By reconstructing a DNA sequence that existed more than 450 million years ago, Joe Thornton, a research scientist and evolutionary biologist at Columbia's Earth Institute, has revealed how new hormones emerged during evolution, concluding that the female hormone estrogen is the most ancient of all steroid hormones but that its role in differentiating the sexes from each other developed much later.
Thornton's study, reported Apr. 30 in the Proceedings of the National Academy of Sciences, completes a piece of a deep evolutionary puzzle by explaining the emergence of new hormones and other molecules that are parts of complex biological systems.
Steroid hormones are critical to a wide range of biological processes, but their evolutionary origins have been a mystery. They include the sex hormones estrogen, testosterone, and progesterone and the adrenal hormones cortisol and aldosterone, which regulate behavior, immunity, and the body's response to stress and changing environments. These hormones circulate in the blood and diffuse into target cells, where each binds to its own specific receptor, a protein that turns genes on and off when its hormone partner is present. Virtually all vertebrates – from humans to fish – have versions of the same six steroid hormone receptors, but invertebrates like insects and nematodes have none.
The question of how hormones and their receptors evolved poses an evolutionary chicken-or-the-egg puzzle that motivated Thornton's research. "Darwinian theory explains how a feature that improves an organism's fitness – a mutation in some essential protein, for instance – spreads through a population, but it doesn't explain how the body's proteins evolved in the first place," Thornton said. "The function of many proteins is to interact specifically with other molecules in tightly integrated regulatory systems. When the function of each part depends on the prior existence of all the other parts, the Darwinian model breaks down."
This puzzle was elegantly illustrated by the evolution of hormones and their receptors. As Thornton explained: "What's the selective pressure that drives the evolution of a new hormone if there's not already a receptor to give it a function? Conversely, how does a new receptor evolve unless there's a hormone present? I wanted to reconstruct the series of events by which new hormones and receptors emerged during vertebrate evolution to find a solution to this problem."
Thornton studied steroid receptors present in lamprey -- jawless fish that diverged from the rest of the vertebrate lineage about 450 million years ago. Using a gene identification technique called the polymerase chain reaction, Thornton established that lampreys have just three of the six hormone receptors that are found in humans and other jawed vertebrates – an estrogen receptor, a progesterone receptor, and a single corticoid receptor, but no androgen receptor.
Thornton's results indicated that lampreys represent an intermediate stage in the evolution of the endocrine system, before steroids were used to regulate the development of secondary sexual differences, as they do in other vertebrates. "Long before steroids were involved in making the sexes look strikingly different, they served the same reproductive and developmental functions in both males and females," he said, a point supported by the fact that estrogen regulates the transition to reproductive maturity in both male and female lamprey.
Thornton then used molecular phylogenetic methods to reconstruct the evolutionary relationships among the lamprey receptors and those of other dozens of other vertebrate species based on their gene sequences. These results, together with mapping data on the receptors' positions in the human genome, indicated that the full complement of hormone receptors evolved in two complete duplications of the vertebrate genome. One of these genome expansions occurred before the divergence of lampreys from other vertebrates 450 million years ago, and the other took place after that event but before the divergence of fish from the lineage that led to mammals, reptiles, birds, and amphibians, about 50 million years later.
Genome duplication provided a mechanism for evolving new receptors, but the order in which the hormones and their receptors appeared remained unknown. Using a sophisticated computational technique and a statistical model of protein evolution, Thornton back-calculated the gene sequence of the ancestral hormone receptor at the root of the steroid receptor evolutionary tree. Analysis of that sequence made its identity clear.
"It's quite convincing that the ancestral steroid receptor was an estrogen receptor of some sort," Thornton said, "because its gene sequence is extremely similar to that of the estrogen receptors found today, but it's not at all like those of the other hormone receptors." This similarity was particular striking at sites in the sequence that are known to confer the ability to recognize specific hormones and target genes.
The ancient nature of the estrogen receptor suggested a solution to the chicken-and-egg problem. Steroid hormones are produced in a common biochemical pathway in which progesterone is converted to testosterone, which is then transformed into estrogen. Thornton's results showed that the last hormone in the pathway was the first to have a receptor, a surprising result in a field whose accepted wisdom is that complex systems and structures are gradually elaborated and optimized in a step-by-step fashion.
But this same result suggested a solution to the puzzle, because it implied that less ancient steroid hormones – progesterone and testosterone in particular – had been present as biochemical intermediates before the receptors that recognize them evolved. "Once ancient organisms had estrogen and an estrogen receptor, they had to produce the other steroids in the process of making estrogen. When new receptors were created by gene duplication and then evolved affinity for these steroids, they turned what had been mere biochemical stepping-stones into bona fide hormones," Thornton said.
"The solution to the which-came-first problem is that the hormones came first, but they had weren't hormones and had no function per se, until their receptor partners emerged in the vertebrate genome," Thornton explained.
This model has practical implications for scientists studying human receptors that control important aspects of development and disease. "If this is a general dynamic, then hormones for the many orphan receptors whose hormones haven't been identified will be found among intermediates formed in the synthesis of hormones for evolutionarily related receptors," Thornton said.
Further, if estrogen is the most ancient of hormones, it must also be the most widely distributed among organisms. "Sensitivity to estrogenic pesticides and industrial chemicals in the environment may therefore be quite broad, and it's possible that the endocrine systems of a very large variety of animal taxa could be disrupted by these pollutants," Thornton said.
The two-year study, supported by the National Science Foundation, was completed seven months ago. Currently, Thornton is working to chemically synthesize the gene sequence he reconstructed, the functions of which will then be tested with techniques that molecular endocrinologists use to study present-day hormone receptors. He is also working to pinpoint exactly when during evolution the estrogen receptor first appeared.
Thornton's research spans two related areas. In addition to studying gene family evolution, he focuses on the public health and policy implications of global toxic pollution. His book, "Pandora's Poison: Chlorine, Health and a New Environmental Strategy" (MIT Press: 2000) analyzed global chemical contamination and its impact on human and wildlife health. The scientific journal Nature called it a "landmark book" and The Washington Post said it was "one of the most impressive environmental books in recent years."