|
Author
|
Topic: The GUToB
|
peter borger
Member
Member # 722
|
posted 18. April 2003 02:49
Dear all,
One of the arguments for ID seems to be the complexity of biological systems. Although I believe that this is a very strong arguments, IMHO the strongest argument for ID is 'biological redundancy'. If ID is on the basis of biology redundancies can be expected to be present on all levels of organic beings.
And yes it is. Redundant systems are present all around and in us:
1) In the brain several pathways for vision are present. Inactivation of one will be compensated for by the activation of another one.
2) The muse given characteristics of the human brain. Musia, humour, story-telling, gossip, art, selfconsciousness, ornate language, ideologies, relegion, morality and arithmics.
3) Regeneration in amphibians and lizards. Regeneration of bones, and liver in general.
4) Autostable seeds of the Zanonia macrocarpa
5) The raptor suite of the Cuculus canorus (although it is adaptive)
6) The newborn swimreflex in conjunction with the gag reflex.
They have all evolved without ID/guidance? Through random mutation and neutral selection?
Let's have a look at:
7) GENETIC REDUNDANCIES.
These redundancies at the level of the genome are the death blow of Darwinism. Why? No association with gene duplication (Winzeler et al, Science 1999, 285:901) and a similar mutation rate as essential genes (Tautz D, TiG 2000, 16:475).
Genetic redundancies are encountered with a lot of disbelief in the orthodox evolutionary community (Nature 2002, 415:8-9). Evolution of robustness?
A careful look demonstrates that genetic redundancies do NOT have a solution in the Darwinian paradigm. E.g. the redundant gene families of the SRC-kinases, the alpha actinins, and the calmodulins. Don't hesitate to ask voor detailed explanations.
8) Missing genes.
Some genes have been predicted to reconcile gene trees. Now HUGO is finsihed we can immediately check such claims. I checked one of the claims: The proposed gene to reconcile the IL-1beta incongruence is not present in the human genome.
So, Darwin really got it wrong!!
(Well, not entirely, but he made a completely unwarranted extension.)
I have set up a new theory: the GUToB (=general and universal theory of biology).
Essentially, it holds that organism have a multipurpose genome that is able to induce variation through non-random mutations.
There is an increasing amount of scientific evidence for NRM in particular in eukaryota where this mechanism is more important than in prokaryota.
Anybody to discuss the GUToB?
Best wishes, Peter
IP: Logged
|
|
Danpech
Member
Member # 163
|
posted 18. April 2003 07:32
The idea of NRM is long overdue.
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 18. April 2003 08:16
Neutral selection?
There was a very good thread on this awhile back, ISTR something about there being an important distinction between redundancy and degeneracy, or some such.
The short Darwinian answer for some of the things that you bring up is that the functions of the "redundant" systems are not really exactly the same, rather they just overlap somewhat.
Then again, many of the things you bring up (bone healing?) aren't redundancies at all, they have obvious utility in keeping one not dead.
IP: Logged
|
|
Grape Ape
Member
Member # 399
|
posted 18. April 2003 14:59
quote:
A careful look demonstrates that genetic redundancies do NOT have a solution in the Darwinian paradigm. E.g. the redundant gene families of the SRC-kinases, the alpha actinins, and the calmodulins.
I'm not familiar with the others, but are you aware that the three calmodulin genes in mammals are all differentially expressed?
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 18. April 2003 18:53
I would be interested in hearing more about redundancy that could not be selected for.
Certain types of redundancy are expected given natural selection and not expected (though possible) given design: specifically, duplicated genes with different expression patterns that can each replace each other (but don't in most tissues, given where they're actually found). This kind of redundancy is, without additional tricks to make it useful, stupid in terms of intelligent design. Unfortunately, when genes duplicate, there is no longer selection to keep both of their expression patterns in the same location, if they're redundant, so this kind of thing really couldn't be avoided in evolution.
Other types of redundancy might be implemented by an intelligent design, but are (most likely) inaccessible to evolution. For instance, entirely redundant systems involving complex structures that come into play only when one structure is damaged, which very rarely happens. "Five nines" computer systems are a good example of this--redundant drives, processors, networking, and so on, most of which never really need to be used, but are there just in case.
I'm afraid I can't easily classify the examples you've given along the range between the two above. For instance, regeneration of damage is highly selectable since you get damaged a lot when living in a world with predators, so that would fall well towards the "selectable/evolvable" side. Most of the others I'm unfamiliar with. (Which brain pathways were you thinking of? I'm not aware of any truly redundant pathways.)
I looked up your references and I noted that the Trends in Genetics commentary by Tautz said that
quote:
Invoking redundancy is not the only possible explanation for gene knockouts that have no phenotype. One could simply propose that it is just a question of finding the right experimental conditions to detect an effect. . . .one could ask whether it is possible that a gene could have such a small function (or fitness benefit) that it could only be revealed by such a tremendously large experimental effort? In fact, this is exactly what population genetical theory predicts. Kimura s theory of neutral evolution shows that, in large populations, vanishingly small selection co-efficients are sufficient to fix a gene that conveys a selective advantage. More precisely, selection can take effect when the fitness difference conveyed by a new allele is larger than 1/(2N), where N is the effective population size.
So, apparently, since theory actually predicts that we should have vanishingly small effects in some cases, finding that there are cases where there are vanishingly small effects should be no huge surprise.
What is needed, therefore, is a more careful look. One way to be more careful is mathematically: can we estimate the expected frequency of genes whose loss causes a given fitness level? Does this, within our ability to predict, correspond to what is observed? Another way to be more careful is to show in more detail why either there is no selection at all, or why weak selection is not enough, for some example systems. Perhaps you can tell us more about the Src-kinases, alpha-actinins, calmodulins, and so on?
(I will add that I know that Src-kinases, despite having broad in-vitro cross-reactivity, are differentially expressed for the most part. It would be more instructive if additional discussion can take differential expression into account.)
It would also be interesting to hear more about the missing genes topic. For instance, I'm not familiar with the "IL-1 beta incongruence". Can you provide a reference, or tell us what the incongruence is and what non-existant gene was proposed to solve it?
IP: Logged
|
|
|
|
peter borger
Member
Member # 722
|
posted 18. April 2003 19:31
Y: Neutral selection?
There was a very good thread on this awhile back, ISTR something about there being an important distinction between redundancy and degeneracy, or some such.
PB: Could you make a link?
Y: The short Darwinian answer for some of the things that you bring up is that the functions of the "redundant" systems are not really exactly the same, rather they just overlap somewhat.
PB: This is a bit too short for me. Please elaborate a bit.
Y: Then again, many of the things you bring up (bone healing?) aren't redundancies at all, they have obvious utility in keeping one not dead.
PB: Many = bone healing? So you agree on the other examples?
Best Wishes,
Peter
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 18. April 2003 19:42
Grape ape says:
I'm not familiar with the others, but are you aware that the three calmodulin genes in mammals are all differentially expressed?
PB says:
Calmodulines are a family of Ca2+ binding proteins present in all known vertebrates and involved in the regulation of intracellular Ca2+ distribution. They are crucial for maintaining biochemical/cellular homeostasis. The calmodulin gene family exists of three variable genes (due to mutations on silent positions and are probably involved in differential gene regulation) and transcribe 5 different mRNAs. Only one invariable calmodulin protein is expressed. Thus, three redundant genes code for only one invariable protein. All known vertebrates, including humans, rats, chicken, frogs, and fish, encode the same three redundant, non-variable (=100% conserved) calmodulin proteins, i.e. neither inter-species nor intra-species differences exist at the level of amino acid sequence. It should also be noted that cells (in which a particular gene of the calmodulin family has been interrupted and therefore not expressed) do not show significantly decreased viability. (ref: Toutenhoofd et al; Cell Calcium; 2000, 28(2): p83-96). This example of redundant super-conserved genes seriously challenges the concept of mutation-selection of being of significance in the maintenance of redundant genetic information.
pb
[ 18. April 2003, 19:56: Message edited by: peter borger ]
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 18. April 2003 19:51
Rex says:
I would be interested in hearing more about redundancy that could not be selected for.
PB: For instance the SRC kinases?
An intriguing example of genetic redundancy is found in a gene family described in a recent issue of Trends in Genetics. (17) The article attempts to explain the existence of the redundant gene family of SRC-type protein tyrosine kinases from an evolutionary point of view. Let us consider this redundant gene family in detail, since it raises a lot of questions about how it could have come into existence by the mechanisms of gene duplication.
This gene family comprises a group of at least eight genes. The SRC-family consists of four closely related kinases, called SRC, YES, FYN and FGR. They are also similar to another closely related subfamily of kinases, known as BLK, HCK, LCK and LYN. The proteins are involved in signal transduction and are intermediates in conducting signals from outside the cell to the nucleus. They operate as molecular switches that activate growth and differentiation of cells. When a cell is triggered by an external stimulus, proteins of the SRC-family are transiently switched on. The family has been extensively studied because it is among the most notorious genes known to man, since they cause cancer as a consequence of single point mutations. A point mutation is a mutation in a gene that causes one nucleotide to be replaced by another nucleotide. Non-silent point mutations will cause the organism’s protein-making machinery to incorporate a wrong amino acid. Consequently, a mutated protein is generated that will immediately induce uncontrolled growth of the cell: cancer. Non-silent point mutations cannot be overcome by allelic compensation, because it causes the molecular switch to be continuously on.
Despite SRC being expressed in many tissues and cell types, the mouse SRC gene knockout is viable. The only obvious characteristic of the knockout is the absence of two front teeth due to osteoporosis. In contrast, there are essentially no point mutations allowed in the SRC protein without severe phenotypic consequences. “Point mutations in most, presumably all, of them can lead to uncontrolled cellular replication, hence cancer “. (17) To study the function of SRC gene products, they have been knocked out in mice. Four of the eight SRC knockouts do not have a detectable phenotype, demonstrating these genes to be redundant.
According to the theory of evolution, the redundant SRC gene family members have its origin in gene duplication. But, truly redundant genes are expected to be reduced to a single copy over time through the random accumulation of mutations that damage the duplicated genes. The existence of the redundant gene family is explained as follows. ”In the redundant gene family of SRC-like proteins, many, perhaps almost all point mutations that damage the protein also cause deleterious phenotypes and kill the organism. The genetic redundancy cannot decay away through the accumulation of point mutations”. (17) Thus, the SRC genes are destined to reside in the genome for ever.
I can think of several convincing arguments why the existence of the two SRC-gene families violates the evolution.
Firstly, the 8 genes are located on 5 different chromosomes. SRC and HCK are located next to each other on chromosome 20, and FRG and LCK are next to each other on chromosome 1. Why surprising? Because, evolution’s mechanism of random accumulation of mutations predicts that similar genes of a gene family that are next to each other, have recently duplicated, and are expected to share the most sequence homology. However, SRC and HCK belong to different subfamilies and share less homology than they do to members of their own families. Exactly the same holds true for FRG and LCK. Despite the putative recent duplication the proteins are members of disparate subfamilies.
Secondly, I would expect that over time duplicated genes plus their regulatory sequences upstream of the genes (promoters) will randomly accumulate mutations by mutational drift, leading to the cessation of expression of the gene. If a gene is not expressed, nothing would withstand accumulation of mutations within in the coding region of the gene, providing the cell an elegant mechanism to get rid of the potentially harmful genes without inducing cancer. However, all SRC-like proteins of both families are expressed and stringently regulated.
Finally and most importantly, if these genes are so potentially harmful that point mutations in the genes are lethal, the existence of this redundant gene family can not be explained by gene duplication. If the first gene duplicated to give rise to the second gene, it is not allowed to mutate because it will invoke a deleterious phenotype that will kill the organism. So, over time, the second gene has to be perfectly stable to avoid deleterious mutations. As a consequence, it will never change. Exactly the same holds true for second, third and additional gene duplications: they are not allowed to change to avoid lethal mutations in the organism. Hence, the evolution theory predicts the SRC-gene family members to be identical genes. Yet, despite sequence similarities, the genes of the SRC- families are far from identical.
17) Toby, J.G. and Spring, J. Genetic redundancy in vertebrates: polyploidy and persistence of genes encoding multidomain proteins. Trends in Genetics 1998, Volume 14(2): p46-49.
pb
[ 18. April 2003, 20:00: Message edited by: peter borger ]
IP: Logged
|
|
Moderator
Administrator
Member # 1
|
posted 18. April 2003 22:52
Peter,
You've redeemed yourself in the last few posts, but Brainstorms discussions need to be self-contained. Please keep from posting external links and rather, spend more time summarizing their main point.
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 18. April 2003 23:11
Rex says:
It would also be interesting to hear more about the missing genes topic. For instance, I'm not familiar with the "IL-1 beta incongruence". Can you provide a reference, or tell us what the incongruence is and what non-existant gene was proposed to solve it?
PB: The molecular genealogy of interleukin 1 beta demonstrates a similar aberration from the species tree as observed for cytochrome c (Zhang, X.H. and Chinnappa C.C. Mol. Biol. Evol 1994, 11: p365-375). ). A careful sequence comparison reveals that the human interleukin-1beta gene is more closely related to the mouse than to the homologue genes in pig and sheep (From: R. Page, Molecular evolution, A phylogenetic approach). Evolutionists must admit that this is not in accord with the species tree and postulate that a fourth gene duplication event is required that caused the aberration. Yet, a thorough scrutiny for IL-1-related genes in the human genome doubts that this event ever took place.
Eight members of the IL-1 related genes in man’s chromosome 2, to be precise in location 2q11-2q13 (OMIM). Sequence comparison of the IL-1 related genes does not present evidence that a recent duplication of IL-1 beta took place in this region. On the contrary, the family tree of the IL-1 genes clearly demonstrates that the common ancestor copy of the IL-1 beta gene duplicated 3 times maximally, and gave rise to IL-1 alpha and IL-1 beta. (Smith, D.E. et al. J Biol Chem, 275: 1169-1175). Thanks to the human genome project, the evolutionary explanation of the aberration of mouse and human IL-1 beta from the species tree can readily be falsified. It should be realised that once a hypothesis is falsified, it remains falsified forever. A falsification is like a mathematical theorem. It cannot be overthrown by new experiments, and it doesn't become old-fashioned.
Best wishes,
Peter
[ 18. April 2003, 23:15: Message edited by: peter borger ]
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 18. April 2003 23:20
Hi Moderator,
You say:
Peter,
You've redeemed yourself in the last few posts, but Brainstorms discussions need to be self-contained. Please keep from posting external links and rather, spend more time summarizing their main point.
PB: Sorry for that. I will not post links anymore but make my point clear. That is: the complete falsification and overturn of the current concept of evolutonary theory and the introduction of a new theory.
Best wishes,
Peter
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 19. April 2003 01:08
Unfortunately for both your point and the authors', much of the Src family is obviously important for fitness. Additional three members of that branch of the pathway have phenotypes:
quote:
Fyn knockouts have some brain defects but are viable. Lck knockouts are T-cell deficient, while Lyn knockouts are B-cell deficient. However, knockouts of Blk, Fgr, Hck and Yes all lack an observable phenotype. Double mutants usually do show a phenotype, for example, the Hck/Fgr double knockout shows a neutrophil adhesion deficiency.
I'd say that missing two front teeth, having brain defects, and missing T- or B-cells are each factors that would seriously impede an organism's fitness in the wild. Lab animals rarely receive the kind of serious immune challenges that their wild counterparts face, so the lack of single knockout phenotypes for the four that show no phenotype can't be taken to very strongly indicate a lack of selectable function. It's a hint, and perhaps the "point mutants tend to be dominant negatives" suggested by Gibson & Sprung is the answer. But it's at best very speculative at the moment.
I think you have a few misconceptions about the genetics of duplication and mutation. Firstly, duplicated genes are sometimes near each other on chromosomes, but certainly not always. Src is around 35700K on chromosome 20, and Hck is way off around 30400K. This isn't so close as to virtually guarantee a duplication event. Secondly, note that promoters tend not to lose their exact expression pattern all at once with a single mutation. Indeed, researchers often engage in "promoter bashing" which involves intentionally mutating a promoter to try to modify it so it will drive expression in similar but different areas to where it already expresses. As such, mutations in promoters would be expected to change expression patterns before abolishing expression completely--and if one copy no longer expresses in tissue X, and the other no longer expresses in tissue Y, both copies are then selected for. However, early stop codons render the gene with the stop unselectable. (I am not familiar with calmodulin expression, but this sounds like the explanation for the calmodulin observation, in addition to the large number of absolutely critical processes that depend on calmodulin.)
I do agree with your point that lethality of point mutations presents a problem for evolution of divergent family members. What remains to be shown is that an extremely high fraction of point mutations are, in fact, lethal. It's a shame that this question is being addressed in vertebrates, as it is hard to generate a large set of mutants easily there, which is the most obvious way to test the hypothesis.
As to the IL-1 beta phylogenetic tree, I have to agree that it's really screwy. I don't really know what is up with that. Frogs cluster with chickens, dolphins with pigs, horses with cats, and humans with rabbits and mice. It would be nice to see the pairwise comparison scores to see if the data is well-represented by a tree at all.
(Note: the authors' names are Gibson, T.B. and Sprung, J..)
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 19. April 2003 03:34
Hi Rex,
R: I do agree with your point that lethality of point mutations presents a problem for evolution of divergent family members. What remains to be shown is that an extremely high fraction of point mutations are, in fact, lethal. It's a shame that this question is being addressed in vertebrates, as it is hard to generate a large set of mutants easily there, which is the most obvious way to test the hypothesis.
PB: It is not only a problem, it is a falsification of a putative mechanism of evolution. It is not the only kinase that gives activated phenotypes in heterozygous point mutants. Something similar holds for pointmutations in phosphatases that terminate signaling pathways involved in cell division. They cannot be explained by gene duplication and random accumulation of mutations.
R: As to the IL-1 beta phylogenetic tree, I have to agree that it's really screwy. I don't really know what is up with that. Frogs cluster with chickens, dolphins with pigs, horses with cats, and humans with rabbits and mice. It would be nice to see the pairwise comparison scores to see if the data is well-represented by a tree at all.
PB: I know what's up with that. Falsifications and false predictions are common for the ToE at the molecular level. It tells me that the theory is wrong and we require a theory that better decribes what we see. That's what a theory is all about: to describe the observed. And to properly predict from it.
Best wishes,
PB
IP: Logged
|
|
peter borger
Member
Member # 722
|
posted 19. April 2003 03:50
Hi Rex:
I think you have a few misconceptions about the genetics of duplication and mutation. Firstly, duplicated genes are sometimes near each other on chromosomes, but certainly not always.Src is around 35700K on chromosome 20, and Hck is way off around 30400K. This isn't so close as to virtually guarantee a duplication event.
PB: No, I do not have a misconception. What is the molecular mechansims underlying the gene duplication? Unequal cross overs? It will yield a second gene next to the one its been derived from Whether they are a couple of Kb or even Mb apart is not relevant. Relevant is what is between them. That determines whether or not the genes have been derived from each other.
R: Secondly, note that promoters tend not to lose their exact expression pattern all at once with a single mutation.
PB: Perhaps not with a single mutation. Acummulation of mutatoins in the DNA elements that induce the gene will, and thus provides the organism an opportunity to get rid of the harmfull genes.
Best wishes,
Peter
IP: Logged
|
|
|
Printer-friendly view of this topic