|
Author
|
Topic: Building an IC System
|
Josh
Member
Member # 405
|
posted 30. July 2003 15:14
The ability of systems to accumulate changes with Hsp90 and exhibit large variations due to stress is quite an interesting mechanism hypothesized to expand evolutionary pathways. This is one possible molecular mechanism that could support Gould’s staltutory evolutionary change models (mass acting RM&NS?). The following article deals with some parameters of this mechanism. Could this be part of the indirect mechanism capable of building "IC" in organisms?
Nature 424, 549 - 552 (31 July 2003); doi:10.1038/nature01765
Evolutionary capacitance as a general feature of complex gene networks
AVIV BERGMAN1,2 AND MARK L. SIEGAL1
1 Department of Biological Sciences, Stanford University, Stanford, California 94305-5020, USA
2 Center for Computational Genetics and Biological Modeling, Stanford University, Stanford, California 94305-5020, USA
Correspondence and requests for materials should be addressed to M.L.S. (mlsiegal@stanford.edu).
An evolutionary capacitor buffers genotypic variation under normal conditions, thereby promoting the accumulation of hidden polymorphism. But it occasionally fails, thereby revealing this variation phenotypically. The principal example of an evolutionary capacitor is Hsp90, a molecular chaperone that targets an important set of signal transduction proteins. Experiments in Drosophila and Arabidopsis have demonstrated three key properties of Hsp90: (1) it suppresses phenotypic variation under normal conditions and releases this variation when functionally compromised; (2) its function is overwhelmed by environmental stress; and (3) it exerts pleiotropic effects on key developmental processes. But whether these properties necessarily make Hsp90 a significant and unique facilitator of adaptation is unclear. Here we use numerical simulations of complex gene networks, as well as genome-scale expression data from yeast single-gene deletion strains, to present a mechanism that extends the scope of evolutionary capacitance beyond the action of Hsp90 alone. We illustrate that most, and perhaps all, genes reveal phenotypic variation when functionally compromised, and that the availability of loss-of-function mutations accelerates adaptation to a new optimum phenotype. However, this effect does not require the mutations to be conditional on the environment. Thus, there might exist a large class of evolutionary capacitors whose effects on phenotypic variation complement the systemic, environment-induced effects of Hsp90.
IP: Logged
|
|
John Bracht
Member
Member # 5
|
posted 12. August 2003 14:51
Josh,
This is really interesting stuff, and I definitely haven't done enough research to be knowledgeable in this area; I just wanted to offer my initial impressions.
It seems to me that this mechanism of capacitance is actually an information-suppressing mechanism. After all, if we imagine a Darwinian fitness function as having gently sloping sides that gradually ascends toward some peak (or network of connected peaks), we can imainge this evolutionary capacitance as effectively flattening out the landscape by requiring adaptive change to occur by a succession of accumulated, but unselected and "blind" changes that suddenly get expressed all at once. In this mechanism, finding adaptive peaks isn't guided along gradually (as one might imagine the Darwinian mechanism gradually moving up a fitness peak) but rather is achieved by large "jumps" on the landscape with the hopes of stumbling upon the rare butte sticking out of the desert. The flattening of the landscape comes not because the landscape is actually flat, but because the search algorithm doesn't allow one to move about on that landscape--instead, it hides (or buffers) the effects of individual mutations, and relies upon extremely low-probability fortuitous combinations of mutations, expressed co-incidentally, to take the organism to a new region of functionality. It stays in one place, then "leaps" to another place. In so doing, the entire concept of intermediate forms is eliminated, since the system is either buffered by the capacitative protein (in which case mutational changes have no functional effect, ie, they don't move one about on the fitness landscape), or they are not (in which case the organism makes a sudden large jump to a new region on the landscape as built-up change is unleashed) By not allowing local searching on the fitness landscape (ie, through small functional changes through mutations), the information encoded in the fitness landscape is completely irrelevant to the origin of biological complexity by this means. This mechanism of evolutionary capacitance seems highly non-Darwinian in nature, and I'd like to see some examples of the sorts of biocomplexity that it is capable of producing.
John
IP: Logged
|
|
|
|
Grape Ape
Member
Member # 399
|
posted 13. August 2003 13:04
John--
quote:
It seems to me that this mechanism of capacitance is actually an information-suppressing mechanism. After all, if we imagine a Darwinian fitness function as having gently sloping sides that gradually ascends toward some peak (or network of connected peaks), we can imainge this evolutionary capacitance as effectively flattening out the landscape by requiring adaptive change to occur by a succession of accumulated, but unselected and "blind" changes that suddenly get expressed all at once.
My understanding is that the capacitance doesn't affect the fitness landscape at all, but rather affects the area of the landscape over which a population might be represented. For example, if you imagine a population as a conglomeration of points on a fixed landscape, the capacitance would keep the conglomeration tight. When the canalization mechanisms get disturbed, the conglomeration of points spreads out, making them more likely to overlap a slope which some of them might then start climbing.
Of course the problem with this example is that the fitness landscape is not fixed; the disturbance of the canalization mechanism means that the landscape was shifted in some way as to move the population off a peak, hence causing environmental stress, and hence causing Hsp90 to become functionally compromised. But while a shift in the landscape is a necessary precursor to the distubance of canalization, that disturbance by itself does not alter the landscape. And in any case, there is no reason to suspect that the landscape is being flattened.
I don't yet know how this would relate to the evolution of biological complexity or IC in particular. At this point I think the most relevant result of this research is to demonstrate a mechanism for puncuated periods of stasis and rapid change.
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 15. August 2003 13:00
Let's see if we can get the metaphor of "capacitance" straight here. Bergman and Siegel say that "An evolutionary capacitor buffers genotypic variation under normal conditions, thereby promoting the accumulation of hidden polymorphism." In electronics, a capacitor is a device for storing charge: it hoards electrical charge. Under certain conditions, as when the voltage threshold of a Zener diode downstream from the capacitor is exceeded, that charge can be (more or less suddenly) released. In Bergman and Siegel's usage, then, I interpret the "capacitance" analogy to mean that Hsp90 in effect allows the accumulation of genotypic variability while not permitting it to be expressed phenotypically and thus be exposed to selection.
Then, under some circumstances - environmental stress is the example given - Hsp90's expression-suppressing action is overwhelmed and the underlying 'stored' genotypic variability is released to phenotypic expression and is thus exposed to selection. That increased expressed phenotypic variability produces a wider range of candidate adaptations and thus increases the probability that the population will find a new 'optimum' (though I intensely dislike that word in this context).
Finally, Bergman and Siegel suggest that there is a large class of such genotypic-variability-storage devices that under certain circumstances 'release' stored genotypic variability to be expressed in phenotypes and thus be exposed to selection.
Therefore, contrary to John Bracht's suggestion, Bergman and Siegel's capacitance analogy is not to be interpreted as "an information-suppressing mechanism," but rather as (if anything) an information-accumulating mechanism. The underlying unexpressed genotypic variability forms a latent pool of variability that (when environmental conditions enable it) is released to phenotypic expression and hence to selection.
As GA suggested, the increased phenotypic variability in effect relatively rapidly spreads the population of phenotypes (organisms) around on a (possibly deforming due to environmental stress) fitness landscape, so that relatively rapid movement of the population on the landscape is enabled, but I see no reason for the process to invoke "saltational jumps" - great leaps over chasms. John's locution "the organism makes a sudden large jump" (emphasis added) is not appropriate. The population may be enabled to shift rapidly on a deforming landscape, due to the increased phenotypic variability, and thus evolution might proceed at a relatively rapid rate, but an individual organism (seen as a vehicle for expressing the latent genetic variability) does what it always did: live or die and reproduce differentially as a function of its relative fitness.
Finally, John wrote quote:
By not allowing local searching on the fitness landscape (ie, through small functional changes through mutations), the information encoded in the fitness landscape is completely irrelevant to the origin of biological complexity by this means. This mechanism of evolutionary capacitance seems highly non-Darwinian in nature, and I'd like to see some examples of the sorts of biocomplexity that it is capable of producing.
In fact, evolutionary capacitance does not seem to me to be "non-Darwinian." It allows the accumulation of genotypic variability without constant selective pruning. However, the dominant Darwinian mechanism, natural selection, is alive and well in this kind of scenario. Just as soon as the underlying genotypic variability is released to be expressed in phenotypes it is subject to selection, and it is selection that will drive the population on the deforming fitness landscape. What Bergman and Siegel suggest is that since genotypic variability can be accumulated and then released, evolutionary "capacitance" such as that mediated by Hsp90 provides a mechanism for spurts of (relatively) rapid Darwinian evolution. This is one kind of mechanism that can produce what is often observed both in nature and in evolutionary simulations running in sufficiently complex and dynamic environments, namely changing rates of evolution of a population over time, one of the mechanisms that refute the "constant speedism" caricature of Darwinian evolution.
RBH
Added in edit: I should note that the evolutionary simulation linked above did not explicitly employ evolutionary capacitance; its variations in rate of evolution are due mostly (I believe) to the dynamics of deformation of the underlying (real-world) selective environment, though the simulation did allow for 'neutral' genotypic variation. Nevertheless, it does illustrate one sort of general pattern of varying rates of fitness change through time one would also expect from a system with buffered genotypic variability that is intermittently released to be exposed to selection.
=========
In still later edit: This is a better illustration of the variability in rate of change of fitness, and indirectly the variability of the rate of evolution in the simulation mentioned above. Note that it is an indirect indicator and depends both on the properties of the population and on the topography of the fitness landscape. Nevertheless, as noted, it illustrates the potential variability in rate of change of a population under selective pressure.
[ 15. August 2003, 15:40: Message edited by: RBH ]
IP: Logged
|
|
|
Printer-friendly view of this topic