The growth of a fetus is one of the great dramas in biology, but there is another performance unfolding beneath the swelling dome of a pregnant woman’s belly. The placenta—the fleshy mattress that nourishes and supports the sac-encased fetus—undergoes striking changes of its own. During the first two trimesters, as the fetus is rapidly taking shape, the placenta is a hotbed of angiogenic activity as it races to produce new blood vessels that will support the fetus’s need for oxygen and nutrients. As the thicket of placental vessels fills out, the demand for oxygen diminishes and angiogenesis stops.
“You need a lot of pruning, a lot of modification, but you don’t need more vessels to form,” said Raghu Kalluri, HMS professor of medicine at Beth Israel Deaconess Medical Center and an affiliated faculty member of the Harvard–MIT Division of Health Sciences and Technology.
This pattern appears to be upset in women suffering from pre-eclampsia. First described in the 19th century, this oddly named illness refers to a set of life-threatening symptoms, including high blood pressure along with kidney and liver problems, that can suddenly appear late in pregnancy. In some women, it can lead to seizures, or eclampsia—a Greek word meaning “bolt from the blue”—which can be fatal. More recently, researchers have discovered that the placenta of pre-eclampsic women remains hungry for oxygen well into the third trimester, but it is not clear why this happens. They have been looking for a molecular defect that causes the placenta to remain hypoxic.
Kalluri and Keizo Kanasaki, working with the late Judah Folkman and colleagues, have identified such a flaw. They found that a protein that is normally expressed at high levels during the third trimester fails to rise in pre-eclampsic pregnant women. The researchers further found that mice lacking the ability to make the protein, 2-methoxyoestradiol (2-ME), developed all the symptoms of pre-eclampsia. These symptoms vanished when the mice were given 2-ME. Conversely, normal pregnant mice developed pre-eclampsia when given an inhibitor of a precursor of the 2-ME protein.
The findings, reported in the May 11 online Nature, appear to pull together more than a century’s worth of observations and theories. Many of these theories have focused on pre-eclampsic symptoms such as hypertension and excess protein in the urine, or proteinuria, attributing them to the release of vasoconstrictive factors and inhibitors of vascular endothelial growth factor (VEGF), which are required for heart and kidney function, respectively. The hypoxic placenta also has been ascribed to an imbalance in circulating angiogenic factors.
But the discovery that pre-eclampsia is associated with an absence of 2-ME could explain how such damaging substances are released into the maternal bloodstream in the first place. Normally, 2-ME dampens the placenta’s response to hypoxia by inhibiting the angiogenesis-promoting protein, hypoxia-inducible factor-1alpha (Hif-1alpha). When 2-ME is absent or low, Hif-1alpha is freer to go about its business. Unfettered, it continues to create new blood vessels, resulting in an unwieldy and leaky vascular web. Meanwhile, cells of the placenta, the cytotrophoblasts, respond to the chaotic situation by secreting an array of substances that seep into the maternal system through the network of vessels.
Kanasaki, an HMS research fellow in medicine, Kalluri, and colleagues have identified at least 15 such soluble factors in the circulation of the pre-eclampsic mice, including vasoconstrictors, VEGF receptors that interfere with the functioning of the kidneys, and inflammatory molecules.
“It’s this vascular dysfunction in the placenta setting in motion a cascade of events that then lead to these different systemic problems,” said Kalluri. “There’s a unifying theme here because all the pathways that other people speculated were disrupted are, in fact, disrupted. Everybody can be happy with this model because it doesn’t go against anybody’s theory.”
The findings also have potentially satisfying clinical implications. Though maternal deaths have declined in this country since the 19th century, pre-eclampsia remains a serious threat. Five in a hundred women develop its symptoms, and this proportion is probably two or three times higher in developing countries. And the treatment for pre-eclampsia—a combination of continual monitoring and bed rest—has not changed much over the past century. Administering 2-ME to these women could be a viable treatment, one that could be quite safe. “Pregnant women and babies are exposed to high levels of this protein,” said Kalluri.
Kalluri’s quest to understand the origins of pre-eclampsia was inspired, like many projects, by the fertile imagination of Judah Folkman. In 2002, the two were sitting next to each other at a meeting on the results of a phase 1 clinical trial of bevacizumab (Avastin), an angiogenesis inhibitor. “They said some patients getting Avastin were developing hypertension and protein in their urine,” said Kalluri. “So Judah looks at me and says, ‘You’re an angiogenesis researcher, but I know there’s another hat you wear. You’re a kidney researcher.’ And he said, ‘You should figure this out.’”
Bevacizumab works by inhibiting VEGF, so Kalluri was especially intrigued when he read that pregnant women with pre-eclampsia have high levels of VEGF-sopping proteins coursing through their blood. Folkman, a pediatric surgeon as well as a researcher, was familiar with the mystery of pre-eclampsia and urged Kalluri to continue investigating. Kalluri decided to block VEGF in mice. Sure enough, the animals exhibited hypertension and high levels of protein in the urine.
But the results did not get at what causes pre-eclampsia. During one of many late-night phone conversations, Folkman and Kalluri hit upon an approach. “It all came down to the commonality that the placentas are hypoxic in these women,” said Kalluri. Rather than focus on the symptoms, Kalluri set out to see what was causing the hypoxia. He began by looking for hypoxia-related genes and discovered, through a literature search, that 2-ME and its precursor, catecho-O-methyltransferase (COMT), were upregulated during the third trimester of pregnancy. Further sleuthing showed that 2-ME works by inhibiting the hypoxia-sensing, angiogenesis-promoting Hif-1alpha, leading him to wonder: “What if hypoxia doesn’t come down in the pre-eclampsic placenta because 2-ME levels are not going up?”
Kalluri and colleagues gathered serum and plasma samples from severely pre-eclampsic women and found their levels of 2-ME were unusually low. It turns out, the precursor COMT is a marker for schizophrenia, and neurologists had developed a mouse lacking COMT. “One of the greatest e-mails I ever made was to ask these people if they could give me their mice,” he said. “They said, ‘Sure you can have these mice, but they have some difficulty breeding.’” Kalluri suspected the difficulty was due to a pre-eclampsia–like condition.
His hunch was confirmed by Kanasaki, who carried out an arduous series of experiments. After mating, the mice were separated and monitored for the 21 days of pregnancy. The mice exhibited the telltale symptoms of high blood pressure and proteinuria late in their pregnancy. “Suddenly it hit,” said Kanasaki. They also exhibited shorter gestation lengths, a higher level of fetal wastage, a more disordered placental vasculature, and, significantly, low levels of 2-ME. The condition was entirely reversed by administering 2-ME, in relatively small doses.
Though the mice received the protein through subcutaneous injection, Kalluri envisions an easier delivery system for women. “The exciting thing is, it’s orally available so it can be taken as a pill,” he said. Also, women excrete 2-ME in the urine, which means they might someday be able to test their likelihood for developing pre-eclampsia. “It could be a urine dipstick,” Kalluri said.
