Theory under construction

Model explains how natural selection creates groups of traits that vary in unison

Marcelo CipisWhen a newspaper article depicts a newly discovered dinosaur hunting in the middle of a prehistoric forest, it’s hard to believe that the whole animal may have been reconstructed from a single tooth.  But this is often the case.  In part, this is possible because the proportions among body parts remain mostly unchanged across a wide range of organisms due to the concerted action of certain traits.  “Evolution plays with building blocks, remodeling living beings as if they were ‘Living Legos’,” says biologist Gabriel Marroig from the Biosciences Institute of the University of São Paulo (IB-USP).

His group at the Mammal Evolution Laboratory has been researching the workings of that game, by studying how different species of animals transmit these “building blocks” from one generation to the next.  But their most recent advance, which in some ways serves as a basis for other projects, did not focus on actual species: it came from theoretical simulations on a computer.  The master’s research results of biologist Diogo Melo showed that the emergence of new evolutionary blocks of grouped traits requires a little push from natural selection – a push that evolutionists call directional evolution, as described in a paper published in January 2015 in the journal PNAS.

To exemplify, Marroig mentions the stable ratios of size and shape between mandible and maxilla, respectively the bones of the lower and upper jaw that anchor the teeth of most mammals.  For an animal to obtain and chew food efficiently, these bones must be proportional.  Because their function – namely, eating – is essential for survival, variations in the size of one part will necessarily trigger changes in the other.  In other words, the lower and upper jaw bones make up a single building block, in terms of evolution.  “Unless it suddenly started raining baby food,” the researcher conjectures.  “In such case, it might be better for the animal to have a larger lower jaw in relation to its maxilla, which would allow it to effortlessly collect the food falling from the sky.”  Using the Lego analogy, evolution would have to create new blocks instead of reshaping existing ones.

Genetica_Fapespgenetica2alta copyMarcelo CipisWhimsical as it may be, the example is not far from the truth.  Like the shape of Lego blocks, which doesn’t vary much, the cranial structure of mammals is also extremely stable.  The work of Marroig and Melo shows that a strong selective pressure – like a change in the type of food available and how it can be obtained – will cause the module to be broken down and a new one to be established within a few generations.

This modularity exists because the relationship between genes and traits is rarely as simple as what is taught in school.  There is usually a direct relationship between a specific gene and a given trait.  But there can also be variations, in any direction, connecting groups of genes and blocks of traits – thus the modules.

By running the simulations for weeks at a stretch, Melo managed an unprecedented feat in the search for understanding how these blocks appear: he created a scenario in which 10,000 generations of a population were subject either to different kinds of natural selection or to no selective pressure at all.  More importantly, this theoretical evolution involved over a thousand genes, responsible for dozens of traits.  “All studies published up to now were based on two-trait systems,” says Melo.  He and Marroig decided to invest in a multi-dimensional, more life-like scenario despite the immensely greater computational effort it would require.  This was made possible by spending a quarter of the funding for Marroig’s project on a powerful server, whose use is shared with other researchers.

Genetica_Fapespgenetica3alta copyMarcelo CipisBy testing different types of natural selection, in addition to a selection-free scenario in which genes appear or disappear randomly within a population (a process known as genetic drift), the simulations showed that actual processes seen in nature can only be replicated by combining two types of natural selection: directional and stabilizing.  Directional selection favors the survival of organisms whose traits are advantageous in a changing environment – such as a protruding lower jaw when food starts raining from the sky.  The presence of this type of selection was a necessary condition for new blocks of traits to appear in the simulated populations.

After a period of directional selection, stabilizing selection entered the scene.  This second type of selection confers an advantage on organisms that preserve a given trait across generations.  What was new becomes the norm.

Although the experiment was based on simulated populations in a computer program, its conclusions mirror the empirical results Marroig obtained in previous studies, such as in his research on the evolution of body size in Neotropical monkeys (see Pesquisa FAPESP Issue No. 141), as well as current work conducted at the laboratory.

Melo’s work emphasizes the importance of an idea that usually gets little attention in evolutionary biology: epistasis, or the influence that some genes have on others.  “Epistasis is the ugly duckling of genetics and evolution, but it has started to play a central role,” says Marroig.  The concept of epistasis has only been discussed for the past 20 years, which is not enough time to give it much space in the specialized textbooks.  But according to Marroig, it explains most genetic variations found today in nature.  This makes sense: when each of a thousand genes controls a single trait, their scope of action is limited.  But if they operate through various combinations of the pieces available in the genetic repertoire, the possibilities increase considerably.  This explains how evolution can react to environmental changes after just a few generations by breaking down old building blocks and creating better-suited ones.  “Things are not as linear as biologists usually imagine,” concludes the researcher.

Modularity and its evolutionary consequences (11/14295-7); Grant mechanism: Thematic Project; Principal investigator: Gabriel Marroig (USP); Investment: R$1,006,189.94 (FAPESP).

Scientific article
MELO, D. and MARROIG, G. Directional selection can drive the evolution of modularity in complex traitsPNAS.  V. 112, No. 2, pp. 470-75.  Jan. 13, 2015.