by Nancy Fliesler

It's the scourge of daycare, preschools, and playgroups in the U.S., causing more than half a million annual doctor visits, parent absences from work and tens of thousands of hospitalizations. In the developing world, it's a killer to be feared, responsible for over 400,000 deaths each year.

It's called rotavirus, and it's the leading worldwide cause of severe, dehydrating gastroenteritis in children under age 5, eventually striking nearly all infants and toddlers. In the U.S., almost all children recover with rehydration therapy to replace lost fluids and electrolytes, but in developing countries, where access to care can be limited, rotavirus causes up to 6 percent of all deaths in young children.

Phil Dormitzer, MD, PhD

Despite being such a well-known foe, scientists never knew how rotavirus broke into cells to wreak its havoc until a Children's Hospital Boston researcher took a much closer look.

In the 1980s, Phil Dormitzer, MD, PhD, now a physician and a structural virologist in Children's Laboratory of Molecular Medicine, was an anthropology student doing fieldwork. During his travels on the Indian subcontinent, he had several episodes of diarrhea, some severe, and later learned in one of his classes that diarrhea is a major cause of global mortality.

"I was an idealistic kid," Dormitzer recalls, "and it struck me as stupid that in the late 20th century, hundreds of thousands of children should be dying of diarrhea. I saw solving the problem as a real opportunity to make an impact on world health."

He decided to go to medical school, and at Stanford University met his mentor Harry Greenberg, MD, who was studying rotavirus. "If you want to make a vaccine against diarrhea," Greenberg told Dormitzer, "that's what my lab does."

Vaccines are especially attractive in developing countries where the health infrastructure is weak. Rehydration therapy, though simple and effective, requires a health system that can deliver timely treatment and health education, whereas with a vaccine, medical workers can sweep through a community and vaccinate everyone in a military-style operation. Dormitzer began working on rotavirus vaccines with Greenberg in 1987.

ìWe aim to use this knowledge to design a cheap, safe and effective vaccine.î

Greenberg was trying to develop a vaccine based on individual rotavirus proteins. In theory, such vaccines do the same job as vaccines made from whole viruses, but are easier to manufacture, more likely to be stable without refrigeration (a boon in developing countries) and less likely to cause side effects.

Dormitzer's first task was to isolate and manufacture a single protein on rotavirus's surface, called VP7, and to try to immunize animals with it. It didn't work, Dormitzer discovered, because VP7 changes its shape when rotavirus breaks into a cell. The animals' immune systems weren't "seeing"óor making antibodies againstóthe right form of the protein.

Dormitzer realized that to build an effective vaccine, he needed to get a structural understanding of VP7 so he could see how its parts moved and reoriented during the process of infection. He decided to relocate to Boston, where Stephen Harrison, PhD, and the late Don Wiley, PhD, were studying viruses at the newly formed Children's Hospital BostonñHarvard Medical School structural biology unit. Now called the Laboratory of Molecular Medicine, the unit was, and still is, supported by the prestigious Howard Hughes Medical Institute. "I came here to learn how to work on the rotavirus problem structurally," Dormitzer says.

Rotavirus is a large, 20-sided, soccerball-shaped virus. Its surface layer, stripped off in the course of entry, somehow gets the virus's inner "payload"óthe genes and the replication machineryópast the intestinal cell's outer membrane to the inside, where it can start reproducing.

Projecting from the surface of rotavirus are 60 "spikes," made up of another protein known as VP4. Dormitzer has lived and breathed VP4 since arriving at Children's in 1997, and he's now shown that the VP4 spikes are what enable rotavirus to enter intestinal cells, through an elaborate molecular gymnastics and two consecutive shape changes. In the August 26, 2004, edition of the journal Nature, Dormitzer published high-resolution images that give evidence of these maneuvers (see sidebar, below).

First, when rotavirus arrives in the intestines, digestive enzymes cause some of the VP4 molecules to rigidify into their spike form. This positions the spike's "head" to attach to an intestinal cell's outer membrane. Second, the spikes bend back, and VP4 takes on a folded-umbrella structure. This folding motion, Dormitzer believes, causes the spike's "body" to punch through the cell membrane, allowing the virus free passage inside.

VP4's head and body proteins contain many of the targets that prime the immune system to attack rotavirus. As a first step in revealing their structureóknowledge that will help in building an optimal vaccineóDormitzer and colleagues manipulated bacteria and insect cells into making large quantities of the proteins in their different configurations. They then crystallized the proteins in a process akin to making rock candy: you put the protein in a liquid solution, tinker to optimize the conditions, then waitósometimes for yearsófor crystals to form.

The researchers then brought the crystals to be X-rayed at a particle accelerator, of which only a handful exist in the U.S. As electrons whiz around a track about 1 kilometer (or 0.6 mile) long, almost at the speed of light, intense X-rays are given off and sent down pipes into adjoining rooms, where investigators from around the world plug in and do their experiments. "You don't sleep very much," Dormitzer says. "Beam time is very valuable."

The crystals scatter the X-ray beams in different directions, making spots on a detector. Back in Boston, these spots were analyzed to build images of the VP4 spikes and their component parts, precise down to the atom. Dormitzer's team then fitted these to reconstructed images of the spikes based on direct electron microscopy of rotavirus particles. This gave a more fleshed-out picture of the head and body components.

The key for vaccine development is that these components not only trigger the immune system, but appear to be stable at room temperature and, Dormitzer believes, could be relatively easy and cheap to mass-produce. Children's has applied for patents on each, and Dormitzer is now collaborating with several other institutions to fashion them into a vaccine. The VP7 protein may also be used if Dormitzer can get it to hold a stable shape that the immune system will recognize. "Ultimately, we'll make a cocktail of these different components," he predicts.

Other vaccines are in the pipeline, but Dormitzer and Harrison believe their approachóengineering vaccines to have exactly the characteristics needed, based on an intimate knowledge of a virus's structureóis the wave of the future.

"We've gained important new insights into how this virus attacks cells," says Dormitzer. "We aim to use this knowledge to design a cheap, safe and effective vaccine to keep children from dying of an illness that should be readily preventable."