|
Author
|
Topic: John A. Davison: An Evolutionary Manifesto: A New Hypothesis For Organic Change
|
Nel
Member
Member # 614
|
posted 14. June 2003 21:11
Rex,
I didn't simply mean that multiple roads can lead you to the same phenotype. The references show that
1. The same genotype results in different phenotypes.
2. Changing the genotype does not cause the phenotype to change.
3. One can rewire the developmental circuitry and yet cause no change to the phenotype.
I also never said that the gene has absolutely nothing to do with phenotypic change, my post was meant to show scientific evidence that contradicts the notion that phenotypic change is gene-centric, gradualistic, etc.
Another example of this might be where related clones of E. Coli exhibited vastly different phenotypes . When they were exposed to an attractant, their chemotactic response time were extremely different. Obviously the reason for this is not genetic.
Spudich, JL and Koshland, DE,
Jr (1976) Non-genetic individuality: chance in the single cell Nature.
What this means is that this area is completely open to a design-theoretic view of evolution (or of phenotypic origination). We can even downplay the role of mutations to genes that provide a selective advantage, instead, in some examples like tissue specificity, we may be looking at a complex web of connections among thousands of signaling pathways which are needed for differentiation.
[ 14. June 2003, 21:32: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 14. June 2003 23:37
Nelson:
quote:
I didn't simply mean that multiple roads can lead you to the same phenotype. The references show that
1. The same genotype results in different phenotypes.
2. Changing the genotype does not cause the phenotype to change.
3. One can rewire the developmental circuitry and yet cause no change to the phenotype
The exact genotype results in different phenotypes? COuld you give some specific examples?
2./3. is not really relevant to the argument, in fact there are many known neutral substitutions for instance.
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 15. June 2003 12:13
quote:
Nelson:
No. The gene they've identified not only shows many genetic changes, but Nup96 causes death when its forced to interact with another (still unidentified) gene from the other species. So the amount of genes required for this type of isolation to occur is at least two. It is by far not the result of changes to a "single gene".
I truly do not understand your reasoning. By this token, there would be monogenic traits at all, since of course each gene can only act in the context of a genome (and many single gene traits express differentially in the context of different genomes). The fact is, Nup76 mutations can cause hybrid sterility by themselves (in the context of a Drosophila simulans or melanogatser genome), as demonstrated in the Presgraves paper. quote:
But the fact that this paper is just elaborating on the first of 200 chromosome regions that was discussed much earlier was what made your statement about timing kind of funny.
This is also unclear. First of all, the number of mapped hybrid sterility loci in melanogaster x simulans is 20, not 200. If you really want to nitpick, since they only screened 70% of the genome, that represents the minimal number, but of course that matters little, since each of those loci is, like Nup76, sufficient for hybrid sterility. Second, I am also unsure what you are saying that was discussed much earlier. If you are referring to the pseudoobscura Bogota x USA crosses, the estimate there is up to 30 loci. In other Drosophila species the estimate is larger (eg, >100 in simulans x mauritiana), but again as I mentioned before this is likely due to the fact that hybrid sterility-causing mutations likely accumulate during divergence after reproductive isolation. Thus, the older the divergence, the higher the number of loci causing reproductive isolation. quote:
I didn't simply mean that multiple roads can lead you to the same phenotype. The references show that
1. The same genotype results in different phenotypes.
2. Changing the genotype does not cause the phenotype to change.
3. One can rewire the developmental circuitry and yet cause no change to the phenotype.
I also never said that the gene has absolutely nothing to do with phenotypic change, my post was meant to show scientific evidence that contradicts the notion that phenotypic change is gene-centric, gradualistic, etc.
In fact, I just read the two references you quoted (the Nature and Dev Biol) and they do nothing of the sort. Maybe you want to be a bit more specific about what you think those 2 works deal with, and what data in particular allow the conclusions you seem to draw.
Finally, I think it's worth pointing out that things like incomplete penetrance, variable expressivity, norm of reaction, phenocopies etc are decades-old (many decades old) concepts in basic genetics. They were well known to evolutionary biologists already at the time of the New Synthesis, and have little if any effect on the validity of current evolutionary models. Johansen himself came up with the terms "genotype" and "phenotype" precisely to underline the non-univocal relationship between the two.
No biologist worth his/her salt subscribes to the strawman one gene-one phenotype model that some ID advocates here claim is supposedly the dominant paradigm. Perhaps picking up a genetics textbook and leafing through it would help define the issue in more realistic terms.
[ 15. June 2003, 12:42: Message edited by: charlie d. ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 16. June 2003 22:55
Charlie d., there is a difference between hybrid inviability and phenotypic traits. Hybrid sterility can in no way shape or form be determined by one gene (monogenic). This is not a nitpick either, you wouldn't pass peer review writing what you did about Nup96, think about it charlie, if changes to Nup96 alone in one species is all that it took to cause sterility, that would mean that it does so regardless of what gene it interacts with, and it wouldn't be selectively advantageous at all (it would not spread). The reproductive isolation that is reported in the paper cannot be the result of a single gene, and actually reading the paper will tell you why. The paper shows that D. simulus Nup96 only causes death when it interacts with D. melanogaster X chromosome.
quote:
The D. simulans Nup96 protein is no longer compatible with an (unknown) interacting factor(s) encoded by the D. melanogaster X chromosome
They are still looking for this other gene.
Charlie wrote:
quote:
This is also unclear. First of all, the number of mapped hybrid sterility loci in melanogaster x simulans is 20, not 200.
Yes but again, thats not the whole story. If you read the paper, they couldn't measure other kinds of interactions, (for example, autosome interactions) and they couldn't do the reverse of the experiment they did where D. melanogaster causes sterility when it interacts with D. simulus. So taking these things into consideration yields about 200 chromosome regions.
Charlie wrote:
quote:
In fact, I just read the two references you quoted (the Nature and Dev Biol) and they do nothing of the sort.
Interesting. Sure I can be a bit more specific, the nature paper shows that messing with signals regulating development may not lead to deleterious changes, because the same signals cause pretty different animals to arise, instead:
quote:
These results demonstrate the evolutionary lability of regulatory genes that are widely viewed as conservative.
This is an example of the same genotypes resulting in different phenotypes, notice that Pim's own quote from one of the author's of the paper supports this interpretation:
quote:
An interesting evolutionary puzzle has recently emerged, involving an apparent disconnection between the evolution of genes and the evolution of anatomy. Anatomically, animals are an enormously diverse group, including such disparate forms as jellyfish, hummingbirds, and lobsters. Despite their anatomical differences, these animals contain, to a first approximation, the same complement of genes.
They would have to posit that master genes actually can change. But the lability is hard to reconcile with why they thought it was conserved. In other words, the hierarchical view of genes (master genes are conserved and therefore cannot be changed, but downstream elements can and therefore natural selection can change them). Now maybe you (or Pim) can respond to my post by actually discussing the data and my interpretation of it.
As for the Dev Biol they write:
quote:
We find a specific range of evolutionary variations at distinct developmental steps. (1) Unlike Caenorhabditis elegans and relatives, the vulva is formed from the four precursor cells P(5-8).p or, exceptionally, from P(6,7).p only. (2) The vulval competence group is restricted to these four cells or is larger. (3) The fates of more anterior and posterior Pn.p cells vary between closely related species (mostly cell death versus epidermal fate). (4) The mechanism of vulval cell fate patterning varies within a single genus, even between strains of the same species. (5) We describe the first example of a vulval cell lineage that is asymmetric between the anterior and the posterior sides of the vulva.
So although things like fate patterning and other developmental steps are different, it results in the same phenotype.
Charlie wrote:
quote:
Finally, I think it's worth pointing out that things like incomplete penetrance,
Whether these things are well known to biologists or not is irrelevant. The problem is that even examples of incomplete penetrance and ve challenge the dogma of Darwinian evolution of phenotypic changes being the result of genes only.
We are in the genomic era now and things look worse than it probably was at the beginning of the modern synthesis. I've seen people attempt to rewrite history when it came to junk DNA, lateral gene transfer, etc, and this is no different. However, it's hard to rewrite history when just a few posts ago you and Rex were arguing for gene changes alone being responsible for phenotypic changes (in that you thought that sterility and traits were somehow identical, and thus individual traits arise through simple changes to single genes and Rex was much more explicit).
I can't explain the situation better than Franklin Harold did in the Way of the Cell, page 199
quote:
Genes and gene products do, of course, retain a role in the evolutionary drama. Catalysts and structural molecules determine the numerical parameters that enter into the physical specification of each system, and they stabilize its organization. Much of that exploration of the range of possible forms is, in fact, carried out by mutation and recombination of genes. But it is system dynamics, not the genetic program, that gives rise to biological forms and function...selection has been ejected from its throne as the dominant creator of biological form.
[ 16. June 2003, 23:43: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 16. June 2003 23:54
Nelson quotes
quote:
Despite their anatomical differences, these animals contain, to a first approximation, the same complement of genes.
to support his claim that "This is an example of the same genotypes resulting in different phenotypes, notice that Pim's own quote from one of the author's of the paper supports this interpretation:"
But he seems to have glossed over the term "to a first approximation".
Nelson continues
quote:
The problem is that even examples of incomplete penetrance and ve challenge the dogma of Darwinian evolution of phenotypic changes being the result of genes only.
How?
So far nothing much that Nelson has said seems to support his claims.
Specific examples would have been helpful but even there he resorts to quoting some comments by the researchers which do not help us resolve the issues.
As far as the Nature paper is concerned it seems that they looked at one specific gene for which "... inviability has evolved between species as a by-product of adaptation of a protein with an essential function. The gene involved encodes a protein component of the nuclear pore complex that has evolved by positive natural selection in both species independently."
Or as the Scientist reports:
quote:
Presgraves et al. analyzed rescued crosses between D. melanogaster and its closest relative, D. simulans, and mapped a hybrid inviability gene to region 95AB on chromosome 3R. They analyzed all 12 loci within this region and identified a single-copy, dicistronic gene encoding nucleoporins Nup98 and Nup96, which function in RNA export and which contain a region surrounding a cleavage site that is perfectly conserved across many species, including humans. Analysis of the gene sequences suggested that positive selection accounted for the changes in both species and that the adaptive event was not recent.
What is fascinating is the 'hitchhiking' observed here, just like in the Avida simulation.
quote:
These results show that a lethal hybrid incompatibility has evolved as a by-product of adaptive protein evolution.
[ 17. June 2003, 00:02: Message edited by: Pim van Meurs ]
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 17. June 2003 00:29
Two German geneticist published a paper in a leading scientific journal where they postulate a similar idea concerning variation induction and/or speciation. They show that all elements to induce variation are already preexisting in the genome and only require activation. In other words a multipurpose genome that can easily induce variation through rearrangements of genetic elements. If you are interested you can find it here:
"Chromosome rearrangements and transposable elements." Annu Rev Genet. 2002;36:389-410.
Their message is clear and it was not unnoticed that it pertains a "creationist's" theory (see: comments in Nature 8th may 2003). Such mechanisms demonstrate how microevolution takes place. And it is NON-Darwinian, since selection is NOT per se involved, often not even relevant.
All that is required to get new species is a bit of shuffling, and (in)activation of preexisting DNA elements. This hypothesis explains a lot more than NeoDarwinian theory ever did.
Apparently what Darwinists take as evolution is already present in the genome. What evolved here exactly? Is shuffling of preexisting DNA elements evolution? I don't think so. It only gives the impression of evolution.
best wishes
Peter
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 17. June 2003 02:38
Although the Lonnig and Saedler paper is a little long on speculation and short on evidence, it makes an excellent companion to Davidson's article. In particular, TE-mediated chromosomal rearrangement is an alternate mechanism to semi-meiosis that will generate a large number of chromosomal rearrangements without requiring switching from sexual to asexual reproduction. (Though the former has the disadvantage of many inviable offspring, the latter has the disadvantage of needing to overcome not only the normal mechanisms restricting development to after fertilization but also maternal and paternal imprinting.)
So it's a highly relevant addition, and I thank Peter for bringing it to my attention.
That said, the authors do not, as Peter claimed, "show that all elements to induce variation are already preexisting in the genome". Rather, the authors argue that in many cases, phenotypic change could be a result of transposable-element induced rearrangements (following McClintock's suggestions, apparently):
quote:
If we have learned anything over the past two decades about animal and plant development, it is that there are as yet no clearcut results to support the proposition that any of the major ontogenetic steps are caused and governed by any known transposable elements, i.e., by means of transposon movements accompanied by small TE-mediated chromosome rearrangements that successively inactivate or activate different suites of genes. To the contrary, this task has been largely assigned to gene classes such as the homeobox, MADS-box, and other regulatory gene families. Current data indicate that, during development, these classes of "controlling elements" do not normally move from one place to another within a given set of chromosomes.
However, active TEs can indeed modulate and modify plant and animal development, as demonstrated in many examples from the work of McClintock and more recent investigators. Yet, the ontogenetic modifications observed and identified to date are not the causes of normal ontogeny; rather they appear to constitute a set of factors that occur more or less concomitantly (and often even disruptive) and are tolerated during development as long as they do not become too heavy a burden for the affected organisms [for details, see (69, 84)].
Furthermore, Peter asks, "Is shuffling of preexisting DNA elements evolution?", to which I answer: Yes!
However, the question then becomes: what constrains the shuffles, and where did those pre-existing elements come from? Did the constraints evolve, and did the elements evolve? If the answer is no, then although the shuffles of existing elements is evolution, the setup did not, and presumably we care about how the setup came to be.
I'd note, though, that firstly, TE-based speciation is less well established than gene-based speciation (since differences in morphology and origins of hybrid lethality are more frequently attributable to changes in single genes than in chromosomal structure right now--though admittedly, our tools to study changes in genes are better, so we're biased towards finding those).
And I'd note that secondly, there isn't any particular reason why TE hotspots can't be accidental, or evolve, or why TE-based chromosomal rearrangements can't just add to the mix of variation created by other types of mutation. In other words, the TE hypothesis advocated by the authors isn't that new or that contradictory to the general idea of evolution--it's more of a question of how significant of a role is played by different factors. As the authors say:
quote:
The most important problems still to be resolved on this topic have to do with causal relationships between chromosome rearrangements and species formations in the wild, selection, and the possibilities and limits of extrapolating from existing results.
1. What is the ratio between "normal" and TE-induced chromosome reshuffling in species in the wild?
2. To what extent is selection involved in the birth of chromosome races or are the majority of these phenomena simply the result of genetic drift (further points below)?
3. What are the exact possibilities and limits of the origin of species by TE-mediated gross and small chromosome rearrangements?
Added in edit: I forgot to mention that with regard to Nup96, I think charlie d. has overlooked the fact that we are trying to understand the difference between two species. The simplest case is where it boils down to a single gene product or chromosomal rearrangement. However, a change in gene product alone is not enough for hybrid sterility or lethality unless Nup96 has two (or more) alleles with the property that the homozygote is viable and the heterozygote is not.
This is not the case; rather, in one background, one version of Nup96 is lethal and the other is not. Thus, there must also be a difference in backgrounds. And so the difference between the species that leads to hybrid sterility/lethality is not adequately explained by Nup96--it only answers half of the question of how to get reproductively incompatable lineages. What Nup96 interacts with is the other (missing, but being worked on) half.
[ 17. June 2003, 02:47: Message edited by: Rex Kerr ]
IP: Logged
|
|
nosivad
Member
Member # 767
|
posted 17. June 2003 06:55
I too thank Peter for the interesting post. I may be alone in this perspective but I really don't see macroevolution going on today, at least through obligatory sexual means. The advantage of the semi-meiotic model is that it does not require that evolution was restricted to small inbreeding populations. It could have occurred anywhere and anytime. The Cambrian explosion might simply represent an initial burst of derepression of the many possible combinations. Similarly, if we are to take Romer's diagrams seriously, the various mammal orders emerged virtually simultaneously. The sexual model just doesn't fill the bill. nosivad
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 17. June 2003 08:10
quote:
Nelson:
Charlie d., there is a difference between hybrid inviability and phenotypic traits. Hybrid sterility can in no way shape or form be determined by one gene (monogenic). This is not a nitpick either, you wouldn't pass peer review writing what you did about Nup96, think about it charlie, if changes to Nup96 alone in one species is all that it took to cause sterility, that would mean that it does so regardless of what gene it interacts with, and it wouldn't be selectively advantageous at all (it would not spread). The reproductive isolation that is reported in the paper cannot be the result of a single gene, and actually reading the paper will tell you why. The paper shows that D. simulus Nup96 only causes death when it interacts with D. melanogaster X chromosome.
quote:
Rex:
And so the difference between the species that leads to hybrid sterility/lethality is not adequately explained by Nup96--it only answers half of the question of how to get reproductively incompatable lineages. What Nup96 interacts with is the other (missing, but being worked on) half.
Maybe I am not making myself clear.
Hybrid sterility is a trait, which by necessity expresses itself after the generation of a hybrid, i.e. in a hybrid genome (how else would one determine hybrid sterility?). In this, appropriate context, Nup76, and each of the other 19 loci in the melanogaster x simulans crosses discussed in the paper, are sufficient to determine the trait.
Note that this is in fact fundamentally true for any genetic trait, even monogenic ones. Heterozygous inactivating mutations in the retinoblastoma gene, Rb, are sufficient to cause a dominant, hereditary, almost completely penetrant form of retinoblastoma, even though such mutations have to operate in the context of a whole cell's replication and cell cycle control machinery, and at the cellular level neoplastic transformation requires other mutations to occur. Furthermore, the retinoblastoma trait requires a human genome to express itself, as mice with heterozygous Rb mutations develop other kinds of tumors (pituitary, IIRC). So, is dominant familial retinoblastoma a monogenic trait? Is Rb mutation sufficient for the trait? Or do we have to conclude that dominant familial retinoblastoma is a polygenic trait, that requires other specific gene(s) in the human genome to express itself?
And Nelson, I can guarantee you that the concept of Nup76 sufficiency for hybrid sterility can pass peer review, because it already has.
The rest later, if I have time.
IP: Logged
|
|
nosivad
Member
Member # 767
|
posted 17. June 2003 08:18
I must disagree with Peter on one crucial point. He questions whether shuffling preexisting information is evolution. I say of course it is, or rather was. Evolution was the production of new life forms and that has certainly occurred. The issue is simply how it occurred.
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 17. June 2003 09:02
There are two distinct questions one could ask regarding the genetic basis of hybrid inviability:
(1) Starting from a common ancestor, what changes are necessary in each lineage to produce reproductive isolation?
(2) Given two reproductively isolated species, which genes in one are lethal to the other, and how did those genes come to be lethal?
The Presgraves et al. paper addresses the second question: it identifies Nup96 (96, not 76), a nuclear pore subunit, as a gene that causes inviability in species hybrids. Furthermore, they show that the divergence in this gene between the two species shows the statistical hallmarks of adaptive evolution.
However, answering the second question only partially answers the first. This is a very important distinction, because the whole point of speciation is that originally, the context is the same in the ancestor of the two species.
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 17. June 2003 21:37
Charlie d.,
Speaking broadly, all genes work within the context of the organism's genome. However, the distinction between hybrid sterility and a trait is pretty clear. For a trait what matters is that the expression depends on the form of the gene in the context of the organism's genome whereas for hybrid sterility it depends on the form of the gene when combined with the form of another gene in the context of the organism's genome.
So back to the fly paper:
D. simulans Nup96 + D. melanogaster X chromosome = hybrid inviability
D. simulans Nup96 + D. simulans X chromosome = diddly squat
It's not monogenic, plain and simple. Whereas, your Rb example is more like what would be described as monogenic. If you are unforutnate enough to inherit the mutated Rb, you'll get retinoblastoma as a consequence and it has absolutely nothing to do with simultaneously inheriting another variant at a second gene.
[ 17. June 2003, 22:21: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 17. June 2003 22:19
Pim writes:
quote:
But he seems to have glossed over the term "to a first approximation".
No I didn't gloss over that term. "First Approximation" means "more or less" or "about the same" or "essentially the same". It does not detract from the main point, which is consistent with my description of the paper.
Pim writes:
quote:
So far nothing much that Nelson has said seems to support his claims.
But you havn't responded to any of my claims nor to any of my references. All you do is assert that I'm wrong in most of your posts. Which is why I rarely respond to you.
You say "specific examples would have been helpful" as if I gave none. In fact, I gave many specific examples, from E. Coli to nematoda.
As far as the fly paper is concerned, you continue to quote material that supports my claims. For example, in the Scientist article they quote the authors in saying:
quote:
"The D. simulans Nup96 protein is no longer compatible with an (unknown) interacting factor(s) encoded by the D. melanogaster X chromosome," conclude the authors.
That is at least two genes.
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 17. June 2003 22:21
quote:
D. simulans Nup96 + D. simulans X chromosome = diddly squat
Oy wey. Are you demanding that a hybrid sterility phenotype be expressed in non-hybrids, in order to define its genetic origin? ![[Roll Eyes]](rolleyes.gif)
Really, this is just semantics at this point. The issue here is that reproductive isolation can clearly arise as a result of straightforward genetic changes, as nucleotide substitutions in gene sequences, and that these changes may have indepenent selectively advantageous effects. At this level, it doesn't even matter if it's one gene, or two, or even a handful. Altogether, this certainly puts a damper on the "mystery of speciation"/"insormountable barrier" schtick. That's the take-home message here.
One more of your misconceptions, before I walk the dog: quote:
Whether these things are well known to biologists or not is irrelevant. The problem is that even examples of incomplete penetrance and ve challenge the dogma of Darwinian evolution of phenotypic changes being the result of genes only.
First of all, incomplete penetrance is perfectly compatible with evolutionary mechanisms. Because evolution acts stochastically, the only result of incomplete penetrance is to quantitatively change the effects of selection. For instance, a 100% penetrant allele that causes a 25% reduction in fertility (fitness 0.75) and a 50% penetrant allele that causes a 50% reduction in fertility (fitness 0.75) are exactly identical as far as evolutionary processes go. Both can be plugged into evolutionary calculations, both will change in population frequency over generations, both can be studied with the same approach. No difference, no problem.
Second, incomplete penetrance can be the result of genes only (e.g., of modifier genes), but even in the cases in which it isn't, again it makes no difference to evolutionary mechanisms. Evolution acts on heritable variation, and as long as a trait is at least partially heritable, the heritable fraction is subject to evolutionary forces.
Finally, "the dogma of Darwinian evolution of phenotypic changes being the result of genes only" exists only in the imagination of those who know little about genetics and even less of evolutionary biology. The role of the environment on the expression of many traits, and the stochastic, non-deterministic nature of evolution were well known long before the New Synthesis, and perfectly compatible with it, let alone with modern evolutionary theory.
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 17. June 2003 23:08
Charlie,
The portion of my quote that you posted has nothing to do with what you asked. That portion states that D. simulans Nup96 does not cause inviability when combined with genes from the D. simulans X chromosome. This is very different from what your Rb example showed.
I don't understand what you mean by "mystery of speciation", there may be, however, the main point is that this is certainly not the result of changes to a single gene. How exactly is this semantics?
As far as incomplete penetrance and Darwinian dogma go, the fact of the matter is that various phenotypes can arise from the same genotype. Population genetics assumes that the phenotype is determined from the genotype and enviornment. Now, it does matter, and matter very much, if examples such as different phenotypes arising from the same genotype or vice versa (and other examples) is not the result of genes only, as my quote from the Way of the Cell and my discussion of the various references showed. The problem is that we can no longer talk in terms of natural selection acting on changes to genes. If you notice, no genes found to date encode huge amounts of information for development. We see gene products in molecular machines, and other constructs, and we see them throw switches but thats about it.
[ 17. June 2003, 23:36: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
|
Printer-friendly view of this topic
