About two years ago, Valerie Galton, a professor at Dartmouth College, was proceeding along a straightforward line of scientific inquiry. She and her colleagues had developed a knockout mouse deficient in type 2 deiodinase (D2), an enzyme that was thought to be responsible for converting the prohormone thyroxine (T4) to the active hormone, triiodothyronine (T3). The role of type 2 deiodinase in converting thyroid hormone, which Galton helped establish, was considered "conventional wisdom," Galton says. And then the mice were born.
She expected the knockouts, with their extremely low levels of brain T3, to be quite sick. Hypothyroid animals, for example, are also deficient in T3, have learning problems, don't reproduce, and die prematurely. But these knockouts were almost normal, says Galton, a petite woman in her 70s with a gentle British accent and expressive blue eyes. The data were quite puzzling, she says: The T3 levels in the brain were the same in both knockouts and the hypothyroid animals. But her knockout mice had none of the problems afflicting the hypothyroid mice.
They predicted that knockout mice would have impaired learning, yet the mice did fine in the spatial memory test. Hypothyroid animals typically have mobility and balance problems, but the D2 knockouts could hang on to a spinning "rotarod" apparatus for as long as the wild-type mice could. Hypothyroid animals have significantly lower serum T3 levels than wild-type animals, but in D2 knockouts, T3 serum levels were normal (Endocrinology, 148:3080-8, 2007).
Naturally, Galton was initially disappointed by her results, which shook up more than a decade of her research. But now, Galton recalls her findings with an ever-widening smile: "Our theory now is that there's much more going on than we realized." "The field of thyroidology always assumed T3 was the principle hormone doing all the work," says Galton's collaborator and fellow Dartmouth professor, Donald St. Germain. But if D2 isn't available to convert T4 to T3 in these mice, why aren't the mice more damaged?
Galton is now excited to find out why these mice don't resemble hypothyroid mice. As one approach, she and St. Germain plan to look at the effects of other factors related to thyroid hormone signaling. For instance, in the type 2 deiodinase knockout mice, T4 levels in the serum and cerebral cortex were significantly higher than in wild-type or hypothyroid mice. "It is possible that T4 has some effects that were not appreciated," says St. Germain.
Since T3 levels in serum were also normal in the knockout mice, the circulating hormone might still be contributing to brain development. Antonio Bianco from Harvard University points out that the thyroid produces more T4 and T3 into circulation to compensate for the lack of D2 enzyme. "It's incredibly homeostatic," he says. There's also the possible role of a T3 transporter protein, which shuffles locally converted T3 from glial cells to neurons. Galton says this transporter could be moving T3 from serum to neurons.
St. Germain says the data indicate a need for rethinking the essential role of local production of T3 in neural development. In humans, hypothyroid levels of T4 in fetuses lead to cretinism. Current treatment is the immediate administration of T4, which presumably would get converted to T3 in the brain. But perhaps, as the knockout animals suggest, circulating T3 levels are also important to preserving neural function. St. Germain says the findings could contribute to better treatments of hypothyroidism in infants.
Galton's work was not for naught, Bianco says: She has proven that the type 2 deiodinase enzyme is indeed responsible for converting T4 to T3 in the brain, and he uses the animals in his laboratory. "Every time I want to confirm the effect is due to D2, I use the D2 knockout."