Questions about recurring mutations and phenotypic convergence

[Buckaroo Banzai]

I think I know one thing, but I'm not so sure.
Convergent pheonotypes don't have proportional genetic convergence, right?
I think that's maybe possible at some extent, in related species, the more the development paths are similar, of course; I'd not expect a genetic modified squid to have shorter tentacles if we insert the gene that makes the legs shorter in some sorts of dog breeds, like basset hounds.

But in closely related species, such as wolves and foxes, it could work, I think... so if we make a genetic modified wolf or even in a GM fox, they would have shorter legs like a basset...
But, does this genetic convergence happens naturally? I don't think so, but I'm not sure... the closest thing I can imagine that happens genetically in two relatively distant related organisms (but yet with similar developmental pathways) when they converge phenotypically is to them have different mutations, even in different genes, but causing similar effects, maybe even identical effects in the protein level, if this convergence is due to supression of a gene shared from a common ancestor.


But, in the other hand, Chris Colby's "introduction to evolutionary biology" mentions recurrent mutations.... what does it means, exactly? That the a specific genetic mutation can occur many times and eventually be fixed, or is he talking about different mutations at the genetic level, but that cause the same/similar phenotype?
If is the first one (which I think that isn't), how frequent would it be? Is there something like a "mutational bias" - but not with respect to fitness; or would it be purely random?

 

[QuarkHead]

I think I know one thing, but I'm not so sure.
Convergent pheonotypes don't have proportional genetic convergence, right?

Absolutely right. Similar phenotypes can have a different genetic bases (and vice versa)

But, does this genetic convergence happens naturally? I don't think so, but I'm not sure...

Absolutely not! It is a well-founded principle of evolutionay genetics that the genetic difference (= genetic distance) between two species is a measure of how close they are related in an evolutionary sense. The greater the genetic distance, the longer ago they diverged. Divergence is a one way street. There are some very restricted situations where genotypic convergence can occur. Ah.. do I need to explain the difference between "genetic" and "genotypic"? Hohum.
Genetic refers to a) the process of inheritence and b) the properties of a single gene (I don't like that, but it's embedded). So, the sequence information of a gene in populations will always diverge over evolutionary time. Genotypic refers to the genetic make of an organism. Note that for a diploid organism, this will include information about both alleles.
Here's an example: you can have blood group A (your phenotype) but your genotype could be AA or AO (not a good example, I know because O is the null product of H. Not important here, though).
OK. I referred above to some restricted cases where genotypic convergence can occur. Here's one. Take two previously non-interbreeding populations of the same species. Let's say they are to some extent genotypically divergent. Now let them interbreed. Over time the offspring population is genotypically convergent relative to the founder populations (it actually happens much faster than you might expect, due to something called the Hardy-Weinberg equilibrium. It sounds scary, but it's nothing more that the 2nd order polynomial based on relative allele frequencies).

the closest thing I can imagine that happens genetically in two relatively distant related organisms (but yet with similar developmental pathways) when they converge phenotypically is to them have different mutations, even in different genes, but causing similar effects..

different genes? Well it can happen. But due to genetic heterogeity at the level of phenotype, yes, different mutations in the same gene frequently give rise to similar phenotypes.

maybe even identical effects in the protein level...

No, never.

recurrent mutations.... what does it means, exactly?

For a mutation to be recurrant means the exact same mutation (A to G or whatever) occurs at the exact same place in the exact same gene. Although I don't agree with Colby's estimate of mutation rate, it is still sufficiently low that we need not expect recurrant mutations to occur.

That the a specific genetic mutation can occur many times ..

Oh right. Let's say

and eventually be fixed....

fixation is a tricky one. Colby is good on that, but it eventually comes down to taking sides. It's all a matter of modelling - can mutations without positive selection potential ever become fixed, even assuming the most extreme statistic for drift?

...would it be purely random?

Mutations are random, fixation in the absence of selection is random! What do you think?

 

[Maddad]

Danniel
Colby was talking about deletion mutations, which are a special type. They happen because a DNA or RNA sequence is left out during replication. It can happen either at the end of a strink or in the middle. By "recurrent mutation" he is saying that the same coding gets left out in the mutation process. We've seen this process demonstrated dramatically in the 1960's when the 4,500 base pair RNA HIV retrovirus was induced to repeatedly duplicate in a test tube containing free base pair and the replicase protein. In 74 generations the RNA shed by deletion mutation all but the 220 base pair needed to encode for more replication. The experiment was repeated by others and showed the same results; the RNA consistently shed by repeated deletion recurrent mutation the same base pair sequences, winding up with the same biological end product.

 

[Hercules Rockefeller]

Colby was talking about deletion mutations, which are a special type.

No he wasn’t. Where did you get that idea? I see no instances of references to deletion mutations, or implications thereof. In the three instances in that large amount information where he mentioned “recurrent mutations”, he was clearly referring to any mutation that causes a phenotype, deletion or otherwise.<P>

 

[Maddad]

Unless they mate with another carrier, the allele may simply continue to be passed on. Deleterious alleles also remain in populations at a low frequency due to a balance between recurrent mutation and selection.

Where did I get the idea? I read the link Danniel provided. All references to recurrent mutation were in the context of deletion mutations. The one I quoted above was just the first.

 

[Hercules Rockefeller]

All references to recurrent mutation were in the context of deletion mutations.

No, they aren't. I've read it too. There is no indication that the phrase "recurrent mutation" is rerferring specifically to deletions. There are other types of mutations that cause phenotypes and that contribute to evolutionary change, you know? Perhaps you could try to explain your reasoning? Why couldn't his description of recurring mutations be referring to point mutations?

The one I quoted above was just the first.

In no way does that quotation specifically implicate deletions. You didn't mis-read the word "deleterious" for "deletion", did you?<P>

 

[QuarkHead]

Danniel
Colby was talking about deletion mutations, which are a special type.

Well he wasn't, but no matter. You're right, they are a special type ...in that many geneticists, me included would hesitate to call them mutations. Deletions and insertions are just that and, although we have a ball-park for the rate of point mutations, there is none, as far as I know, for the rate of deletions, insertion, inversion etc. We know there are hot-spots (as there are for point mutations in prokaryotes) but these are to do with what follows.

They happen because a DNA or RNA sequence is left out during replication.

No!! They happen for any number of reasons. The ones we know most about are reciprocal exchange mis-match and inappropriate gene conversion.

It can happen either at the end of a strink or in the middle.

What's a "strink?

...the 4,500 base pair RNA HIV retrovirus was induced to repeatedly duplicate in a test tube ...the RNA shed by deletion mutation all but the 220 base pair needed to encode for more replication.

You really think this a good model for evolutionary genetics? I can think of any number of reasons why deletions occur in vitro. If I may be a bit rude - before replying to a post here, especially if it is the originating post, read the question twice before replying. I find that quite good exam technique too!

 

[Buckaroo Banzai]

Absolutely not! It is a well-founded principle of evolutionay genetics that the genetic difference (= genetic distance) between two species is a measure of how close they are related in an evolutionary sense. The greater the genetic distance, the longer ago they diverged. Divergence is a one way street. There are some very restricted situations where genotypic convergence can occur.

But despite of the phylogenetic distance, various genetic similarities can be maintained by selection, and if there are similar genes involved in the development (and there are), a similar mutation would probably lead to a similar phenotype...
as far as I know, genetic distances are evaluated by somewhat "random" gene sequences, not by specific parts... well, may be specific sequences, but not in the sense I'm talking... I'm thinking about sequences which are likely to have been highly conserved, such genes linked to development of the legs on dogs and cats... where, maybe, their lenght, for example, maybe could be similarly controlled, and then we find that a similar gene, or a similar mutation in the same gene, is responsible both for the long legs of the cheetah and of the greyhounds. Or something in this sense...

No, never.

... but yer there are nearly identical proteins that evolved twice, at least one case is known:

Convergent evolution of antifreeze glycoproteins in Antarctic notothenioid fish and Arctic cod
(I haven't read the paper carefully yet, so I may be saying just bovid feces right now... all I know is that the genes aren't equal, but the protein is quite similar, or equal, and works equally to the same function)

Mutations are random, fixation in the absence of selection is random! What do you think?

Well..... I was thinking if couldn't exist something like a mutational bias, somewhat non-random mutations, but the non-randomness would not be in respect to fitness (despite of generally speaking of non-random mutations in the sense of adaptive mutations). Something like some gene sequences just having a tendence to mutate more than others, maybe even to specific mutations. But I have no idea of how could it be, if could.
But a real tendence to mutate, not that mutations there would just be neutral, then allowed to mutate randomly by absence of selection... then in two "neutral regions" we would find more mutations than in the other, more than what would be expected if mutations where completely random. And, not only that, but maybe some specific mutations could happen more likely than others....


Other paper I haven't read, but seems interesting in the subjetc of genetic convergence:

Exceptional Convergent Evolution in a Virus

Abstract:
Replicate lineages of the bacteriophage C#JX 174 adapted to growth at high temperature on either of two hosts exhibited high rates of identical, independent substitutions. Typically, a dozen or more substitutions accumulated in the 5.4kilobase genome during propagation. Across the entire data set of nine lineages, 119 independent substitutions occurred at 68 nucleotide sites. Over half of these substitutions, accounting for one third of the sites, were identical with substitutions in other lineages. Some convergent substitutions were specific to the host used for phage propagation, but others occurred across both hosts. Continued adaptation of an evolved phage at high temperature, but on the other host, led to additional changes that included reversions of previous substitutions. Phylogenetic reconstruction using the complete genome sequence not only failed to recover the correct evolutionary history because of these convergent changes, but the true history was rejected as being a significantly inferior fit to the data. Replicate lineage subjected to similar environmental challenges showed similar rates of substitution and similar rates of fitness improvement across corresponding times of adaptation. Substitution rates and fitness improvements were higher during the initial period of adaptation than during a later period, except when the host was changed.

http://www.genetics.org/cgi/reprint/147/4/1497.pdf

Some time ago I posted here a question of mutation rates, time to go back and read the answers and references again...

 

[Maddad]

Colby was talking about deletion mutations, which are a special type. . . They happen because a DNA or RNA sequence is left out during replication.

No!! They happen for any number of reasons. The ones we know most about are reciprocal exchange mis-match and inappropriate gene conversion.

You are confusing why deletion mutations happen with what they are. A deletion mutation, by definition, is a genetic sequence left out during replication. An insertion mutation inserts coding into the middle of the sequence (usually - it can also insert it at the ends as well.) If the deletion or insertion does not involve a whole codon, then you get a fram shift mutation, which for obvious reasons is usually immediately fatal to the replication process.

The post of yours above wasn't technically wrong, really. You were describing a simultaneous insertion and deletion action which is common in the exchange of entire alleles. The swapping though of these genes is fundamentally a combination of insertion and deletion mutations.

 

[QuarkHead]

You are confusing why deletion mutations happen with what they are.

Confusing? No just explaining why I disagree with this.....

A deletion mutation, by definition, is a genetic sequence left out during replication.

and this....

Danniel
They happen because a DNA or RNA sequence is left out during replication. .

If the deletion or insertion does not involve a whole codon, then you get a fram shift mutation, which for obvious reasons is usually immediately fatal to the replication process.

Huh? What has frame shift got to do with replication? Cells will happily relicate complete garbage. If you mean it's usually fatal to transcrption or translation you are still wrong. These only fail if the frame shift introduces a stop codon. It needn't do so.

The post of yours above wasn't technically wrong, really.

So kind!

You were describing a simultaneous insertion and deletion action which is common in the exchange of entire alleles. The swapping though of these genes is fundamentally a combination of insertion and deletion mutations.

I have no idea what this means. Gene conversion is a one way process by definition - that's why it's called non-reciprocal exchange. It almost never involves entire alleles. If it goes wrong (and there are many examples of this in Man) you can end up either with an insertion or a deletion.
Reciprocal exchange (recombination) doesn't involve simultaneous insertion and deletion . It involves a two way exchange. Again what do you mean when you say it involves entire alleles? It involves entire chunks of chromosome carrying several alleles usually

 

[Hercules Rockefeller]

A deletion mutation, by definition, is a genetic sequence left out during replication. An insertion mutation inserts coding into the middle of the sequence (usually - it can also insert it at the ends as well.)

Yeah, so????????? This doesn’t explain why you think Colby was specifically referring to deletions when he discussed “recurrent mutations” contributing to the allele frequencies. His description could (and does) equally refer to point mutations etc.

If the deletion or insertion does not involve a whole codon, then you get a fram shift mutation, which for obvious reasons is usually immediately fatal to the replication process.

That’s just plain wrong. During transcription the RNA pol complex will proceed until it hits a polyA signal in the 3’UTR. During translation the ribosome complex will make the polypeptide until it hits a STOP codon. The resulting protein may be garbage due to a frameshift, but the frameshift is not “immediately fatal to the replication process”. Transcription and translation proceed oblivious to the frameshift.<P>

 

[QuarkHead]

Let's leave out the quibbles with Maddad and get back to your question

But despite of the phylogenetic distance, various genetic similarities can be maintained by selection, and if there are similar genes involved in the development (and there are), a similar mutation would probably lead to a similar phenotype...

Ho! Let me stop you right there! Ignoring your slight misuse of the word "phylogenetic"....remember that selection works on the phenotype. This may or may not have a consequence for the genotype. For example, whereas a deleterious mutation would quickly be removed from the gene pool, mutations (= genetic differences) leading to similar phenotypes may be maintained by selection (among other mechansims). If you re-read my post you will see that I by no means discounted the possibility that diverged genotypes could result in converged phenotypes. We know of many examples of this

... genetic distances are evaluated by somewhat "random" gene sequences, not by specific parts... well, may be specific sequences, but not in the sense I'm talking... I'm thinking about sequences which are likely to have been highly conserved

Genetic distance can be established using "non-random" sequences as well. One of the first really good phylogenies was constructed using protein sequence data from cytochrome c, an extremely essential protein for life on this planet!

...such genes linked to development of the legs on dogs and cats...

I strongly advise you (and me for that matter!) not to worry too much about such things. Not enough is known to make sensible comments

... but yet there are nearly identical proteins that evolved twice, at least one case is known:

Again, I never discounted that possibility, at least I hope not! But "nearly identical proteins" need not have nearly identical genes. Of course there will be similarities, but the selection pressure would have to be immensly strong.

Something like some gene sequences just having a tendence to mutate more than others, maybe even to specific mutations.

Well yes - we know this is true in prokaryotes, we also know that, statistically, some base changes occur more frequently than others. I know of no example in mammals, but there may be.

 

[QuarkHead]

During transcription the RNA pol complex will proceed until it hits a polyA signal in the 3’UTR.

No - the polyA is added after transcription. The transcription stop signal is varied and not interesting (meaning I can't remember the details!!)

Transcription and translation proceed oblivious to the frameshift.

Yes, transcription may, but translation isn't always "oblivious". Maddad didin't realise it, but the probability of getting a UAG, UGA etc point mutation from a coding triplet. is ...er... well it happens often.

 

[Buckaroo Banzai]

Let's leave out the quibbles with Maddad and get back to your question Ho! Let me stop you right there! Ignoring your slight misuse of the word "phylogenetic"....remember that selection works on the phenotype. This may or may not have a consequence for the genotype. For example, whereas a deleterious mutation would quickly be removed from the gene pool, mutations (= genetic differences) leading to similar phenotypes may be maintained by selection (among other mechansims). If you re-read my post you will see that I by no means discounted the possibility that diverged genotypes could result in converged phenotypes. We know of many examples of this

But should exist a limit, based on probabilities or something, to the changes in the genes controlling homologous parts of the body to be different... for example, we wouldn't expect to find, looking at the genes, that the tails evolved convergently in multiple instances in the canids or even twice in the marsupials and in the placentals... it would be much more likely to have recognizable genes dealing with the development of the tail in vertebrates in general I guess, or at least in a few big branches, such as mammals and reptiles... and yet with the vast majority of the genes conserved...

Genetic distance can be established using "non-random" sequences as well. One of the first really good phylogenies was constructed using protein sequence data from cytochrome c, an extremely essential protein for life on this planet!
“
...such genes linked to development of the legs on dogs and cats...
”
I strongly advise you (and me for that matter!) not to worry too much about such things. Not enough is known to make sensible comments

mmm.... disappointing.... my whole point was to know if we could trace wrong phylogenies (by tracking only specific genes, disregarding a lot of other stuff) indicating false phylogenetic proximity due to some convergence ...in something that still is homologous, anyway... I would not expect a squid to be any genetic similarer to a dolphin or a shark than a octopus because of the squid's convergence to torpedo shape.

 

[Hercules Rockefeller]

No - the polyA is added after transcription. The transcription stop signal is varied and not interesting (meaning I can't remember the details!!)

No QuarkHead, I was not referring to the actual poly(A) tail added to a mRNA. That is added post-transcriptionally, as you say. I was referring to the DNA sequence that serves as the polyadenylation <B><I>signal</I></B> ie. the signal that tells the RNA pol complex to stop transcribing and for the relevant terminal transferase to add the adenines. The consensus poly(A) signal is AATAAA, although sometimes the signal is cryptic and cannot be identified. More often than not it is the poly(A) signal that is the transcription stop site.

Yes, transcription may, but translation isn't always "oblivious".

The translation machinery doesn’t care what the mRNA triplet sequence is; it keeps going until it hits a stop codon. Thus, it is oblivious to frameshifts.
<P>

 

[QuarkHead]

The translation machinery doesn’t care what the mRNA triplet sequence is; it keeps going until it hits a stop codon. Thus, it is oblivious to frameshifts.

Well of course. Frame-shifts might yield a garbage protein, but where they generate a stop (usually down the line, as it were) then they will be translated to a product which will almost always be degraded intracellularly.
So what has this to with evolutionary genetics? Tell you what. Let's talk direct to Danniel, the originator of this thread

 

[QuarkHead]

....tails evolved convergently in multiple instances in the canids or even twice in the marsupials and in the placentals... it would be much more likely to have recognizable genes dealing with the development of the tail in vertebrates in general I guess...

I can only geuss at what you are saying here. Let me say first, though, I know almost nothing about development in general. But - I do know that it is a popular fallacy that because something is there, then it must have been selected for. Example - look in the mirror, there, right beneath your nose, do you see a double fold? (maybe you have a beard!!). That fold is not "selected for", it is just a consequence of the way the embryo develops.
I would imagine (again - I'm no expert here) that tails are the default design for all vertebrates, and they are only "discarded" in the evolutionary sense when they either become a nuisance or can easily be dispensed with.

....if we could trace wrong phylogenies (by tracking only specific genes, disregarding a lot of other stuff) indicating false phylogenetic proximity due to some convergence ...

Ahhh! I'm starting to get suspicious here. Why do you want to trace wrong phylogenies? Have you a better one than the standard? Does it matter?
In one sense, no it doesn't - phylogenies are merely our way of collecting stamps and keeping the natural world in our sort of order. On the other hand, if you choose to re-classify our existing phyla and their relationships, say "all things with tails or flagella in this box, all things with light receptors in that one" then you would have a very very very hard time coming up with an evolutionary model.

 

[Maddad]

Huh? What has frame shift got to do with replication?

Frame shift stops replication.

If you mean it's usually fatal to transcrption or translation you are still wrong. These only fail if the frame shift introduces a stop codon.

You must have slept though your microbiology classes. Each base-pair triplet codes for an amino acid. A string of amino acids builds a protein. A frame shift changes each triplet, starting at the shifted triplet and continuing to the stop codon. Because the stop codon will be changed as well, it will not be located in the correct place in the sequence. It will be sooner or later, randomly placed. Even if the shift didn't mess with the stop codon, the replication will code for the wrong protein, meaning that it is utter garbage and will not support life.

Yeah, so????????? This doesn’t explain why you think Colby was specifically referring to deletions . . .

Because I read the article and that was the context he used.

If the deletion or insertion does not involve a whole codon, then you get a fram shift mutation, which for obvious reasons is usually immediately fatal to the replication process.

That’s just plain wrong.

That's just plain basic biology. Basic right biology, and I'm always right, remember?

The translation machinery doesn’t care what the mRNA triplet sequence is; it keeps going until it hits a stop codon. Thus, it is oblivious to frameshifts.

You partially right but also partially wrong. The frame shift means that what used to be a stop codon no longer is one. What wasn't a stop codon now is. Where will the new stop codon be? Where it doesn't belong. Transcription doesn't end when it's supposed to, so it is not oblivious to the frame shift.

It doesn't even matter though because the transcription will be building a junk protein anyway, which is also not be oblivious to the frame shift.

 

[qwerasdfzxcv]

I got a ?. Can personality passed on by genes? I've never heard anything bout that but it's a question i have. I don't mean as in that you're going to act a certain way becuase the way your parents are. Well can any these smart ass people answer my question? Thanks for ya time.

 

[QuarkHead]

Frame shift stops replication.

No it does not. The replication machinery simply doesn't know or care what is is doing. Consider this...no replication, no cell division, no meiosis, no gametes, no offspring, no frame-shifts passed through the germ line. Do you really think this is the case?

You must have slept though your microbiology classes.

Microbiology??

A frame shift changes each triplet, starting at the shifted triplet and continuing to the stop codon. Because the stop codon will be changed as well, it will not be located in the correct place in the sequence. It will be sooner or later,

sooner, later....or never!! That's the point.

the replication will code for the wrong protein,

Ah! Now I see your problem. Replication has nothing to do with coding for proteins, "right" or "wrong" ones. Replication merely copies DNA. The code as you so quaintly call it is broken only by the translation machinery...and only that.
Look - I don't really care whether you understand this or not. But in case anybody else is looking in let me say this. You have no idea what you are talking about.

The frame shift means that what used to be a stop codon no longer is one. What wasn't a stop codon now is. Where will the new stop codon be? Where it doesn't belong.

Please compute for us the probability of generating a transcrition or translation stop signal from any arbitrary pair of codons after a frame shift. Can't do it? Well let me tell you it is significantly <1