Is it possible to engineer a creature capable of Lamarckian evolution?

Factual Questions
1

As I understand it, Lamarck’s theory of evolution basically says that an animal will adapt directly to fit its surroundings - the giraffe’s neck became long because giraffes were constantly stretching for high leaves, and their genetic coding changed as the neck became longer, allowing the modified genes for long necks were thus passed along to offspring.

We now know that it doesn’t work that way- that animals can only modify their behaviors to adapt to their surroundings, and that evolution only happens because the short-necked giraffes did not reproduce as successfully as the long-necked ones.

Since we know that 90% of our DNA doesn’t actually do anything, would it be possible to engineer an animal with extra genetic coding that the animal’s physical characteristics could be modified by? Say, take the basic sequences that make a horse and add material that allows the horse to grow a longer neck if necessary, or a shorter one?

I’m putting this in GQ because I expect the factual answer is “no”- I can’t imagine how the extra code would be “triggered”.

I’m not really concerned about whether the animal would be able to produce viable offspring. Also, don’t worry too much about the limits of available technology - if we will probably be able to build a supercomputer capable of the necessary sequencing, that’s good enough.

2

“Since we know that 90% of our DNA doesn’t actually do anything,”

Not knowing yet exactly what a sequence does is not the same as knowing it does nothing.

3

[QUOTE=DrFidelius]
“Since we know that 90% of our DNA doesn’t actually do anything,”

Not knowing yet exactly what a sequence does is not the same as knowing it does nothing.
[/QUOTE]

Fair enough.

Let’s assume that at least 50% of our code is either redundant or dormant, then.

4

Okay. It was just a quibble.

Personally, I think that a mechanism to “write back” instructions from the individual phenotype to the genome would be practically impossible. It is not enough to say “Create more muscle mass” or “self-closing nostrils are needed.”
Development of any morphological feature is the result of the balance between any number of different growth factors, and it may be a failure of my imagination but I cannot see how to set up a mechanism to translate an altered end result back to the directions.

5

[QUOTE=DrFidelius]
Okay. It was just a quibble.

Personally, I think that a mechanism to “write back” instructions from the individual phenotype to the genome would be practically impossible. It is not enough to say “Create more muscle mass” or “self-closing nostrils are needed.”
Development of any morphological feature is the result of the balance between any number of different growth factors, and it may be a failure of my imagination but I cannot see how to set up a mechanism to translate an altered end result back to the directions.
[/QUOTE]

I had not thought of that.

Let’s suppose we have a highly developed brain (read “human”) capable of conscious control of hormone production.

Is that theoretically feasible in itself?

The Google ads are for something called Little Giraffe Boutique…

6

I don’t know if it’s possible.

I had a thought though… Assuming you could figure out which genes stimulate neck growth, could you put a gene or genes in a tree which would make a chemical that would trigger the promoters for the genes that stimulate neck growth in the horse?
Do horses eat leaves?

7

I’m not sure that it could be done using our current biology: What you’re talking about would require a feedback mechanism to allow body state to affect the genes, directly. And I’m skeptical enough of our ability to play with the interaction between genes and their expression in phenotype to think that could be done.
Now, if you want to get really SF, however, consider the concept of the Von Neumann machine. Basically, it’s a machine that can mimic life: It can, by itself, out of local resources, replicate itself. I can imagine that if we can engineer a Von Neumann machine, it could be designed to take on aspects of Lamarckian evolution: i.e. responding to changed environments, by changing how it reproduces itself.

On preview: DrFidelius has already addressed my major point, better than I did.

8

The key problem is not that organisms cannot modify their phenotype in response to their environment. They clearly do. The problem is that the phenotypic changes aren’t written back into the genome, and given the way genetics works, there doesn’t seem to be any way to do so. A giraffe has a long neck, but how do you translate the need for a long neck back into the giraffe’s genes? How do the cells in the stretched neck communicate back to the germ line cells? Cells do communicate with each other through hormones, but how do the germ line cells know how to modify their DNA in response to those hormones, such that the changes are inheritable?

9

[QUOTE=Lemur866]
The key problem is not that organisms cannot modify their phenotype in response to their environment. They clearly do. The problem is that the phenotypic changes aren’t written back into the genome, and given the way genetics works, there doesn’t seem to be any way to do so. A giraffe has a long neck, but how do you translate the need for a long neck back into the giraffe’s genes? How do the cells in the stretched neck communicate back to the germ line cells? Cells do communicate with each other through hormones, but how do the germ line cells know how to modify their DNA in response to those hormones, such that the changes are inheritable?
[/QUOTE]

In the OP I noted that I’m not worried about whether the creature could produce viable offspring- don’t worry about whether it can pass along its genome.

What organisms modify their phenotypes in response to the environment?

10

[QUOTE=Really Not All That Bright]

Since we know that 90% of our DNA doesn’t actually do anything, would it be possible to engineer an animal with extra genetic coding that the animal’s physical characteristics could be modified by? Say, take the basic sequences that make a horse and add material that allows the horse to grow a longer neck if necessary, or a shorter one?

[/QUOTE]

As I understand it, 90% of our DNA (if that’s the number) is non-coding, which I gather means it does not contain the genes that produce all the various proteins used by our bodies. However, non-coding DNA does produce chemicals (I recall one called micro RNA or mRNA) that seem to play a large part in turning genes on and off, and how gene-encoded proteins do their jobs.

Gene expression can also be altered by markers (I believe called methyl groups) that are themselves subject to change by environmental conditions, and, once changed, are inherited along with the genes they are attached to. As I recall it, a single gene can code for a variety of proteins, or be turned off entirely, depending on where various markers are placed.

So it could be that something like Lamarckian evolution does exist, at a chemical level at least, in the sense that alterations in gene expression that are caused by environmental conditions can be carried forward to a creature’s offspring.

I’m not sure if this is an example of the same thing, but I have read that releasing a domestic pig into the wild results in rapid physical changes in the individual pig (like skin color, hair texture, and tusk size) that distinguish it as a feral rather than a domestic pig within a short time (weeks or months, as I recall). If that’s true (I just read it somewhere), then perhaps this is a case of physical environment (adrenaline levels?) altering the pig’s gene expression adaptively.

You could still say that this adaptive gene expression is built into the genome, and therefore not the same as genetic evolution. Still, if it’s all true, it seems to make the whole topic of genetics even more complex and variable than what they told us in high school. If we eventually learned enough about how it works, perhaps we could manipulate gene expression more directly than we already do by selective breeding.

BTW: I think I read the stuff about methyl group markers and micro DNA in Science News over the last few years. The stuff about feral pigs was in some state wildlife document I read a while ago.

11

[QUOTE=Really Not All That Bright]
In the OP I noted that I’m not worried about whether the creature could produce viable offspring- don’t worry about whether it can pass along its genome.

What organisms modify their phenotypes in response to the environment?
[/QUOTE]

But Lamarckian evolution would effect the genome in some way - of course it doesn’t consider the genome, since that wasn’t discovered yet.

Alligators (or is it crocs) change their sex in the egg based on the temperature. To change the DNA of an existing animal, perhaps some kind of virus that inserted the modified DNA strands in the right places would work. You do say engineer, so I’m not feeling limited to what might actually evolve. Where this new coding would come from I can’t figure out yet.

As for passing it on, it would be simple enough to use the same method to change the sperm and eggs, or to modify their production. The problem would be how to ensure that two animals with the same modification would have compatible coding.

12

But then you’re not talking about Lamarckian evolution. The essence of Lamarckian evolution is that phenotypic changes to an organism have to be heritable. The giraffe stretches his neck his whole life, and therefore his offspring are born with longer necks. The blacksmith pounds on the anvil his whole life and develops massive arm muscles, and his children are born with larger arm muscles because of it. A peasant works in the sun his whole life and develops a deep tan, and his children are born darker colored because of it.

As for phenotypes changing based on environment, the blacksmith’s arms and the peasant’s tan are examples, and there are millions of others. But as far as we can tell, these phenotypic changes have no way of being passed on to the changed organism’s offspring, because those phenotypic changes don’t alter the DNA in the organism’s gametes, in fact, the phenotypic changes don’t even alter the DNA in the directly affected cells and tissues and organs. The potential for phenotypic plasticity is already present.

13

[QUOTE=Lemur866]
But then you’re not talking about Lamarckian evolution. The essence of Lamarckian evolution is that phenotypic changes to an organism have to be heritable. The giraffe stretches his neck his whole life, and therefore his offspring are born with longer necks. The blacksmith pounds on the anvil his whole life and develops massive arm muscles, and his children are born with larger arm muscles because of it. A peasant works in the sun his whole life and develops a deep tan, and his children are born darker colored because of it.

[/QUOTE]

Rather than Lamarckian evolution, then, call it direct adaptivity or something. I just excluded the possibility of offspring to make things easier (apart from anything else, you’d have to engineer a breeding population of these creatures to produce a viable population, anway).

14

[QUOTE=Lemur866]

As for phenotypes changing based on environment, the blacksmith’s arms and the peasant’s tan are examples, and there are millions of others. But as far as we can tell, these phenotypic changes have no way of being passed on to the changed organism’s offspring, because those phenotypic changes don’t alter the DNA in the organism’s gametes, in fact, the phenotypic changes don’t even alter the DNA in the directly affected cells and tissues and organs. The potential for phenotypic plasticity is already present.
[/QUOTE]

In my previous post, I might be overstating the potential for inheritance of methyl group states, but it seems to be possible. I think the following quote from Science News is where I got the idea:

I can’t find anything else in the article about epigenetic marks being inherited. BTW: Here’s the cite (sorry, subscriber link):

Registered subscribers only: Nurture Takes the Spotlight (6/24/2006)
What a person eats, what chemicals he or she is exposed to, and other features of a person’s environment chemically modify chromosomes, thereby changing how genes are ultimately expressed.

15

Total sci-fi speculation: Over a long span of time, I imagine some super-cool alien race could “evolve” such a creature, by engineering its environment such that there are certain laws governing how the environment varies over time. More complicated versions of things like “After a thousand years of short shruberies, begin growing tall ones instead.” Then creatures could evolve which are prepared to react to these kinds of changes by switching on or off various genes in its genome. But this wouldn’t be what the OP is asking after, because if you took such a designed creature and put it on the earth (where there are no such laws about how environments change over time) its design features would not work as intended.

-FrL-

*That’s not to say environmental changes happen completely haphazardly, just that you can’t be sure what the environment in a particular place will be like in ten thousand years based on measurements of that place’s environment right now. I think…

16

[QUOTE=Lemur866]
The key problem is not that organisms cannot modify their phenotype in response to their environment. They clearly do. The problem is that the phenotypic changes aren’t written back into the genome, and given the way genetics works, there doesn’t seem to be any way to do so. A giraffe has a long neck, but how do you translate the need for a long neck back into the giraffe’s genes?
[/QUOTE]

I see it as being even more difficult than that… somehow the organism’s physiology has to recognize the condition “a longer neck really would have worked better for me in this life” before it writes that feedback into the DNA. I cannot possibly see how this would work without a super-intelligent organism that can analyze its own internal state like this, and can modify its own internal state. Any organism this smart would have had learned how to do this with lab equipment eons before it could ever do it autonomously. (Which sorta describes us, actually).

17

[QUOTE=mwbrooks]
Gene expression can also be altered by markers (I believe called methyl groups) that are themselves subject to change by environmental conditions, and, once changed, are inherited along with the genes they are attached to. As I recall it, a single gene can code for a variety of proteins, or be turned off entirely, depending on where various markers are placed..
[/QUOTE]

This topic has been discussed by the SDMB in far greater depth in the past, but IMHO, I’d give unqualified support to the statement that a variety of Lamarckian evolution does exist. It isn’t nearly as far reaching as Lamarck might have envisioned it (you can’t grow a big-nose into an elephant trunk), but patterns of genetic expression governing metabolic strategies, etc. can be passed down between generations without alterations of the genome.

This is now a well established and demonstrated result in modern biological sciences.

The key phrase is epigenetics, and the Wikipedia article is a decent primer, although a number of other excellent reviews are available in the lay press and scientific literature.

So, in response to the OP, yes, and it has already happened.

18

It was discovered a decade or so ago that under the right conditions, E. coli will induce systems that will put it into a hypermutable state. So when it’s stressed, it can essentially speed up its own evolution. This was controversial for a while because it was seen as smacking of Lamarkism, but now that it’s better understood, it’s pretty well widely accepted. I interviewed with one of the discoverers of this process when I was applying for grad school.

19

[QUOTE=Smeghead]
It was discovered a decade or so ago that under the right conditions, E. coli will induce systems that will put it into a hypermutable state. So when it’s stressed, it can essentially speed up its own evolution. This was controversial for a while because it was seen as smacking of Lamarkism, but now that it’s better understood, it’s pretty well widely accepted. I interviewed with one of the discoverers of this process when I was applying for grad school.
[/QUOTE]

It still requires individual cells to die, right?

20

Well, yes. It activates a DNA polymerase with an unusually high error rate and deactivates or downgrades some of the repair mechanisms. Thus, every time it replicates its genome and creates two daughter cells, they have a higher rate of mutations. As with any mutations, some are favorable, some are unfavorable, and many are neutral. The odds of generating a cell with the mutation(s) it needs to survive go up, though, overall.