Scientists have found a way to get antibody-based therapies across a key barrier in the brain and deliver a payload of drugs that take aim at an elusive Alzheimer's target.

The researchers at Roche Holding AG's biotechnology unit Genentech said the findings from two studies, reported on Wednesday in the journal Science Translational Medicine, could open the door to new treatments for diseases like Alzheimer's, schizophrenia, Parkinson's and even autism.

"This really opens a whole new frontier for antibody therapies," said Mark Dennis, a senior antibody scientist at California-based Genentech, a company known for its antibody-based treatments for cancer.

"Before, the brain was considered off limits," Dennis said in a telephone interview.

He and colleagues discovered a reliable way of getting antibody drugs across the blood-brain barrier, a protective fortress that only allows select molecules or nutrients from the bloodstream to enter the brain.

"It's protecting the brain from toxins," Ryan Watts, associate director of neuroscience at Genentech who worked on both studies, said in a telephone interview.

Small molecules, including some pills, can cross this barrier, but large molecules, such as engineered antibodies, get stuck in the tight mesh of cells that line blood vessels in the brain.

Companies are already developing Alzheimer's drugs that use antibodies to attack the Alzheimer's-related protein beta amyloid, but the problem is that only small amounts can get into the brain.

Watts estimates less than 0.1 percent make it across. "This technology significantly improves that," he said.

The discovery came through studies of a new targeted antibody drug for Alzheimer's disease that works by blocking beta-secretase 1 or BACE, an enzyme required for chopping up amyloid beta proteins that go on to form sticky plaques in the brains of Alzheimer's patients.

Studies in mice and monkeys showed the engineered antibody effectively reduced the amount of amyloid beta in the animals' blood, but it had only a modest effect at reducing levels of the protein in the brain.

To overcome this, the team decided to use a Trojan horse approach. They knew that iron crosses easily into the brain, so they made the antibody specifically target transferrin receptors, which transport iron across the blood brain barrier.

But the large molecules kept getting stuck on the receptors and couldn't make it across the blood brain barrier.

Watts likens it to a ski lift: The molecules got on the lift, but never got off.

To overcome this, Dennis figured out a way to make the antibodies less sticky - by reducing their affinity for the transferrin receptors.

That allowed enough of the antibodies to fall off the receptors and enter the brain.

Tests in mice show the antibodies hit their target, and significantly reduced levels of amyloid beta in the brain.

Several companies, including Eli Lilly and Co, have been working to develop drugs that target BACE, which is thought to be a safer target than gamma-secretase.

Lilly pulled the plug on late-stage studies of its gamma-secretase drug semagacestat last summer because some people saw declines in brain function, and some developed a form of skin cancer.

Watts said their beta-secretase 1 antibody is very specific, and it may prove to be safer, but the big discovery for him is getting antibodies into the brain.

"We're going after this very aggressively," Watts said, adding that the company wants to study antibody treatments for neurodegenerative diseases other than Alzheimer's.

"The list is quite long," Watts said.