PDA

View Full Version : Deep Homology


lpetrich
28 Mar 2009, 02:30 PM
A remarkable discovery of evolutionary developmental biology ("evo-devo") over the last few decades, has been deep homology, homology across animal-kingdom phyla that has been totally unexpected.


An important part of specifying identity along the anterior/posterior (head-tail) body axis is the expression of Hox genes, whose proteins control other genes. They are expressed in zones along that axis, and their proteins control the expression of other genes.

They were first identified in Drosophila fruit flies, often used in genetics research because they are easy to raise in the lab. Mutations in them and their expression control produce conditions like antennapedia, where their antennae develop like legs, and the the third thoracic segment developing like the second one, making 4-winged flies.

Since then, they have been found over much of the animal kingdom, and with similar function, as far as can be determined. Snakes lost their front limbs when some of their Hox zones got moved forward, keeping those limbs from growing.

Hox genes most likely originated by gene duplication, with different copies becoming specialized for different subdivisions of the original regions. These genes have relatives known as the ParaHox and the NK genes. By embryonic germ layer:

Ectoderm (skin, nerves) | Hox
Mesoderm (heart) | NK
Endoderm (gut) | ParaHox

Hox gene
A brief overview of Hox genes (http://scienceblogs.com/pharyngula/2006/04/a_brief_overview_of_hox_genes.php)
Hox genesis (http://scienceblogs.com/pharyngula/2006/05/hox_genesis.php)
Bilateral symmetry in a sea anemone (http://scienceblogs.com/pharyngula/2006/05/bilateral_symmetry_in_a_sea_an.php)
Loss of legs in Snakes is linked to Hox gene expression (http://www.hoxfulmonsters.com/2008/05/loss-of-legs-in-snakes-is-linked-to-hox-gene-expression/)


Turning to the next body axis, the dorsal-ventral (back-to-belly) one, evo-devo research has revived a remarkable hypothesis put forth by Geoffroy St. Hilaire in 1822: that vertebrates' body plans are inverted relative to arthropods' body plans. In effect:

Arthropod | Vertebrate
Dorsal | Ventral
Heart | Heart
Gut | Gut
CNS | CNS
Ventral | Dorsal

where CNS = central nervous system. Annelids have the same arrangement as arthropods, which suggests that the arthropod-annelid arrangement is ancestral and that vertebrates are the inverted ones.

However, there are so many details that get in the way of this nice hypothesis that it was quickly rejected. Every twenty years or so, someone would revive it, only to see it get picked apart.

But recently, some remarkable homologies have been discovered among the molecules responsible for dorsoventral patterning. Vertebrate ones were discovered in the embryos of Xenopus frogs, often used as lab animals, and were discovered to be homologous to Drosophila ones.

Fly | Frog
Dorsal | Ventral
dpp | BMP
sog | chordin
Ventral | Dorsal
dpp = decapentaplegic
sog = short gastrulation
BMP = Bone Morphogentic Protein

dpp/BMP makes cells have a dorsal/ventral fate, but is antagonized by sog/chordin, which makes cells have a ventral/dorsal fate. What is especially interesting is that the frog version of sog/chordin works in flies and vice versa.

Going further, the marine annelid worm Platynereis was discovered to have a similar patterning mechanism -- and to have a patterning mechanism for its central nervous system that closely parallels that of Drosophila and vertebrates. So despite the differences in overall shape of their respective central nervous systems, the patterning mechanisms are remarkably similar.


We have the brains of worms (http://scienceblogs.com/pharyngula/2007/05/we_have_the_brains_of_worms.php)
Geoffroy's lobster and the animal common ancestor (http://chancenecessity.blogspot.com/2009/02/geoffroys-lobster-and-animal-common.html)


Going from nervous systems to eyes, yet another bit of deep homology is in a molecule involved in early eye development. The Drosophila eyeless gene is homologous to the mouse Pax6 gene, and Pax6 can make Drosophila flies grow ectopic eyes. However, those eyes resemble fly eyes rather than mouse eyes.

And even weirder is the two types of photoreceptors in the animal kingdom, "rhabdomeric" and "ciliary". Invertebrate eyes always have rhabdomeric photoreceptors, which grow out of the skin, while vertebrate ones always have ciliary ones, which grow out of the brain.

But a remarkable "missing link" has recently been discovered. Platynereis worms have rhabdomeric-photoreceptor eyes, like other invertebrates, but they also have ciliary photoreceptors in their brains, which are possibly for setting their circadian clocks. And these ones are closest to vertebrate photoreceptors.

So the ancestor of the bilaterians likely had both kinds of photoreceptors, rhabdomeric ones on their skins and ciliary ones in their brains.


And limbs? Fruit flies have a gene called distalless, which is involved in the growth of their legs. Mutations of it can interfere with that growth, thus the name. Similar genes are involved in the growth of various appendages in other phyla, like fish fins / land-vertebrate limbs, starfish tube feet, and annelid parapodia. Though these structures are not homologous across phyla, their growth mechanisms thus have some cross-phylum homology.

And a heart? Yes, that also. Drosophila has a gene involved in heart growth called tinman, named after what mutations of it can do. Vertebrates have homologues of tinman also, the NK2 genes.

From all this, one can infer that the ancestor of most of the animal knigdom had had a front-to-rear body axis patterned with Hox genes, a belly-to-back axis, a simple central nervous system, photoreceptors, a contractile vessel for pumping internal fluids, and some sort of appendages.

Jobar
28 Mar 2009, 05:27 PM
From all this, one can infer that the ancestor of most of the animal kingdom had had a front-to-rear body axis patterned with Hox genes, a belly-to-back axis, a simple central nervous system, photoreceptors, a contractile vessel for pumping internal fluids, and some sort of appendages.
Are there any ideas about what this earliest 'reversed' ancestor might be? Some sort of ur-fish?

lpetrich
29 Mar 2009, 08:11 AM
That question has indeed been addressed. Vertebrates' closest relatives are invertebrate chordates, Cephalochordata (amphioxus) and Urochordata (tunicates, etc.), and they share vertebrate dorsoventral inversion.

So we have to look in the remaining deuterostomes, the hemichordates and echinoderms.

Chordates have a gene called Pitx2 that is expressed on the left-hand side of their early embryos, thus marking out the third body axis: left-right. But some people found the sea-urchin and starfish larvae express a homologous gene on the right-hand side. This means that echinoderms are inverted relative to chordates, which means that they are not inverted relative to most non-chordates.

Pitx homeobox genes in Ciona and amphioxus show left–right asymmetry is a conserved chordate character and define the ascidian adenohypophysis (http://www3.interscience.wiley.com/journal/118932869/abstract) (anterior pituitary)
Phylogenetic correspondence of the body axes in bilaterians is revealed by the right-sided expression of Pitx genes in echinoderm larvae. (http://www.ncbi.nlm.nih.gov/pubmed/17118013?ordinalpos=2&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsP anel.Pubmed_DefaultReportPanel.Pubmed_RVDocSum)


Looking at the remaining deuterostomes, the hemichordates, there is evidence that chordates are inverted relative to them also. They have the same dorsoventral patterning system, and some homologous CNS-layout molecules, though their nervous system is diffuse rather than concentrated. But here also, this patterning system is inverted relative to chordates.

Dorsoventral Patterning in Hemichordates: Insights into Early Chordate Evolution (http://biology.plosjournals.org/perlserv/?request=get-document&doi=10.1371/journal.pbio.0040291)
Hemichordates and the origin of chordates. (http://www.ncbi.nlm.nih.gov/pubmed/15964754?ordinalpos=4&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsP anel.Pubmed_DefaultReportPanel.Pubmed_RVDocSum)
The deuterostome ancestor. (http://www.ncbi.nlm.nih.gov/pubmed/17001683?ordinalpos=1&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsP anel.Pubmed_DiscoveryPanel.Pubmed_Discovery_RA&linkpos=1&log$=relatedreviews&logdbfrom=pubmed)
Evolutionary biology: hedgehog crosses the snail's midline. (http://www.ncbi.nlm.nih.gov/pubmed/12075342?ordinalpos=6&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsP anel.Pubmed_DefaultReportPanel.Pubmed_RVDocSum) (A certain gene expressed dorsally in vertebrates has a homologue in limpets that is expressed ventrally)


So as far as can be determined, only chordates have this inversion. So some proto-chordate must have flipped over. And that proto-chordate was likely much like the present-day amphioxus, mid-Cambrian Pikaia, or early-Cambrian Yunnanozoon.

Berthold
29 Mar 2009, 09:24 AM
One thing that amazes me: How do the molecular mechanisms "know" where the front end is; I mean, in deutero- vs. protostomes?

DMB
29 Mar 2009, 09:28 AM
I liked "Man is but a worm". Or I suppose we could say "Man is but an upside-down worm".

lpetrich
29 Mar 2009, 06:04 PM
In response to Berthold's question, This page (http://www.nature.com/embor/journal/v2/n12/fig_tab/embor266_f1.html) and this page (http://nobelprize.org/nobel_prizes/medicine/laureates/1995/illpres/nuss-wiesch.html) show some diagrams of what happens in Drosophila.

The flies' eggs start off with the mother fly creating gradients of proteins called bicoid and nanos. The varying amounts of these proteins trigger and suppress the expression of various other genes, making them expressed in thick bands. The proteins they produce then control the expression of some other genes, which then get expressed in thinner bands, until one reaches the laying out of segments.

However, Drosophila is rather unusual in being a "long-germ" insect, one where all the segments are laid out at once. They share this feature with other Diptera (flies, mosquitoes), Lepidoptera (butterflies, moths), Hymenoptera (bees, wasps, ants), and some Coleoptera (beetles). Most other insects are "short-germ", with only the first few segments forming all-at-once, or "intermediate-germ", with around half of the segments forming all at once. The other segments form from a rear-end segment-making growth zone.

This rearward growth is what other arthropods do, and also what the other big segmented groups do: annelids and chordates.


There is a curious conundrum about segmentation. Did it evolve separately three times? Or did it evolve only once in the shared ancestor of most of the bilaterian phyla, only to be lost in many of its descendants? We don't know enough about segmentation mechanisms to be able to say one way or the other, but there is evidence of some interesting similarities, like And Lophotrochozoa makes three: Notch/Hes signaling in annelid segmentation. (http://www.ncbi.nlm.nih.gov/pubmed/19011897)

lpetrich
14 Apr 2009, 08:28 PM
From PZ's excellent blog Pharyngula: Snails have nodal ! (http://scienceblogs.com/pharyngula/2009/04/snails_have_nodal.php) reporting on Nodal signalling is involved in left-right asymmetry in snails. (http://www.ncbi.nlm.nih.gov/pubmed/19098895)

The gene nodal is involved in producing left-right asymmetry in chordates, being expressed more on the left rather than the right side. Its protein activates other left-right genes like Pitx.

But the new result here is that snails also have nodal and Pitx, and that it is also involved in left-right asymmetry. Snails with dextral (right-handed) coiling have nodal and Pitx expressed on the right side, while snails with sinistral (left-handed) coiling have nodal and Pitx expressed on their left side.

Ancestral snails were likely dextral, having nodal and Pitx expressed on their right sides.

Interestingly, sea urchins also express nodal and Pitx, in the right side of their larvae -- Phylogenetic correspondence of the body axes in bilaterians is revealed by the right-sided expression of Pitx genes in echinoderm larvae. (http://www.ncbi.nlm.nih.gov/pubmed/17118013)

So nodal and Pitx were likely involved in specifying left-right asymmetry in the ancestral bilaterian, with the chordates getting a 180-degree rotation around their nose-tail body axis:

Non-chordates (composite):
- | BMP/dpp, heart | -
----- | gut | nodal, Pitx
- | chordin/sog, CNS | -

Chordates:
- | chordin/sog, CNS | -
nodal, Pitx | gut | -----
- | BMP/dpp, heart | -

By comparison, ecdysozoans (arthropods, nematodes, etc.) have no Nodal, and their expression of Pitx is not involved with left-right asymmetry. Checking on more species would be good, of course -- especially more lophotrochozoans (mollusks, annelids, etc.), like the polychaete annelid worm Platynereis.

Overall taxonomy:

Bilateria: (Protostomia, Deuterostomia)
Protostomia: (Ecdysozoa, Lophotrochozoa)
Deuterostomia: (Chordata, (Hemichordata, Echinodermata))