THE idea that evolution is an essentially random process has been dealt another blow. When related species independently evolve similar physical traits they tend to use the same genes to do so.
Nature often comes up with similar solutions for a particular problem, a phenomenon known as convergence. For instance, different species of butterflies in the genus Heliconius have evolved similar wing patterns to warn potential predators that they are poisonous, as have the famous monarch and viceroy butterflies. 鈥淏ut we don鈥檛 know whether similar butterfly patterns have evolved with changes in the same genes or by different means,鈥� says Sean Carroll of the University of Wisconsin-Madison.
Now separate studies by Carroll鈥檚 team and a group led by David Stern of Princeton University show that evolution mostly relies on the same genes to arrive at similar traits in related species. 鈥淚t鈥檚 really the first peek inside the mechanisms of convergence,鈥� says Carroll.
Advertisement
His team studied fruit flies that have independently evolved pigmented patterns on their abdominal segments. For instance, each segment in the species Drosophila melanogaster has a dark band, and two terminal segments in males are darker than in females.
When the researchers studied 13 species of fruit flies they found that a gene called bab2 controlled pigmentation in nine of them: the higher the levels of expression of bab2, the fainter the pigmentation. 鈥淭he same gene is involved in a whole bunch of diverse patterns across the whole Drosophila group,鈥� says Carroll. Evolution, it seems, arrived at the same genetic solution several times (Nature, vol 424, p 931).
But in two species of fruit fly, D. santomea and D. serrata, the bab2 gene is not correlated with pigmentation, Carroll鈥檚 team found. Instead it is correlated with patterns of hair-like projections from individual cells called trichomes. It鈥檚 as if bab2 had already been allocated to a particular function, so evolution had to come up with another way to produce the pigmentation 鈥� one that researchers have yet to find. 鈥淭here鈥檚 more than one way to skin the cat,鈥� says Carroll. 鈥淏ut there are biases that have evolution driving through the same genes again and again.鈥�
Further evidence of a similar effect is provided by the larvae of nine different Drosophila species that have trichomes on their backs. Some of them have patches of cells with no trichomes, making those areas look bald, and there is an enormous variation in the patterns of trichomes, giving rise to everything from hairy larvae to mostly bald ones. The common ancestor of these species did not have these traits, so they must have evolved independently in each one. Stern and his colleagues found that the bald patches were directly correlated with a gene called shavenbaby. Cells that expressed this gene had no trichomes (Nature, vol 424, p 935). 鈥淚t鈥檚 a very, very precise match across nine species,鈥� says Stern.
The findings are very significant, says Michael Richardson of Leiden University in the Netherlands. 鈥淚t means that evolution isn鈥檛 a shotgun approach, where any old gene can be changed,鈥� he says. 鈥淪omething about the way in which an animal is put together limits its options. It supports the whole idea of developmental constraints. In other words, there are only a certain number of pathways that evolution can go down.鈥�
But Richardson cautions that while the same genes may be correlated with the same traits across species, it may not be the same mutations that are responsible. It is possible that different mutations in a number of genes could have led to the expression, say, of the shavenbaby gene in a cell, and that such genes represent hot spots in a genetic network that are expressed time and again in evolution.
The studies also suggest that researchers have to be careful when drawing up evolutionary trees for species. 鈥淛ust because you observed the same genes making the same morphology doesn鈥檛 mean they evolved from a common ancestor,鈥� says Stern.