When Cristian Cañestro set out in the early 2000s to study how animals with brains and backbones evolved, he picked a sea squirt called Oikopleura as a useful subject. Like all sea squirts, it has a tiny brain and nerve cord, but unlike the others, Oikopleura doesn’t undergo a metamorphosis on its way to maturity. Cañestro thought that Oikopleura had perhaps retained simpler, more ancestral features than other sea squirts and could be a guide to what they had evolved from.
“And that was the start of my frustration,” said Cañestro, a professor of genetics, microbiology and statistics at the University of Barcelona and a group leader at its Institute for Research on Biodiversity. His team was unable to find certain genes within Oikopleura’s genome that should have been there because they are very conserved across animals. In particular, none of the genes involved in the synthesis, modification or degradation of retinoic acid were present. Nor was the receptor for retinoic acid. Yet retinoic acid signaling was thought to be essential for making a brain, nerve cord and other vital features. Furthermore, Oikopleura also lacks a gene that seemed critical for triggering the development of heart tissue.
“If you imagine a car in your mind, of course it has wheels, right? Now, what if I told you I found a car that has no wheels?” Cañestro asked. “We found a situation in which the things we thought were essential are not there, even though the structure [they make] is still there. And that makes you rethink the essentiality of some of the genes.”
Two surprising analyses that appeared in Nature Ecology & Evolution early this year have hammered home just how inessential genes can be, and how creatively evolution can deal with losing them. By analyzing hundreds of genomes from across the animal kingdom, researchers in Spain and the United Kingdom showed that a startling degree of gene loss pervades the tree of life.
Their results suggest that even early animals had relatively complex genomes because of an unprecedented spurt of gene duplication early in life’s history. Later, as lineages of animals evolved into different phyla with distinct body plans, many of their genes began to disappear, and gene loss continued to be a major factor in evolution thereafter. In fact, the loss of genes seems to have helped many groups of organisms split away from their ancestors and triumph over new environmental challenges.
Until recently, gene losses in evolution were difficult to study because “if you don’t see something, it might be because it’s not there, but it also might be that you can’t find it,” said Günter Theißen, a plant biologist at the Friedrich Schiller University Jena in Germany. Scientists thought gene losses might be most common among symbiotic or parasitic species, which can simplify themselves by outsourcing a lot of their functional needs to their partners or hosts.
The availability of more and higher-quality genomes, however, enabled researchers to examine patterns of gene loss across the entire animal kingdom and made it clear that the phenomenon is not confined to simplified or parasitic lineages and animal groups.“There were periods in the evolution of the animal kingdom where gene loss was not going together with periods of morphological simplification,” said Jordi Paps, an evolutionary biologist at the University of Bristol who studies comparative genomics, and a co-author of one of the two big genome analyses.
Recognition that gene loss has been important to evolution throughout the animal kingdom opens new doors for research. When geneticists need to understand what genes do, they can create laboratory mice with “knockout” mutations and see whether and how the animals cope with the loss. The discovery that nature has in effect been running its own extensive knockout experiments — not just with Oikopleura but with all kinds of complex organisms — should afford rich insights into how evolution shapes development (and vice versa), the focus of a discipline known as evo-devo.
Use It or Lose It
Gene losses in evolution may sound like damaging events, since genes confer the traits that make life and health possible. It’s true that if individuals lose a genuinely essential gene, they may die or fail to flourish, and natural selection will weed them out of the population. But in reality, the majority of gene losses during evolution are likely to be neutral, with no fitness consequences for the organism, says Michael Hiller, an evolutionary genomicist at the Max Planck Institute of Molecular Cell Biology and Genetics in Dresden, Germany.
The reason is that evolutionary gene losses often occur after some change in the environment or behaviors makes a gene less necessary. If a key nutrient or vitamin suddenly becomes more available, for example, the biosynthetic pathways for making it may become dispensable, and mutations or other genetic accidents may make those pathways disappear. Losses can also occur after a chance gene duplication, when the superfluous copy degenerates, since selection no longer preserves it.
Plants offer abundant examples of this “use it or lose it” strategy, because many plant species have undergone whole genome duplications followed by waves of gene loss, explains Lydia Gramzow, a plant biologist at the Friedrich Schiller University Jena. Sometimes the duplicate copies persist for many millions of years before being lost, for reasons that Gramzow and Thießen are still investigating.
In a recent study that looked at different forms of genes in Arabidopsis plants from all over the globe, researchers in China and California found that about 66% of protein-coding genes had broken versions, known as loss-of-function variants. Surprisingly, 1% of these less functional genes were under positive evolutionary selection — that is, the plants with the missing or broken genes thrived better than those with working versions. These results validate the intriguing idea, proposed by the genetics researcher Maynard Olson of the University of Washington back in 1999, that “less is more”: Sometimes, losing a gene can be adaptive.
One of the best examples of adaptive gene loss in animals can be seen in cetaceans (the order of aquatic mammals including whales and dolphins), which have lost 85 protein-coding genes seen in other mammals, as Hiller reported last year. Many of these losses are probably neutral, but some seem linked to diving-related adaptations, like the narrowing of blood vessels during diving. One of the lost genes, KLK8, is interesting because it is involved in the development both of sweat glands in the skin and of the hippocampus in the brain; cetaceans lost it during their transition from land back to water. The loss of this gene is linked to the development of a thicker epidermis and the loss of hair (hair is not adaptive in aquatic environments, where it creates drag and does not preserve body heat as it does in terrestrial animals).
To investigate how repeatable and predictable gene loss is, Hiller’s group studied convergent gene losses in lineages of carnivorous and herbivorous mammals. Many of the gene losses involved traits that the animals no longer needed, but Hiller proposes that at least one loss was adaptive. There is a certain protein, designated PNLIPRP1, that inhibits an enzyme for digesting fat in the diet: Many groups of herbivores have independently lost the gene that codes for this protein, but dedicated carnivores retained it. In experiments, when this gene is knocked out in mice (which are omnivorous), the animals become overweight because they derive too many calories from their food. It’s possible that because herbivores need to get the most that they can out of their low-fat diets, the animals have little reason to hang on to PNLIPRP1.
Similarly convergent losses have occurred in yeasts that inhabit similar ecologies. Gregory Jedd, a senior investigator at the Temasek Life Sciences Laboratory at the National University of Singapore, became interested in Neolecta, an obscure group of organisms that have all the traits of multicellular fungi, though they are grouped with yeasts. After Jedd and his colleague Jason Stajich at the University of California, Riverside sequenced the genome of a Neolecta species, they were able to identify hundreds of ancestral genes that Neolecta and other multicellular fungi had retained but that two unicellular yeasts, budding yeast (Saccharomyces cerevisiae, well known to brewers and bakers) and fission yeast (Schizosaccharomyces pombe, used to make banana beer in Central Africa), had each separately lost.
These findings suggested that the yeasts independently evolved their unicellular way of life from a multicellular ancestor. Since many of the lost genes are involved in oxygenic metabolic reactions, the budding and fission yeasts may have each hit on the functional elimination of the same genes to thrive in oxygen-poor habitats. The convergent genetic changes might reflect optimal solutions to the yeasts’ unicellular and “facultative anaerobic” lifestyle. “This is interesting because it suggests that evolution may be more predictable and deterministic than we thought,” Jedd said.
A later, more comprehensive analysis of yeast genomes showed that gene loss is pervasive throughout the yeast phylum. As Antonis Rokas of Vanderbilt University, Chris Todd Hittinger of the Wisconsin Energy Institute and their co-authors wrote in their paper, “Our results argue that reductive evolution is a major mode of evolutionary diversification.”
Of course, the risk of evolving by jettisoning genes is that even if a gene is dispensable in particular environmental conditions, it might be needed again millions of years later, Jedd says. What then? It turns out that yeasts, at least, can sometimes get genes back.
Carla Gonçalves, a postdoctoral researcher at the University of Lisbon, works with a lineage of yeast that has lost the enzymes for alcoholic fermentation. This capability was restored, she discovered, when the yeasts acquired bacterial versions of those genes via horizontal gene transfer. In fact, she says, yeasts have lost a variety of genes involved in diverse metabolic pathways and reacquired them from multiple bacteria.
New Solutions to Old Problems
Yeasts are not alone in their metabolic virtuosity. Dolphins and whales, Old World fruit bats, and elephants — three lineages with relatively big brains — have all lost a gene, HMGCS2, required for ketogenesis, a metabolic process that scientists had thought was required to support the activity and growth of large, energy-hungry brains. Brain cells consume glucose, but when that is unavailable, they fuel themselves with ketone bodies from fatty acids. HGMCS2, the enzyme that converts fatty acids into ketone bodies, becomes especially important during fasting.
Animals without this enzyme are often starvation-sensitive: Fruit bats that have lost this gene can die after being starved for just 24 hours. Yet cetaceans and elephants can fast for much longer, “and this somehow tells us that they must have found different ways to fuel their brain during periods of starvation,” Hiller said.
In fact, the evolutionary record indicates that the loss of HMGCS2 occurred before the independent evolutionary expansion of brain size in the elephant and cetacean lineages. “In mammalian evolution, large brains evolved at least twice without having ketogenesis as a metabolic process,” Hiller said. “It shows that energy metabolism is probably more flexible than previously appreciated.”
How elephants and cetaceans feed their hungry brains without ketogenesis is still unknown, but they seem to have evolved alternative ways to address the physiological challenge. “You wouldn’t have known this is an exceptional lineage without having observed that this key gene is lost,” Hiller said.
These instances are fascinating, Jedd says, and raise the question of how those novel solutions, which presumably weren’t optimal when they first arose, came to replace the ancestral way of fueling the brain.
The different solutions to metabolic or developmental puzzles that evolution has achieved by subtracting key genes could do more than reveal new biological insights; they could inspire new biomedical interventions for human disease.
Hiller has looked at what happens to animals that lack genes whose inoperability is linked to diseases in humans. In some intriguing cases, that loss of genes isn’t known to cause disease symptoms in any other mammals. For example, when the gene for the TBX22 transcription factor is inoperative in humans, it can cause a cleft palate. Yet guinea pigs, dogs and cape golden moles do not have that gene. Studying how they develop without palate defects could be a promising direction for biomedical research.
This approach turns the usual experimental model on its head: Typically, researchers study disease mutations by introducing them into a mouse or other model organism to reproduce a disease state. But identifying evolutionary knockouts could reveal “how to not get sick despite having lost the same genes,” Hiller explained. “It is conceptually a different direction.”
More generally, the pervasiveness of gene loss in the tree of life points to an inversion of a classic theme in evolutionary developmental biology. In the 1970s and ’80s, “the big shock was to find that flies and humans use the same genes,” Cañestro said. Replace the fly Pax6 gene with the human version, and the fly can still make an eye. “Now we are finding that sometimes the structures [that grow] are the same, but the genes responsible for making the structures have many differences,” he said. “How is it possible that there are so many different genes, and still the structures are the same? That’s the inverse paradox of evo-devo.”