three million base pairs of DNA.

The Hox D locus, on human chromosome 2. It contains eight distinct hox genes, (numbered black boxes at bottom), flanked by enormous control regions on either side which drive expression of some cluster genes in the hand (blue) and some in the arm (red), responding to transcription regulators earlier in the cascade of developmental patterning and differentiation. What are those fancy-looking blue and red cubic designs? That reflects a separate study where the authors physically tested which DNA was close to which other DNA in embryonic cell chromosomes. And they found that the right and left regions form their own knotted-up domains each hooking up with the central hox D gene, but not touching anything on the opposite side.

A recent paper is one of a pair that find that two clusters, hox D and hox A, are both flanked by very large regulatory regions that in fish have only slight differentiation, one directing slightly more distal (towards the outside) expression than the other one (red). The large regulatory region downstream (red) which originally specified expression in fish fins, has pretty much retained the same function in tetrapods, specifying the arm.

But the large regulatory region on the other side (blue) in fish only adds a little bit of extra expression to some cluster members towards the outside of the limb. In tetrapods, however, it specialized to direct expression of hox D genes in the hand, quite exclusively from directing expression anywhere else. The basic finding is that fish fins are not proto-fingers, really, but are related principally to our arms. The fingers arose from a mostly new regulatory program established by the blue areas in the genome shown above. And the wrist ... that is specified in the gap, partly by the lack of hox expression. It is interesting to note as an aside that the hox B and hox C clusters seem to have regulatory control only from one side, not both sides.

Inference of the paper, that the hand-specifying regulatory  regions of hox D and hox A (blue) developed from earlier regions (yellow) that had relatively minor roles in fish, and which specified the margin of the fin, rather than a separate structure.

What is some of their evidence? Well, first, let's see some of the native expression of mouse hox A genes:

Expression of individual genes from the mouse hox A cluster, showing finger-specific expression for 9, 10, 11as, and 13. The exception of hox A11 is striking, as a departure from the hand-specific pattern of its nearby siblings, and in its well-defined zeugopod, or lower-arm expression pattern.

One obvious experiment was to transplant the fish hox DNA into mice to ask where it gets expressed. It always gets expressed in the same place- where the arm expression happens, at the base of the limb bud, not where finger expression happens. This makes the case pretty strongly that finger expression and development was, as one might imagine, a novel evolutionary development.

Mouse embryonic limb buds showing the expression of a transgenic zebrafish hox A cluster, with regulatory regions and genes it contains, including each of the ones as labeled. They all get expressed in the near, or arm region, not in the finger region. This was true no matter which regulatory region of the zebrafish hox A cluster was used, whether the upstream or the downstream side.

Even more striking, the researchers show expression patterns in complete embryos. Below is a stage E11.5 mouse embryo with transgenic fish hox A13, driven by the fish regulatory region corresponding to what would be the hand/finger-specifying region on tetrapods. Its expression appears in many areas of the body, but not in the fingers, as the mouse's own hox A13 does. It is worth noting that in vertebrates, the hox genes are used all over again in specifying brain region development, which does not happen in flies. It is a common theme- that through the accumulation of regulatory complexity, the same genes can be re-used many times to create ever more elaborate phenotypes.

As you can see from the genome locus diagram a few figures above, the regulatory regions controlling the hox D genes are far, far larger than the protein-coding genes themselves. Complexity of control is the theme in all genomes, especially ours. These regions contain many little modular bits of DNA that bind to various other transcriptional regulators that operate from upstream in the developmental cascade, allowing a progressive, step-by-step, though in actuality a stochastic and mix-and-match evolutionary process whereby the silk purse of our present bodies are made out of the sows' ear of a few thousand ancient genes.

• 23 & me genetic testing- another front in privacy and big data.
• Example of another paper on limb formation, in the transcription regulator cascade of development.
• Creationism map.
• The POTUS with the SOTUS- does work pay the worker, or only the CEO?
• These kids just don't understand religion!
• The patent backstory to the Google, Motorola, and Nortel deals.
• Fascism, American style- corporations and the blacklist.
• Economic quote of the week, from John Schmitt:

"Workers today are a lot older than they were in the 1960s or the 1970s, and they are enormously better-educated than they were in the 1960s or 1970s. The fact that most workers are doing barely better, and some workers are doing worse than their counterparts from 40 or 50 years ago … suggest that the problem is that the way the economy converts people’s skills, people’s experience, people’s education and their training, into good jobs is what has deteriorated over this period. Not people’s underlying skills, or work experience, or education."