Evolving proteins in snakes

| 35 Comments
Blogging on Peer-Reviewed Research

We've heard the arguments about the relative importance of mutations in cis regulatory regions vs. coding sequences in evolution before — it's the idea that major transitions in evolution were accomplished more by changes in the timing and pattern of gene expression than by significant changes in the genes themselves. We developmental biologists tend to side with the cis-sies, because timing and pattern are what we're most interested in. But I have to admit that there are plenty of accounts of functional adaptation in populations that are well-founded in molecular evidence, and the cis regulatory element story is weaker in the practical sense that counts most in science (In large part, I think that's an artifact of the tools — we have better techniques for examining expressed sequences, while regulatory elements are hidden away in unexpressed regions of the genome. Give it time, the cis proponents will catch up!)

This morning, I was sent a nice paper that describes a pattern of functional change in an important molecule — there is absolutely no development in it. It's a classic example of an evolutionary arms race, though, so it's good that I mention this important and dominant side of the discipline of evolutionary biology — I know I leave the impression that all the cool stuff is in evo-devo, but there's even more exciting biology outside the scope of my tunnel vision. Also, this paper describes a situation and animals with which I am very familiar, and wondered about years ago.

taricha.jpg

When I was a graduate student in Oregon, I worked now and then with an emeritus faculty member named Jim Kezer — a great guy who was classically trained in natural history, and who would dazzle us benchies by taking us on field trips into the Oregon Cascades, where he could name every weed and insect we'd encounter, and he'd tell us all kinds of stories about these otherwise almost unnoticeable organisms. We made collecting trips up into a remote lake where we'd harvest rough-skinned newts, Taricha granulosa, for histology studies. This lake was swarming with newts — it was pretty much the only large animal you'd find there, and that was because they had a potent biochemical defense mechanism: they oozed a neurotoxin. These newts were not popular denizens of the lakes, because where they were found, the fish and frogs soon disappeared.

ttx.jpg

The toxin they secreted is called tetrodotoxin, or TTX. It's the same nasty substance that the pufferfish, fugu, contains — it binds the sodium channels of the nerves, blocking all electrical transmission. It's notoriously popular in sushi because at low doses it can cause a tingling sensation, similar to what you felt when the novocaine was wearing off after your visit to the dentist, and it also provides the titillating thrill of danger. Overdoses cause a flaccid paralysis, and can be lethal. More than a mild tingle, I suspect it's that entirely psychological frisson that this food might just kill you that lends fugu its culinary notoriety.

The newt has no other defenses. They don't have fangs or claws, they are as soft as noodles, and so these lakes are reduced to big bowls of squirmy delicate amphibian meat that is frustratingly untouchable by most predators because of the unfortunate fact that they are also using a nasty biotoxin in violation of all of the rules of the Geneva Convention. You might expect that if something…evolved…a countermeasure, this would be a situation ripe for exploitation.

And so it is. Some of the most successful predators of small amphibians are another herpetological marvel, the garter snakes, Thamnophis. Unfortunately, if you feed ordinary garter snakes a diet of rough-skinned newts, they tend to move more and more slowly as the innervation of their skeletal muscles undergoes a toxin blockade, and if they eat enough, they die. This is not a good thing from the snake's perspective, although the newts do get revenge and their relatives benefit from the subsequent reluctance of snakes to eat them. It also presents an evolutionary opportunity, in that resistance to TTX in snakes can be a real advantage, since they won't die and they'll be able to feast on squishy purplish-brown and orange tubes of meat.

This is happening right now. Populations of garter snakes, T. sirtalis, in California, Oregon, and Idaho are showing different degrees of resistance to TTX, and these differences are being traced right down to specific changes in the amino acid sequence of the snake sodium channel. It's happening repeatedly, too, with different populations independently acquiring different variations that confer differing degrees of resistance.

We know a lot about the structure and biophysics of the sodium channel — it's one of those universal proteins we find all over the animal kingdom. It's a protein that loops through the membrane multiple times, forming four cylindrical domains. These cylinders pack together, leaving a space at the center that is the pore proper; there are also regions of the protein that act as gates, opening to allow sodium to flow through and generate an electrical current, or closing to block it.

nachannel.jpg

We also know how TTX works. It binds especially strongly to an aromatic amino acid on the outside of the cell, in domain I. In that place, it effectively blocks the pore, making the channel permanently closed so no current flows.

Obviously, the animal that must most effectively resist the effects of TTX is the one that is producing the toxin. Species that make TTX, like fugu, typically replace that aromatic amino acid with one that doesn't bind TTX. It's a testimony to the hit-or-miss nature of mutations and evolutionary change that the snakes haven't stumbled onto that same change—they've instead made other small changes to the protein to reduce binding of TTX. Instead, they've tweaked the pore helix and β-strand from domain IV, which also reduces the effectiveness of TTX binding.

Here's a summary tree diagram of the differences found in these populations. We're looking at 5 different populations of snakes, named after their collection sites; Benton and Warrenton are in Oregon, Willow Creek is in California, and Bear Lake is in Idaho. Illinois represents the ancestral phenotypic state, a population from a state without TTX-secreting newts, and which has no TTX resistance.

TTX resistance is measured in MAMUs, or mass-adjusted mouse units — low numbers mean they have no particular resistance, while large numbers indicate increasing resistance. The Bear Lake and Illinois populations are sensitive, while the others have varying degrees of resistance.

The right side of the figure is the interesting bit: it shows the amino acid sequence of a small stretch of the protein in domain IV, and you can see the differences. All the resistant populations have a valine at position 1561, but notice that it is likely that these represent two independent origins. That valine alone only weakly improves resistance; the Benton population has an additional amino acid substitution that doubles the resistance. Willow Creek snakes have substantially greater resistance, and they also have 3 other different substitutions.

ttx_resist_seq.jpg
Amino-acid sequence differences for four snake populations. a, Phylogeographic relationships based on mitochondrial DNA analysis of 19 North American populations of Thamnophis sirtalis indicate separate origins of elevated resistance to TTX in the Willow Creek population compared with populations from Benton and Warrenton. Bear Lake is from a third lineage and is not resistant to TTX. Whole-animal TTX resistance for each population is reported in mass-adjusted mouse units (MAMU); branch colours reflect statistically distinguishable levels of resistance. Whole-animal TTX resistance was measured as a mass-adjusted dose of TTX (MAMU) that produced an average of 50% decrease in snake sprint speed in each population. b, Amino-acid alignment of part of the domain IV S5–S6 linker that affects TTX binding from tsNaV1.4. Green, pore α-helix; purple, β-strand; asterisk, selectivity filter. Dots indicate identical amino acids and grey shading highlights sequence differences between populations. Despite independent evolutionary histories, all resistant snakes share the substitution of valine for isoleucine at position 1,561.

There are other details — the proteins have been isolated, chimeric proteins generated to isolate specific regions, and they've been expressed in Xenopus oocytes, all demonstrating that these small changes are actually responsible for conferring TTX resistance. The meat of the story, though, is that we have concrete measurements of specific molecular changes that are responses to an evolutionary arms race, and we're seeing these differences emerge in different populations of a single species. This is evolution in action, and the observed appearance of new properties, traced right down to single changes in proteins.


Geffeney SL, Fujimoto E, Brodie ED III, Brodie ED Jr., Ruben PC (2005) Evolutionary diversification of TTX-resistant sodium channels in a predator-prey interaction, Nature 434:759–763.

Soong TW, Venkatesh B (2006) Adaptive evolution of tetrodotoxin resistance in animals. Trends Genet. 2006 Nov;22(11):621-6.

35 Comments

they are also using a nasty biotoxin in violation of all of the rules of the Geneva Convention.

Article 4 of the Geneva Convention says the soldiers have to be in uniform. Since the fish and frogs are buck naked, they are enemy combatants.

Ah, so Myers was a Duck (I wuz a Beaver) … and did time in the People’s Republic of Eugene. Both of us have no doubt lost our webbed feet … in his current environment I imagine he is evolving resistance to mosquitoes. “Do you want to take him to Michigan, or eat him here?” “Nah, if we take him to Michigan the big guys’ll take him away from us.”

Cool stuff on snakes. I trying to avoid running down the garter snakes hiding in my lawn when I’m mowing – I haven’t done one in for a few years, I keep wondering if there’s a selection effect in the matter and I’m evolving snakes that know how to avoid lawn mowers.

White Rabbit (Greg Goebel) http

The different resistance strategies are obvious evidence of front-loading, right?

PZ

I’m shocked that you didn’t mention that TTX is also found in a cephalopod, the Blue-ringed Octopus!

More intriguingly, there is good evidence that TTX is not synthesized by all of the critters who have managed to make it work for them; it is probably a bacterial product. See

http://www.chm.bris.ac.uk/motm/ttx/ttx.htm

Somebody post on this: http://www.badscience.net/2008/06/a[…]nist-pwnage/

shafly takes on R. Lenski – too funny

Albatrossity said:

More intriguingly, there is good evidence that TTX is not synthesized by all of the critters who have managed to make it work for them; it is probably a bacterial product. See

http://www.chm.bris.ac.uk/motm/ttx/ttx.htm

Intriguingly indeed, thanks. Symbiosis is one of the most fascinating evolutionary phenomena.

Kinda nice to make a posting not shooting peas at a Darwin-basher for a change. It gets a little tiresome.

White Rabbit (Greg Goebel) http://www.vectorsite.net/tardarwin.html

The referenced articles, incidentally, said that TTX was as deadly as saxitoxins, which is a little scary. I wrote an online document about chemical and biological warfare and found the spooks got into saxitoxins obtained from marine cone snails (which I think get them in turn by eating dinoflagellates).

When Francis Gary Powers was shot down over the USSR in his U-2 spyplane on 4 July 1960, his Soviet interrogators took a silver dollar off of him. It had a needle inserted into it that was coated with saxitoxins and he told them: “Be careful how you handle that.” They pricked a dog with it and the beast simply fell over and died immediately.

White Rabbit (Greg Goebel) http://www.vectorsite.net/tadarwin.html

(which I think get them in turn by eating dinoflagellates).

If so, it would be via bio-accumulation.

Cones are all predatory, and all eat annelid worms or larger things.

the really toxic ones are all piscivores, IIRC.

Ichthyic said:

Cones are all predatory, and all eat annelid worms or larger things.

the really toxic ones are all piscivores, IIRC.

I vaguely recall some cones can actually shoot poison darts. Myth?

White Rabbit (Greg Goebel) http://www.vectorsite.net/tadarwin.html

The toxin they secreted is called tetrodotoxin, or TTX. It’s the same nasty substance that the pufferfish, fugu, contains — it binds the sodium channels of the nerves, blocking all electrical transmission.

The full story is even slightly more interesting. TTX blocks the voltage-gated sodium channel. This fascinating protein is what allows the propagation of action potentials, and thus, the existence of complex nervous systems in general.

Resting cells are usually have a very slight negative potential relative to the extracellular fluid. http://en.wikipedia.org/wiki/Restin[…]ne_potential. I’m proud to say that I remembered that it was -70 milivolts.

This potential is maintained by ionic pumps, which are pretty interesting in their own right.

Some positive ions, mainly potassium, will flow out of cells anyway, if given the chance, due to a concentration gradient - there’s relatively a lot of it on the inside of the cell membrane.

Na+, on the other hand, will flow in, despite its many superficial similarities to K+, again, mainly because of a concentration gradient (the slight negative polarity of the intracellular fluid is a much lesser factor).

We’re talking about naked ions here, more or less - charge bearing particles, not molecules of salts. So Na+ in (or Ca++, another common and important actor) makes the intracellular side of things more “positive”, relative to the extracellular side, than it was before.

Of interest, both K+ and Cl- specific channels tend to have the opposite effect of Na+ channels, because K+ goes out and Cl- comes in, due to concentration gradients.

When the potential on the intracellular side of the membrane becomes more positive in terms of relative potential, it’s called “depolarization”. When the opposite happens, it’s called “hyperpolarization”.

What the voltage gated Na+ channel does is incredibly interesting. When a threshold depolarization level is reached, at first, it opens more, creating a positive feedback loop as Na+ ions flow in, and a comparitively massive depolarization.

Then an individual voltage gated Na+ channel shuts down and goes into a state that is insensitive to depolarization for a while. But the protein tends to be expressed on the axons of neurons, among other places. So it results in a wave of depolarization traveling down the axon. Because when one voltage gated channel is shut down, the next ones down the axon are just experiencing the depolarization necessary to trigger the positive feedback. And then they’ll shut down but the next ones along will be provoked to open fully, temporarily, and so on. And that’s the major way that many types of neurons communicate with each other, as well as with muscle cells. By propagating a signal that could be conceived of as “digital” at the level of the individual neuron (it happens or it doesn’t) down the axon the the synaptic end, where it will trigger the release of neurotransmitters and neuromodulators.

And when Hodgkin and Huxley discovered all this, studying the squid giant axon, they made use of ion channel agonists - ESPECIALLY tetrodotoxin.

So one could say that TTX attacks one of the most fundamental aspects of our nervous system, but also taught us a lot about how our nervous system works.

I vaguely recall some cones can actually shoot poison darts. Myth?

no myth, fact.

Cones have a modified radula (tongue, essentially) that forms a bag of harpoon-like “teeth” which are hollow and attached to a venom sack.

once they find a prey item, they launch a “tooth” through their proboscis into the target, and venom is pumped in.

a recent paper on the subject:

http://www.biolbull.org/cgi/content/full/207/2/77

using conotoxins:

http://news.bbc.co.uk/1/hi/sci/tech/4846504.stm

video clips showing piscivorous cone capturing goby:

http://www.oceanfootage.com/stockfo[…]e/Cone_Shell

it’s the idea that major transitions in evolution were accomplished more by changes in the timing and pattern of gene expression than by significant changes in the genes themselves.

It would seem logical if anatomical changes were found to be mostly from changes in regulatory sequences, while biochemical changes would be mostly from expressed genes.

These newts were not popular denizens of the lakes, because where they were found, the fish and frogs soon disappeared.

Would that be a case of no newts is good newts?

This is evolution in action, and the observed appearance of new properties, traced right down to single changes in proteins.

But their still garter snakes!!eleven!! (I tried to resist saying that, but it was futile.)

Henry

Two college students in Oregon ate rough skinned newts on a dare at a party. One of them died. Various sources claim that there is enough TTX to kill several humans in one animal.

Neuroscience for kids newsletter:

With a little library detective work, I discovered descriptions of three people who ate toxic newts. In two cases, men swallowed Oregon rough-skinned newts on dares. Soon after eating the newts, these men became weak, vomited and felt tingling in their lips or hands. One man died and the other recovered 24 hours later. The third case involved a two-year-old girl who bit off the tail of the family pet: an Oregon rough-skinned newt! The girl started to cry immediately after biting the newt’s tail and the girl’s mother was able to brush the tail out of the girl’s mouth. Both the girl and the newt survived.

PZ Myers Wrote:

The meat of the story, though, is that we have concrete measurements of specific molecular changes that are responses to an evolutionary arms race, and we’re seeing these differences emerge in different populations of a single species. This is evolution in action, and the observed appearance of new properties, traced right down to single changes in proteins.

PZ,

Is there any estimate on when these snakes on the West Coast began encountering these newts?

This is really cool!

I demand to see the data. And i want samples of the snakes!

Nice article, PZ.

Incidentally, it might be worth mentioning that the I1556L substitution is unlikely to participate in TTX resistance, because I and L are so similar. Of course, unlikely things have happened before, so I could be wrong.

Particularly interesting is the D1568N which, if I correctly recall my single-letter codes, is aspartate - asparigine.

Harold, thanks for the summary of the operation and physiological significance of voltage-gated sodium channels. I must confess, I had forgotten a good deal of that.

Meanwhile

Henry J Wrote:

Would that be a case of no newts is good newts?

I’m sorry, Henry, but I’m going to have to shoot you now.

Has anyone studied the evolving habits of the snakes at the Discotute? They are often found injecting their venom into the minds of young sentient bipedal creatures.

(Sorry…didn’t mean to denigrate the true lurking reptiles that rightfully hiss, slither and lurk.)

Hooray for this article, and the research that it describes!

I have been accused of asserting that the “neo-darwinian synthesis” is dead (and have, indeed, asserted something like that on various occasions). But this article is a very clear example of exactly the kind of microevolutionary change that the synthesis was formulated to explain: apparently random mutations in single loci, resulting in significant changes in reproductive success. Just what Drs. Fisher, Haldane, Wright and Dobzhansky ordered.

That said, I can’t help pointing out that the research described here, beautiful as it is, doesn’t really say anything about the diversification of these species of snakes. As Darwin himself pointed out in chapter 8 of the Origin of Species,

“On the theory of natural selection the case is especially important, inasmuch as the sterility of hybrids could not possibly be of any advantage to them, and therefore could not have been acquired by the continued preservation of successive profitable degrees of sterility.” (http://darwin-online.org.uk/content[…];pageseq=263)

In the cladistic divergence of discrete populations into multiple independently evolving lines it is precisely the “evo-devo” mechanisms that P.Z. most often describes in his always brilliant posts that underly much of macroevolutionary change.

Allen McNeil -

That said, I can’t help pointing out that the research described here, beautiful as it is, doesn’t really say anything about the diversification of these species of snakes.

Although this certainly could be a first step toward more extensive diversification of the snakes.

There are probably physical barriers to snake travel, and the allele conferring TTX resistance is probably concentrated in the local population.

Far traveling snakes from other populations who migrate to the area may be less likely to have the allele, and thus, at a significant relative disadvantage in the newt-enriched area.

In addition, the allele may create some sort of a biochemical “trade-off”; snakes with it may be a tiny bit slower or some such thing. Snakes with the allele might thus be at a slight disadvantage in other areas, and if not, there would still be no reason for the allele to be selected for, in the absence of a large TTX-secreting prey population.

So this allele could, in fact, be a trigger for relative isolation of the population that has it.

Other unique features might arise within the TTX-resistant local population, if there was an absence of in-migration of genetic material from outside garter snakes.

Eventually, TTX-resistant snakes might show a human-recognizable morphologic difference, or the population might develop a chromosomal barrier to reproduction with outside populations. The latter would be accepted by many as “speciation”.

Of course, what I have done above is nothing more than state a rational, testable hypothesis as to how a biochemical change might be the trigger for eventual morphologic or major genetic diversification of the population that contains it, but it is rational and testable.

So although I am not in full disagreement with your post, it strikes me that you may imply too sharp a distinction between various types of genetic diversification events.

Nigel D said:

I’m sorry, Henry, but I’m going to have to shoot you now.

Not sufficient punishment. We need to make him eat a newt.

“Sire, we caught this peasant poaching on the royal preserve!”

“Boil him in oil for three minutes!”

White Rabbit (Greg Goebel) http://www.vectorsite.net/tadarwin.html

Harold, I agree.

Allen, I think you are leaping to conclusions that are not justified.

While changes in genes that regulate devlopment tend to produce more obvious morphological changes, there is no reason to suppose that this change in several snake populations cannot be the start of a speciation event by a cladogenetic process.

If, as Harold points out, the snake populations are isolated, the populations that can eat the toxic newts could very well diverge from their relatives and form a new species. However, evolutionary theory predicts that such gradual change is likely to take many generations (perhaps several thousand or tens of thousands) unless it is forced by a strong selection pressure.

Besides, at what point would we recognise that the various populations of snakes actually are distinct species? It makes no difference to the snakes whether they are a variety, a subspecies, a species or even a distinct genus.

OK, got my details straight(er) … cones synthesize their own class of toxins, reffered to as (duh) “conotoxins”:

www.itg.be/itg/DistanceLearning/LectureNotesVandenEndenE/Teksten/sylabus/46_Marine_biotoxins.doc

There are apparently about 100 different variations on conotoxins, falling into roughly six categories of effect, one category being potassium channel inhibitors.

White Rabbit (Greg Goebel) http://www.vectorsite.net

… and (duh again) sodium-channel inhibitors.

Nigel D said:

Besides, at what point would we recognize that the various populations of snakes actually are distinct species? It makes no difference to the snakes whether they are a variety, a subspecies, a species or even a distinct genus.

That is a problem, especially since garter snake species readily hybridize with each other. My old evolution professor spoke of the headaches garter snake hobbyists caused for herpetologists by producing so many hybrid varieties, and then insisting on having the taxonomy changed accordingly, even though the species the hobbyists hybridized together only hybridize in captivity.

iml8 said:

OK, got my details straight(er) … cones synthesize their own class of toxins, reffered to as (duh) “conotoxins”:

www.itg.be/itg/DistanceLearning/LectureNotesVandenEndenE/Teksten/sylabus/46_Marine_biotoxins.doc

There are apparently about 100 different variations on conotoxins, falling into roughly six categories of effect, one category being potassium channel inhibitors.

White Rabbit (Greg Goebel) http://www.vectorsite.net

And don’t forget: a live cone shell was a murder weapon in an episode of Hawaii Five O.

Stanton said:

And don’t forget: a live cone shell was a murder weapon in an episode of Hawaii Five O.

Yeah, I remember that episode, and I rarely watched the series. “Book ‘em, Danno!”

White Rabbit (Greg Goebel) http://www.vectorsite.net/tadarwin.html

I’m sorry, Henry, but I’m going to have to shoot you now.

“What a senseless waste of human life.”

Allen, I think you are leaping to conclusions that are not justified.

that IS Allen’s forte.

Stanton said:

That is a problem, especially since garter snake species readily hybridize with each other. My old evolution professor spoke of the headaches garter snake hobbyists caused for herpetologists by producing so many hybrid varieties, and then insisting on having the taxonomy changed accordingly, even though the species the hobbyists hybridized together only hybridize in captivity.

Hmm, that’s a hobby I haven’t heard of before. I do have the impression that garter snakes are genetically pretty compatible but have a wild variation in color schemes – Wikipedia seems to bear that impression out – so I would bet the game is to come up with new “designer” garter snakes.

No doubt just to annoy taxonomists. Y’know, that would be an interesting subject for a paper, tracking the evolution of coloration variations in garter snakes. Tracing the evolutionary tree of conotoxins might be interesting too.

Of course, that said, Google before posting … ah, it’s been done, there’s a monograph available on the subject at Amazon.com:

http://www.amazon.com/Garter-Snakes[…]p/0806128208

– AND a dissertation from my alma mater, Beaver State University:

http://ir.library.oregonstate.edu/d[…]ertation.pdf

And there’s considerable reference to the evolution of conotoxins, it seems to be a hot topic:

http://www.pnas.org/cgi/content/full/96/12/6820

I’m hanging around with biologists too much, I’m getting into the habit of citing. Of course it would be nice if I could honestly understand more than half of these articles myself.

White Rabbit (Greg Goebel) http://www.vectorsite.net/tadarwin.html

For my own information, I looked up garter snakes in Wiki and found this:

Garters were long thought to be nonvenomous, but recent discoveries have revealed that they do in fact produce a mild neurotoxic venom. Garter snakes are nevertheless harmless due to the very low amounts of venom they produce, which is comparatively mild, and the fact that they lack an effective means of delivering it. They do have enlarged teeth in the back of their mouth, but unlike many rear-fanged colubrid snakes, garter snakes do not have a groove running down the length of the teeth that would allow it to inject venom into its prey. It is rather spread into wounds, through a chewing action. The properties of the venom are not well known, but it appears to contain 3FTx, commonly known as three-finger toxin, which is a neurotoxin commonly found in the venoms of colubrids and elapids. A bite may result in mild swelling and an itching sensation. There are no known cases of serious injury and extremely few with symptoms of envenomation.

I don’t know anything about toxicology. Is this venom similar enough to TTX to inform research on this developing resistance in garters?

dpr

DPR, Wikipaedia has this paper listed as a source under Toxicofera:

http://www.venomdoc.com/downloads/2[…]x_phylog.pdf

3FTxs are peptide toxins. TTX is a small molecule (even though part of its structure seems to scream out “arginine” as a biosynthetic precursor).

Nigel D said:

DPR, Wikipaedia has this paper listed as a source under Toxicofera:

http://www.venomdoc.com/downloads/2[…]x_phylog.pdf

3FTxs are peptide toxins. TTX is a small molecule (even though part of its structure seems to scream out “arginine” as a biosynthetic precursor).

Thanks Nigel!

Nigel D said:

DPR, Wikipaedia has this paper listed as a source under Toxicofera:

http://www.venomdoc.com/downloads/2[…]x_phylog.pdf

3FTxs are peptide toxins. TTX is a small molecule (even though part of its structure seems to scream out “arginine” as a biosynthetic precursor).

I thought so also.

Whatever happened to Nigel D?

About this Entry

This page contains a single entry by PZ Myers published on June 30, 2008 12:00 PM.

Analysis of Lousiana “Academic Freedom” bill was the previous entry in this blog.

The Crews Favorite Science and Evolution Books for Kids! is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.

Categories

Archives

Author Archives

Powered by Movable Type 4.381

Site Meter