Genetic assimilation

Last updated

Genetic assimilation is a process described by Conrad H. Waddington by which a phenotype originally produced in response to an environmental condition, such as exposure to a teratogen, later becomes genetically encoded via artificial selection or natural selection. Despite superficial appearances, this does not require the (Lamarckian) inheritance of acquired characters, although epigenetic inheritance could potentially influence the result. [1] Waddington stated that genetic assimilation overcomes the barrier to selection imposed by what he called canalization of developmental pathways; he supposed that the organism's genetics evolved to ensure that development proceeded in a certain way regardless of normal environmental variations.

Contents

The classic example of genetic assimilation was a pair of experiments in 1942 and 1953 by Waddington. He exposed Drosophila fruit fly embryos to ether, producing an extreme change in their phenotype: they developed a double thorax, resembling the effect of the bithorax gene. This is called a homeotic change. Flies which developed halteres (the modified hindwings of true flies, used for balance) with wing-like characteristics were chosen for breeding for 20 generations, by which point the phenotype could be seen without other treatment. [2]

Waddington's explanation has been controversial, and has been accused of being Lamarckian. More recent evidence appears to confirm the existence of genetic assimilation in evolution; in yeast, when a stop codon is lost by mutation, the reading frame is preserved much more often than would be expected. [3] Genetic assimilation has been incorporated into the extended evolutionary synthesis. [4] [5] [6] [7]

History

Waddington's experiments

Normal and cross-veinless Drosophila wings Drosophile normale et cross-veinless.jpg
Normal and cross-veinless Drosophila wings

Conrad H. Waddington's classic experiment (1942) induced an extreme environmental reaction in the developing embryos of Drosophila . In response to ether vapor, a proportion of embryos developed a radical phenotypic change, a second thorax. At this point in the experiment bithorax is not innate; it is induced by an unusual environment. Waddington then repeatedly selected Drosophila for the bithorax phenotype over some 20 generations. After this time, some Drosophila developed bithorax without the ether treatment. [8]

Waddington carried out a similar experiment in 1953, this time inducing the cross-veinless phenocopy in Drosophila with a heat shock, with 40% of the flies showing the phenotype prior to selection. Again he selected for the phenotype over several generations, applying heat shock each time, and eventually the phenotype appeared even without heat shock. [9] [10]

Waddington's explanation

Waddington called the effect he had seen "genetic assimilation". His explanation was that it was caused by a process he called "canalization". He compared embryonic development to a ball rolling down a slope in what he called an epigenetic landscape, where each point on the landscape is a possible state of the organism (involving many variables). As a particular pathway becomes entrenched or "canalized", it becomes more stable, likely to occur even in the face of environmental changes. Major perturbations such as ether or heat shock eject the developmental pathway from the metaphorical canal, exploring other parts of the epigenetic landscape. Selection in the presence of that perturbation leads to the evolution of a new canal; after the perturbation is discontinued, developmental trajectories continue to follow the canalized pathway. [10]

A Darwinian explanation

Other evolutionary biologists have agreed that assimilation occurs, but give a different, purely quantitative genetics explanation in terms of Darwin's natural or artificial selection. The phenotype, say cross-veinless, is presumed to be caused by a combination of multiple genes. The phenotype appears when the sum of gene effects exceeds a threshold; if that threshold is lowered by a perturbation, say a heat shock, the phenotype is more likely to be seen. Continued selection under perturbing conditions increases the frequency of the alleles of genes that promote the phenotype until the threshold is breached, and the phenotype appears without requiring the heat shock. [10] [11]

Perturbations can be genetic or epigenetic rather than environmental. For example, Drosophila fruit flies have a heat shock protein, Hsp90, which protects the development of many structures in the adult fly from heat shock. If the protein is damaged by a mutation, then just as if it were damaged by the environmental effects of drugs, many different phenotypic variants appear; if these are selected for, they quickly establish without further need for the mutant Hsp90. [12]

A mutational explanation

In 2017, L. Fanti and colleagues replicated Waddington's experiments, but included DNA sequencing, revealing that the wing phenotypes were due to mutational events, small deletions and the insertions of transposable elements that were mobilised by the heat exposure. [13]

Neo-Darwinism or Lamarckism

Waddington's genetic assimilation compared to Lamarckism, Darwinian evolution, and the Baldwin effect. All the theories offer explanations of how organisms respond to a changed environment with adaptive inherited change. Lamarck Compared to Darwin, Baldwin, Waddington.svg
Waddington's genetic assimilation compared to Lamarckism, Darwinian evolution, and the Baldwin effect. All the theories offer explanations of how organisms respond to a changed environment with adaptive inherited change.

Waddington's theory of genetic assimilation was controversial. [4] The evolutionary biologists Theodosius Dobzhansky and Ernst Mayr both thought that Waddington was using genetic assimilation to support so-called Lamarckian inheritance. They denied that the inheritance of acquired characteristics had taken place, and asserted that Waddington had simply observed the natural selection of genetic variants that already existed in the study population. [14] Waddington himself interpreted his results in a Neo-Darwinian way, particularly emphasizing that they "could bring little comfort to those who wish to believe that environmental influences tend to produce heritable changes in the direction of adaptation." [1] [15] [16] The evolutionary developmental biologist Adam S. Wilkins wrote that "[Waddington] in his lifetime... was widely perceived primarily as a critic of Neo-Darwinian evolutionary theory. His criticisms ... were focused on what he saw as unrealistic, 'atomistic' models of both gene selection and trait evolution." In particular, according to Wilkins, Waddington felt that the Neo-Darwinians badly neglected the phenomenon of extensive gene interactions and that the 'randomness' of mutational effects, posited in the theory, was false. [17] Even though Waddington became critical of the neo-Darwinian synthetic theory of evolution, he still described himself as a Darwinian, and called for an extended evolutionary synthesis based on his research. [18] Waddington did not deny the threshold-based conventional genetic interpretation of his experiments, but regarded it "as a told to the children version of what I wished to say" and considered the debate to be about "mode of expression, rather than of substance". [19] Both genetic assimilation and the related Baldwin effect are theories of phenotypic plasticity, where aspects of an organism's physiology and behaviour are affected by the environment. The evolutionary ecologist Erika Crispo states that they differ in that genetic assimilation decreases the level of plasticity (returning to Waddington's original definition of canalization; whereas the Baldwin effect may increase it) but does not change the mean phenotypic value (where the Baldwin effect changes it). [20] Crispo defines genetic assimilation as a kind of genetic accommodation, "evolution in response to both genetically based and environmentally induced novel traits", [20] which in turn is in her view central to the Baldwin effect. [20]

Relationship to adaptation

Mathematical modeling suggests that under certain circumstances, natural selection favours the evolution of canalization that is designed to fail under extreme conditions. [21] [22] If the result of such a failure is favoured by natural selection, genetic assimilation occurs. In the 1960s, Waddington and his colleague the animal geneticist J. M. Rendel argued for the importance of genetic assimilation in natural adaptation, as a means of providing new and potentially beneficial variation to populations under stress, enabling them to evolve rapidly. [23] [24] Their contemporary George C. Williams argued that genetic assimilation proceeds at the cost of a loss of previously adaptive developmental plasticity, and therefore should be seen as resulting in a net loss rather than gain of complexity, making it in his view uninteresting from the perspective of the constructive process of adaptation. [25] However, the preceding phenotypic plasticity need not be adaptive, but simply represent a breakdown of canalization. [21]

A 2023 transcriptomic analysis revealed that genetic assimilation in environmental adaptations is rare. [26]

In natural populations

Notechis scutatus tiger snakes from islands, like this one from Chappell Island, have larger heads than mainland populations, apparently genetically assimilated. Chappell Island tiger snake.jpg
Notechis scutatus tiger snakes from islands, like this one from Chappell Island, have larger heads than mainland populations, apparently genetically assimilated.

Several instances of genetic assimilation have been documented contributing to natural selection in the wild. For example, populations of the island tiger snakes ( Notechis scutatus ) have become isolated on islands and have larger heads to cope with large prey animals. Young populations have larger heads by phenotypic plasticity, whereas large heads have become genetically assimilated in older populations. [27]

In another example, patterns of left-right asymmetry or "handedness", when present, can be determined either genetically or plastically. During evolution, genetically determined directional asymmetry, as in the left-oriented human heart, can arise either from a nonheritable (phenotypic) developmental process, or directly by mutation from a symmetric ancestor. An excess of transitions from plastically determined to genetically determined handedness points to the role of genetic assimilation in evolution. [28]

A third example has been seen in yeast. Evolutionary events in which stop codons are lost preserve the reading frame much more often than would be expected from mutation bias. This finding is consistent with the role of the yeast prion [PSI+] in epigenetically facilitating stop codon readthrough, followed by genetic assimilation via the permanent loss of the stop codon. [3]

See also

Related Research Articles

<span class="mw-page-title-main">Heredity</span> Passing of traits to offspring from the species parents or ancestor

Heredity, also called inheritance or biological inheritance, is the passing on of traits from parents to their offspring; either through asexual reproduction or sexual reproduction, the offspring cells or organisms acquire the genetic information of their parents. Through heredity, variations between individuals can accumulate and cause species to evolve by natural selection. The study of heredity in biology is genetics.

<span class="mw-page-title-main">Genotype–phenotype distinction</span> Distinction made in genetics

The genotype–phenotype distinction is drawn in genetics. "Genotype" is an organism's full hereditary information. "Phenotype" is an organism's actual observed properties, such as morphology, development, or behavior, and the consequences thereof. This distinction is fundamental in the study of inheritance of traits and their evolution.

<span class="mw-page-title-main">Evolutionary developmental biology</span> Comparison of organism developmental processes

Evolutionary developmental biology is a field of biological research that compares the developmental processes of different organisms to infer how developmental processes evolved.

<span class="mw-page-title-main">Modern synthesis (20th century)</span> Fusion of natural selection with Mendelian inheritance

The modern synthesis was the early 20th-century synthesis of Charles Darwin's theory of evolution and Gregor Mendel's ideas on heredity into a joint mathematical framework. Julian Huxley coined the term in his 1942 book, Evolution: The Modern Synthesis. The synthesis combined the ideas of natural selection, Mendelian genetics, and population genetics. It also related the broad-scale macroevolution seen by palaeontologists to the small-scale microevolution of local populations.

A maternal effect is a situation where the phenotype of an organism is determined not only by the environment it experiences and its genotype, but also by the environment and genotype of its mother. In genetics, maternal effects occur when an organism shows the phenotype expected from the genotype of the mother, irrespective of its own genotype, often due to the mother supplying messenger RNA or proteins to the egg. Maternal effects can also be caused by the maternal environment independent of genotype, sometimes controlling the size, sex, or behaviour of the offspring. These adaptive maternal effects lead to phenotypes of offspring that increase their fitness. Further, it introduces the concept of phenotypic plasticity, an important evolutionary concept. It has been proposed that maternal effects are important for the evolution of adaptive responses to environmental heterogeneity.

Evolvability is defined as the capacity of a system for adaptive evolution. Evolvability is the ability of a population of organisms to not merely generate genetic diversity, but to generate adaptive genetic diversity, and thereby evolve through natural selection.

<span class="mw-page-title-main">Baldwin effect</span> Effect of learned behavior on evolution

In evolutionary biology, the Baldwin effect, a phenotype-first theory of evolution, describes the effect of learned behaviour on evolution. James Mark Baldwin and others suggested during the eclipse of Darwinism in the late 19th century that an organism's ability to learn new behaviours will affect its reproductive success and will therefore have an effect on the genetic makeup of its species through natural selection. Though this process appears similar to Lamarckism, that view proposes that living things inherited their parents' acquired characteristics. The Baldwin effect has been independently proposed several times, and today it is generally recognized as part of the modern synthesis.

<span class="mw-page-title-main">C. H. Waddington</span> British biologist

Conrad Hal Waddington was a British developmental biologist, paleontologist, geneticist, embryologist and philosopher who laid the foundations for systems biology, epigenetics, and evolutionary developmental biology.

Evolutionary capacitance is the storage and release of variation, just as electric capacitors store and release charge. Living systems are robust to mutations. This means that living systems accumulate genetic variation without the variation having a phenotypic effect. But when the system is disturbed, robustness breaks down, and the variation has phenotypic effects and is subject to the full force of natural selection. An evolutionary capacitor is a molecular switch mechanism that can "toggle" genetic variation between hidden and revealed states. If some subset of newly revealed variation is adaptive, it becomes fixed by genetic assimilation. After that, the rest of variation, most of which is presumably deleterious, can be switched off, leaving the population with a newly evolved advantageous trait, but no long-term handicap. For evolutionary capacitance to increase evolvability in this way, the switching rate should not be faster than the timescale of genetic assimilation.

<span class="mw-page-title-main">Facilitated variation</span>

The theory of facilitated variation demonstrates how seemingly complex biological systems can arise through a limited number of regulatory genetic changes, through the differential re-use of pre-existing developmental components. The theory was presented in 2005 by Marc W. Kirschner and John C. Gerhart.

<span class="mw-page-title-main">Canalisation (genetics)</span> Measure of the ability of a population to produce the same phenotype

Canalisation is a measure of the ability of a population to produce the same phenotype regardless of variability of its environment or genotype. It is a form of evolutionary robustness. The term was coined in 1942 by C. H. Waddington to capture the fact that "developmental reactions, as they occur in organisms submitted to natural selection...are adjusted so as to bring about one definite end-result regardless of minor variations in conditions during the course of the reaction". He used this word rather than robustness to consider that biological systems are not robust in quite the same way as, for example, engineered systems.

In biology, saltation is a sudden and large mutational change from one generation to the next, potentially causing single-step speciation. This was historically offered as an alternative to Darwinism. Some forms of mutationism were effectively saltationist, implying large discontinuous jumps.

<span class="mw-page-title-main">Robustness (evolution)</span> Persistence of a biological trait under uncertain conditions

In evolutionary biology, robustness of a biological system is the persistence of a certain characteristic or trait in a system under perturbations or conditions of uncertainty. Robustness in development is known as canalization. According to the kind of perturbation involved, robustness can be classified as mutational, environmental, recombinational, or behavioral robustness etc. Robustness is achieved through the combination of many genetic and molecular mechanisms and can evolve by either direct or indirect selection. Several model systems have been developed to experimentally study robustness and its evolutionary consequences.

<span class="mw-page-title-main">Transgenerational epigenetic inheritance</span> Epigenetic transmission without DNA primary structure alteration

Transgenerational epigenetic inheritance is the transmission of epigenetic markers and modifications from one generation to multiple subsequent generations without altering the primary structure of DNA. Thus, the regulation of genes via epigenetic mechanisms can be heritable; the amount of transcripts and proteins produced can be altered by inherited epigenetic changes. In order for epigenetic marks to be heritable, however, they must occur in the gametes in animals, but since plants lack a definitive germline and can propagate, epigenetic marks in any tissue can be heritable.

<span class="mw-page-title-main">Phenotypic integration</span>

Phenotypic Integration is a metric for measuring the correlation of multiple functionally-related traits to each other. Complex phenotypes often require multiple traits working together in order to function properly. Phenotypic integration is significant because it provides an explanation as to how phenotypes are sustained by relationships between traits. Every organism's phenotype is integrated, organized, and a functional whole. Integration is also associated with functional modules. Modules are complex character units that are tightly associated, such as a flower. It is hypothesized that organisms with high correlations between traits in a module have the most efficient functions. The fitness of a particular value for one phenotypic trait frequently depends on the value of the other phenotypic traits, making it important for those traits evolve together. One trait can have a direct effect on fitness, and it has been shown that the correlations among traits can also change fitness, causing these correlations to be adaptive, rather than solely genetic. Integration can be involved in multiple aspects of life, not just at the genetic level, but during development, or simply at a functional level.

The Extended Evolutionary Synthesis (EES) consists of a set of theoretical concepts argued to be more comprehensive than the earlier modern synthesis of evolutionary biology that took place between 1918 and 1942. The extended evolutionary synthesis was called for in the 1950s by C. H. Waddington, argued for on the basis of punctuated equilibrium by Stephen Jay Gould and Niles Eldredge in the 1980s, and was reconceptualized in 2007 by Massimo Pigliucci and Gerd B. Müller.

State switching is a fundamental physiological process in which a cell/organism undergoes spontaneous, and potentially reversible, transitions between different phenotypes. Thus, the ability to switch states/phenotypes is a key feature of development and normal function of cells within most multicellular organisms that enables the cell to respond to various intrinsic and extrinsic cues and stimuli in a concerted fashion enabling them to ‘make’ appropriate cellular decisions. Although state switching is essential for normal functioning, the repertoire of phenotypes in a normal cell is albeit limited.

In biology, constructive development refers to the hypothesis that organisms shape their own developmental trajectory by constantly responding to, and causing, changes in both their internal state and their external environment. Constructive development can be contrasted with programmed development, the hypothesis that organisms develop according to a genetic program or blueprint. The constructivist perspective is found in philosophy, most notably developmental systems theory, and in the biological and social sciences, including developmental psychobiology and key themes of the extended evolutionary synthesis. Constructive development may be important to evolution because it enables organisms to produce functional phenotypes in response to genetic or environmental perturbation, and thereby contributes to adaptation and diversification.

In evolutionary biology, developmental bias refers to the production against or towards certain ontogenetic trajectories which ultimately influence the direction and outcome of evolutionary change by affecting the rates, magnitudes, directions and limits of trait evolution. Historically, the term was synonymous with developmental constraint, however, the latter has been more recently interpreted as referring solely to the negative role of development in evolution.

Ecological evolutionary developmental biology (eco-evo-devo) is a field of biology combining ecology, developmental biology and evolutionary biology to examine their relationship. The concept is closely tied to multiple biological mechanisms. The effects of eco-evo-devo can be a result of developmental plasticity, the result of symbiotic relationships or epigenetically inherited. The overlap between developmental plasticity and symbioses rooted in evolutionary concepts defines ecological evolutionary developmental biology. Host- microorganisms interactions during development characterize symbiotic relationships, whilst the spectrum of phenotypes rooted in canalization with response to environmental cues highlights plasticity. Developmental plasticity that is controlled by environmental temperature may put certain species at risk as a result of climate change.

References

  1. 1 2 Pocheville, Arnaud; Danchin, Etienne (January 1, 2017). "Chapter 3: Genetic assimilation and the paradox of blind variation". In Huneman, Philippe; Walsh, Denis (eds.). Challenging the Modern Synthesis. Oxford University Press.
  2. Gilbert, Scott F. (1991). "Induction and the Origins of Developmental Genetics". A Conceptual History of Modern Embryology. Plenum Press. pp. 181–206. ISBN   978-0306438424.
  3. 1 2 Giacomelli, M. G.; Hancock, A. S.; Masel, Joanna (13 November 2006). "The Conversion of 3' UTRs into Coding Regions". Molecular Biology and Evolution. 24 (2): 457–464. doi:10.1093/molbev/msl172. PMC   1808353 . PMID   17099057.
  4. 1 2 Pigliucci, M. (2006). "Phenotypic plasticity and evolution by genetic assimilation". Journal of Experimental Biology. 209 (12): 2362–2367. doi: 10.1242/jeb.02070 . ISSN   0022-0949. PMID   16731812.
  5. Pigliucci, Massimo. Phenotypic Plasticity. In Massimo Pigliucci, and Gerd B. Müller (eds), Evolution: The Extended Synthesis (Cambridge, MA, 2010; online edn, MIT Press Scholarship Online, 22 Aug. 2013).
  6. Loison, Laurent (2019). "Canalization and genetic assimilation: Reassessing the radicality of the Waddingtonian concept of inheritance of acquired characters". Semin Cell Dev Biol. 88: 4–13. doi: 10.1016/j.semcdb.2018.05.009 . PMID   29763656.
  7. Aaby, Bendik Hellem (2022). "The Ecological Dimension of Natural Selection". Philosophy of Science. 88 (5): 1199–1209. doi:10.1086/714999.
  8. Waddington, C. H. (1942). "Canalization of development and the inheritance of acquired characters". Nature . 150 (3811): 563–565. Bibcode:1942Natur.150..563W. doi:10.1038/150563a0. S2CID   4127926.
  9. Waddington, C. H. (1953). "Genetic Assimilation of an Acquired Character". Evolution. 7 (2): 118–126. doi:10.2307/2405747. JSTOR   2405747.
  10. 1 2 3 Masel, Joanna (2004). "Genetic assimilation can occur in the absence of selection for the assimilating phenotype, suggesting a role for the canalization heuristic" (PDF). Journal of Evolutionary Biology. 17 (5): 1106–1110. doi: 10.1111/j.1420-9101.2004.00739.x . PMID   15312082. S2CID   11050630.
  11. Falconer, D. S.; Mackay, Trudy F. C. (1998). Introduction to quantitative genetics (4th ed.). Longman. pp.  309–310. ISBN   978-0-582-24302-6.
  12. Rutherford, S. L.; Lindquist, S. (1998). "Hsp90 as a capacitor for morphological evolution". Nature . 396 (6709): 336–342. Bibcode:1998Natur.396..336R. doi:10.1038/24550. PMID   9845070. S2CID   204996106.
  13. Fanti, L.; Piacenti, L. (2017). "Canalization by Selection of de Novo Induced Mutations". Genetics. 206 (4): 1995–2006. doi:10.1534/genetics.117.201079. PMC   5560803 . PMID   28576865.
  14. Gilbert, Scott F. (2013). A Conceptual History of Modern Embryology: Volume 7: A Conceptual History of Modern Embryology. Springer. p. 205. ISBN   978-1-4615-6823-0.
  15. Waddington, Conrad (1953). "Genetic assimilation of an acquired character". Evolution. 7 (2): 118–126. doi:10.2307/2405747. JSTOR   2405747.
  16. Noble, Denis (2015). "Conrad Waddington and the origin of epigenetics". Journal of Experimental Biology. 218 (6): 816–818. doi: 10.1242/jeb.120071 . PMID   25788723.
  17. Wilkins, Adam S. (2015). "Waddington's Unfinished Critique of Neo-Darwinian Genetics: Then and Now". Biological Theory. 3 (3): 224–232. doi:10.1162/biot.2008.3.3.224. ISSN   1555-5542. S2CID   84217300.
  18. Huang, Sui (2012). "The molecular and mathematical basis of Waddington's epigenetic landscape: A framework for post-Darwinian biology?". BioEssays. 34 (2): 149–157. doi:10.1002/bies.201100031. ISSN   0265-9247. PMID   22102361. S2CID   19632484.
  19. Waddington, C. H. (1 November 1958). "Comment on Professor Stern's Letter". The American Naturalist. 92 (867): 375–376. doi:10.1086/282049. ISSN   0003-0147. S2CID   83910453.
  20. 1 2 3 Crispo, Erika (2007). "The Baldwin effect and genetic assimilation: revisiting two mechanisms of evolutionary change mediated by phenotypic plasticity". Evolution. 61 (11): 2469–2479. doi: 10.1111/j.1558-5646.2007.00203.x . PMID   17714500. S2CID   9292273.
  21. 1 2 Eshel, I.; Matessi, C. (August 1998). "Canalization, genetic assimilation and preadaptation. A quantitative genetic model". Genetics. 149 (4): 2119–2133. doi:10.1093/genetics/149.4.2119. PMC   1460279 . PMID   9691063.
  22. Pigliucci, Massimo; Murren, C. J. (July 2003). "Perspective: Genetic assimilation and a possible evolutionary paradox: can macroevolution sometimes be so fast as to pass us by?". Evolution. 57 (7): 1455–64. doi: 10.1111/j.0014-3820.2003.tb00354.x . PMID   12940351.
  23. Hall, Brian K. (1992). "Waddington's Legacy in Development and Evolution" (PDF). American Zoologist. 32 (1): 113–122. doi: 10.1093/icb/32.1.113 . JSTOR   3883742.
  24. Rendel, J. M. (1968). R. C. Lewinton (ed.). Genetic control of developmental processes. Syracuse University Press. pp. 47–68.
  25. Williams, George Christopher (2008). Adaptation and Natural Selection: A Critique of Some Current Evolutionary Thought. Princeton University Press. pp. 71–77. ISBN   978-1-4008-2010-8.
  26. Chen P, Zhang J (2023). "Transcriptomic analysis reveals the rareness of genetic assimilation of gene expression in environmental adaptations". Science Advances. 9 (39): eadi3053. doi:10.1126/sciadv.adi305. PMID   37756399.
  27. 1 2 Aubret, F.; Shine, R. (2009). "Genetic assimilation and the postcolonisation erosion of phenotypic plasticity in island Tiger snakes" (PDF). Current Biology. 19 (22): 1932–1936. doi: 10.1016/j.cub.2009.09.061 . PMID   19879141. S2CID   205091.
  28. Palmer, A. Richard (2004). "Symmetry breaking and the evolution of development". Science. 306 (5697): 828–833. Bibcode:2004Sci...306..828P. CiteSeerX   10.1.1.631.4256 . doi:10.1126/science.1103707. PMID   15514148. S2CID   32054147.