Fluid filled. Compared with a healthy human fetus (left), a fetus with hydrocephalus cannot drain the fluid it needs to from its brain.

J. P. McAllister, Science Translational Medicine

Fat Molecule May Cause 'Water on the Brain'

Many babies born prematurely suffer from bleeding in their still-developing brains. Even when the bleeding stops, another life-threatening condition can strike: hydrocephalus, which occurs when fluid produced to keep the brain healthy builds up because it can't properly drain. For decades, doctors have known that the bleeding and hydrocephalus, also called "water on the brain," were linked, but they weren't sure why. A new study suggests the answer lies in a lipid that's common in blood but that can also profoundly disrupt brain structure and function when it's present in large quantities.

Hydrocephalus strikes about one in 1500 babies, and treatment is imperfect. Doctors usually implant a shunt to drain cerebrospinal fluid out of the brain and into the spinal cord. Shunts fail over time, however, and follow-up surgeries are sometimes needed. The condition itself can also cause lifelong neurological problems. The roots of hydrocephalus remain murky, but for those linked to brain bleeds, the hypothesis was that blood clots—necessary to stop the bleeding—blocked the razor-thin pathways through which cerebrospinal fluid must travel to exit the brain. "We assumed for 100 years that it was just a mechanical block," says James McAllister II, a neuroembryologist at the University of Utah School of Medicine in Salt Lake City, who wasn't involved in the recent work. "Everybody thought that you dammed up the narrow channels."

A group based at The Scripps Research Institute in San Diego, California, recently began to suspect that something else was at work. For years, Scripps neuroscientist Jerold Chun had been studying the embryonic brain and how certain lipids in the blood of both the mother and the embryo affect its development. He and graduate student Yun Yung developed a way to inject one of these lipids, lysophosphatidic acid (LPA), into the ventricles of a fetal mouse brain. When they did, they saw something startling: Every embryo that got the injection was born with hydrocephalus.

Researchers had experimented before with injecting blood into animal brains, to mimic a hemorrhage, and had indeed observed the onset of hydrocephalus. But they hadn't isolated any specific molecular components of blood as the key. To determine whether LPA was it, Chun and his colleagues repeated their mouse brain injections with only red blood cells. (LPA is in the plasma component of blood.) This didn't cause hydrocephalus, but injecting plasma did in some animals. And genetically removing a specific LPA receptor prevented blood from inducing the condition.

In a final experiment, the group gave the animals a compound that prevented LPA from binding to its receptor on mouse brain cells, right before they were injected with hydrocephalus-inducing LPA. Those mice stayed healthy, the researchers report today in Science Translational Medicine.

Why is LPA having this effect? "It's almost like a drug overdose," Chun says. LPA receptors are all over neural progenitor cells, which go on to form neurons and other types of cells in a young brain. When a flood of LPA hits them, "it causes them to do things they wouldn't do." This results in shifts in the structure of the brain as it forms and displacement of cells from where they're supposed to be, including those that line the ventricles and help control fluid flow.

Hydrocephalus researchers, who have long been frustrated at the field's slow progress, are enthusiastic. "It really blew me away," McAllister says of the work. "This is the first time we have recognized that there are bad things in the blood that can affect the cells" and cause hydrocephalus.

Although the work was done only in mice and needs to be confirmed, there's hope that it will be. The LPA receptor is expressed in the brain of human fetuses, just as in mice, and in the same types of neural progenitor cells. "That bodes well," says Pat Levitt, a neuroscientist at the University of Southern California in Los Angeles. "Is a finding like this going to be related to the human condition? ... I think in a case like this, it's fair to say yes."

Equally exciting is that the discovery offers a clear path to therapy. Although there's no LPA receptor blocker on the market right now, the proteins fall into a class that's commonly targeted by drugs. In theory, a compound like this might be given to premature babies from birth in an effort to prevent hydrocephalus should a brain bleed occur, or to pregnant women whose fetuses are at risk of hydrocephalus because of bleeding. Chun's group has started studying cerebrospinal fluid and other human samples, to see what LPA is doing there.