From Alaska down to the Baja Peninsula, the rocky tide pools of North America’s West Coast are separated by hundreds of kilometers of sandy beaches. Inside those tide pools live Tigriopus californicus copepods, small shrimp-like animals that evolutionary biologist Ron Burton has been studying since he was an undergraduate at Stanford University in the 1970s. During those early days of DNA technology, Burton became curious how the genomes of the isolated copepod populations compared.
While still at Stanford, Burton sequenced the mitochondrial gene cytochrome c oxidase subunit one, the standard marker people used at the time for species identification, and discovered that the copepod populations were strongly differentiated: on average, there was a 20 percent sequence divergence in this gene between populations. When he crossed Santa Cruz copepods with animals from San Diego, the hybrids did fine, but when he bred them to one another, their offspring did not do well, taking longer to develop, producing fewer offspring, and having lower survival. “That was the first indication that there was some sort of genetic incompatibility developing between these isolated populations,” says Burton, now a professor at the Scripps Institution of Oceanography in San Diego.
The big change that’s occurred is just the way that people look at mitochondrial variation.—Justin Havird, University of Texas at Austin
The first-generation hybrids have a full set of nuclear genes from each parent, but among their progeny, genetic recombination has mixed and matched the parental genomes. This potentially creates mismatches between the nuclear and mitochondrial genomes that affect fitness. When Burton mated the second-generation female hybrids with male copepods from the parental Santa Cruz or San Diego populations, he found that the direction of the backcross made a difference. Breeding the second generation of hybrids with members of the paternal population produced no improvement in fitness. But a backcross to the maternal population produced offspring that were just as fit as the original natural populations. This happened regardless of whether the maternal population was from the north or from the south. (See illustration on page 46.)
“If you go to the maternal line, you’re crossing to the population that matches the mitochondria, because mitochondria are maternally inherited,” Burton explained. “That was pretty strong evidence that there was something going on between the mitochondrial and nuclear genes.”
Mitochondria are ancient endosymbionts that over time lost many of their genes—some of which migrated to the nuclear genome—and increasingly came to rely on the nuclear genes to supply the basic raw materials necessary for mitochondrial function. In most bilaterian animals, the mitochondria’s pared-down genome contains only 37 genes, 13 of which code for proteins, with the rest encoding various RNAs—all of which play roles in mitochondrial function. “For many years it was thought that all of the variation that we see in mitochondrial genomes had to be neutral because these genes are so important that any mutations that would have affected these functions would have been screened out,” says Justin Havird, a biologist at the University of Texas at Austin.
In fact, that’s exactly why mitochondrial genes are so commonly used to assess the genetics of populations: quantifying neutral variation serves as a molecular clock that allows researchers to estimate how long it has been since the populations diverged. When it came to genome analyses seeking to identify the genetic basis of adaptive change, mitochondrial genes were often ignored.
TWO GENOMES, ONE CELL
For 1.5 billion years, the mitochondrial and nuclear genomes have been coevolving. Over this time, the mitochondrial genome became reduced, retaining only 37 genes in most animal species, and growing reliant on the nuclear genome to fulfill the organelle’s primary function—to produce ATP by oxidative phosphorylation. Mitochondrial gene products interact with those encoded in nuclear genes, and sometimes with the nuclear genome itself. Because the mitochondrial genome mutates faster than the nuclear genome, it takes the lead in the mitonuclear evolutionary dance, while the nuclear genome follows, evolving compensatory mutations to maintain coadapted gene complexes. Researchers are now coming to appreciate that this has consequences for physiology and even macroevolution.
Researchers have long known that many proteins are made of several components, some of which are coded for in the mitochondrial genome, and others being coded for in the nuclear genome. Cytochrome oxidase, the last enzyme in the respiratory electron transport chain, is one example.
Mitochondria require nuclear gene products to continually produce energy for the cell. For example, mitochondrial protein translation requires aminoacyl tRNA synthetases (aaRS) encoded by the nuclear genome to attach amino acids to the corresponding tRNAs encoded by the mitochondrial genome.
Nuclear gene expression
Mitochondrial gene products can influence the expression of nuclear genes, though the mechanisms are as yet unclear.
Over the past 20 years, however, researchers have begun to document the effects of variation in the mitochondrial genome on physiological functions such as growth rate and reproductive success in flies, copepods, and various fish species. Last year, for example, Brown University evolutionary biologist David Rand and colleagues found in fruit flies that the mitochondrial genome regulates the expression of hundreds or even thousands of nuclear genes, including genes that aren’t related to mitochondrial function. “On a per-nucleotide basis, the small mitochondrial genome is disproportionately influential given its very small size,” Rand says. “I think now people recognize that there’s lots of functionally important stuff happening among different mitochondrial DNA. Crosstalk between the two genomes is probably an important part of a lot of physiology.”
The evolution of the mitochondrial genome might also matter, the field has come to realize, for adaptation and speciation. “The big change that’s occurred is just the way that people look at mitochondrial variation,” says Havird.
A mitonuclear species concept
Mitochondria replicate their genomes more than once per cell cycle, and they do so in an environment full of DNA-damaging free radicals produced as a byproduct of the metabolic process that generates ATP within the organelles. These factors contribute to the rapid mutation rate of mitochondrial DNA, which greatly exceeds that of the nuclear genome. In the Tigriopus copepods, for instance, Burton has discovered that the mitochondrial genome evolves 50 times faster than the nuclear genome.
Moreover, mitochondrial genomes don’t undergo recombination, so if a mutation arises in the mitochondrial genome, it’s hard to get rid of it. Because mitochondrial gene products often interact with nuclear-encoded RNAs and proteins, such a mutation exerts a selective pressure for a compensatory mutation in the nuclear genome to keep the cell functioning properly. With mutations accumulating in both the mitochondrial and nuclear genomes, isolated populations drift apart, genetically speaking. Eventually, the mitochondrial genome of one population could become incompatible with the nuclear genome of the other and lead to reproductive isolation.
Ornithologist Geoff Hill at Auburn University in Alabama says he believes this process could explain speciation, and he has formalized this idea into the “mitonuclear species concept.”3 In his view, a species is defined by a set of coadapted mitonuclear genotypes, “so if you mate outside a species boundary, you’re creating bad combinations of nuclear and mitochondrial genes that lead to poor mitochondrial function,” he explains. For example, the closely related species of the golden-winged warbler and the blue-winged warbler, small songbirds that live in southeastern and south-central Canada and in the Appalachian Mountains, share 99.9 percent of their nuclear genome, but they have unique mitochondrial genotypes.
Although the mitonuclear species concept is intriguing, it is far from universally accepted. Some consider Burton’s copepods to be the poster child for the idea, but Burton himself says it’s unclear at this point how important mitonuclear incompatibility is for speciation. “I think it can play a role, but I don’t know that it is the organizing principle behind speciation,” Burton says. The tight population structure of the copepods, with little gene flow between populations, is a key for creating the conditions for mitonuclear coadaptation, Burton says, and most species don’t have such structured populations.
Burton is working to identify exactly which genes are responsible for the problematic mitonuclear interactions in the copepods from different tide pools. “We don’t know if there’s 10 different gene interactions that are responsible, or just one,” he says. To explore this question, he is carrying out hybrid crosses that allow him to map the chromosomal locations causing problems. He then compares the incompatibilities arising between two pairs of closely related (recently diverged) populations with those between more distantly related (anciently diverged) populations to find out whether the same incompatibilities tend to arise repeatedly or if they are random and unpredictable. So far, preliminary findings point to the latter, Burton says.
Havird is intrigued by the mitonuclear species concept but says the jury is still out on how prevalent the phenomenon is. “In some cases, mitonuclear coevolution leads to reproductive isolation, and in other cases, mitochondrial genomes just seem to completely ignore species boundaries,” Havird says. For instance, studies from the University of British Columbia found that Atlantic killifish (Fundulus heteroclitus) are differentiated into northern and southern populations whose mitochondria are functionally divergent in ways that are clearly adapted to different water temperatures. In spite of this, there does not appear to be evidence of mitonuclear incompatibilities between northern and southern populations—studies show the two populations can mate without any untoward consequences.
One of the challenges in studying mitonuclear interactions in the context of speciation is that it’s difficult to establish a causal relationship: it is possible that mitonuclear mismatches could arise due to drift after a speciation event, rather than drive the reproductive isolation in the first place. These are questions that Michi Tobler, a biologist at Kansas State University, is trying to answer. His study system of choice: the fishes of the family Poeciliidae, several lineages of which have adapted to live in hydrogen sulfide–rich springs in Mexico.
CONSEQUENCES OF MITONUCLEAR INTERACTIONS
The intimate relationship between the mitochondrial and nuclear genomes comes into play as populations evolve. For example, the relatively fast mutation rate of mitochondrial DNA (mtDNA) means that the nuclear genome (nDNA) has had to evolve compensatory mutations to keep pace and maintain collaborative functionality. This process causes populations to drift apart due to mitonuclear incompatibilities.
Copepods on the Pacific coast of North America are the best-known example of this phenomenon. Researchers have successfully bred animals from different tide pools, and while the first-generation hybrids do fine, second-generation individuals develop slower and have fewer offspring.
When F2 hybrids are backcrossed to the paternal line, they show no improvement in fitness. When they are backcrossed to their maternal line, however, their fitness is rescued, most likely because the backcross in this direction reintroduces the nuclear genome to the mitochondrial background it is co-adapted with.
F2 hybrid females crossed with paternal line, where mitochondria types do not match, leads to no fitness improvement.
F2 hybrid females crossed with maternal line, which carries the same mitochondrial type, improves fitness.
Hydrogen sulfide is extremely toxic because it binds to and blocks cytochrome c oxidase, thus interrupting mitochondrial function. Tobler wanted to understand how the fish could survive in these conditions, and to figure out whether mitochondrial adaptations to sulfidic environments could cause mitonuclear incompatibilities that would promote speciation. He has discovered that in some lineages, one of the mitochondrial proteins in the cytochrome c oxidase enzyme has evolved a shape that cannot be inhibited by hydrogen sulfide. The modified protein, along with evolved changes in gene expression, allows the fish to survive the toxic environment. This should, in theory, exert selective pressure for compensatory changes in the nuclear components of cytochrome c oxidase—something he is currently trying to suss out across lineages that have been separated from their cousins in non-sulfidic waters for various lengths of time.
This is an ideal study system, he says, because “we basically get snapshots at different time points during speciation,” revealing clues as to how the animals evolved.
The fish adapting to sulfidic springs are a clear example of mutations in the mitochondrial genome leading to changes in the physiology of the evolving lineages. But emerging evidence suggests that mitochondrial variation can influence the physiology of males and females differently. That’s because mitochondria are maternally inherited, meaning that mutations that are harmful to males but beneficial or neutral for females could, in theory, accumulate in mitochondrial DNA.
Researchers first proposed this idea, dubbed “mother’s curse,” in 1996, but it’s only in the last five years or so that researchers have begun collecting data to investigate the question empirically. The logic of this argument seems obvious: “It’s inevitable that the mitochondrial genome will accumulate mutations that are exclusively male-harming,” says Damian Dowling, an experimental evolutionary biologist at Monash University in Australia. But empirical tests have been equivocal, with Dowling’s group finding some evidence for it under some circumstances, while Rand’s work has yielded no support whatsoever.
For example, in a recent study, Rand’s team generated 72 experimental pairs of mitonuclear genotypes in fruit flies to test for mitonuclear gene interactions, or epistasis. When these flies were then exposed to different dietary and oxygen environments, the researchers found evidence of many mitonuclear gene interactions—and these interactions changed depending on the dietary or oxygen treatments the fruit flies were exposed to. But they found no evidence that males were worse off, as would be predicted by the mother’s curse hypothesis.
In another recent study, looking at the transcriptional responses to hypoxia in Drosophila, Rand found that females were generally more sensitive to lack of oxygen than males—a result that squarely contradicts mother’s curse predictions. “Several years ago, I had several papers showing some support for [mother’s curse],” he says, but the new results have forced him “to revise my enthusiasm for the idea. I’d rather have allegiance to my data.”
Moreover, from a physiological perspective, it’s unclear whether mitochondrial mutations with different effects in males and females can arise and be maintained. A priori, one would not expect the fundamental cellular processes encoded in the mitochondrial genome to work differently in males and females. “What is the real capacity for [sexually antagonistic mutations] when the mitochondria is this really core organelle for physiology?” says evolutionary physiologist Kristi Montooth of the University of Nebraska. “You expect that there wouldn’t be a lot of physiological capacity [for mitochondrial mutations that harm males but not females], and so even if we can make this evolutionary argument [for mother’s curse], how pervasive do we really expect the phenomena to be?”
Like Rand, Montooth has found that sometimes the deleterious outcomes are stronger in females, counter to the predictions of mother’s curse. But she is also uncovering evidence suggesting that the physiology of males and females may be more different than previously realized. At the Society for Integrative and Comparative Biology meeting in Tampa, Florida, last January, Montooth presented evidence that mitochondrial variation can have sex-specific effects when fruit flies are reared under long daylight cycles that cause an increase in metabolic rate.
To investigate this further, Montooth and her collaborator Elizabeth Rideout at the University of British Columbia are exploring to what extent metabolism can differ between male and female flies by introducing sex markers to tell larval males from females and measuring their metabolic activity. “It may be that males and females are just really from the get-go using metabolites and storing metabolites differently,” Montooth says. “Even though the process of larval development and metamorphosis seems sexually uniform, it may actually not be.”
Notwithstanding whether variation in the mitochondrial genome can have different physiological consequences in males and females, these and other experiments continue to unveil the organelle’s substantial coordination with the organism’s nuclear genome. “To me” says Rand, “this whole mitonuclear interaction is the really exciting stuff because that tells a story of a billion years of evolution, and how these two genomes make the cell and the organism work.”
Viviane Callier is a freelance science writer based in San Antonio, Texas.
Correction (November 5): The illustration in this story has been updated to correctly label the red copepods as coming from San Diego. The Scientist regrets the error.