Alzheimer's affects one in 20 of those over 65, causing loss of memory, personality changes and, eventually, death.
A hallmark of Alzheimer's, along with diseases such as variant CJD, is the formation of "plaques" composed of clumps of protein, which are believed to damage brain cells.
Now a drug has been found to cut the production of one protein fragment that is prone to build up and stop these fragments from clumping by several separate mechanisms.
What excites neuroscientists is that the drugs, called GSMs, work in a way that has not been seen before, on the protein rather than the enzyme that makes the proteins that form the clumps.
This, says one commentator, is "an astonishing result".
The team - which number 29, from four nations - is led by Prof Todd Golde, Department of Neuroscience at Mayo Clinic in Jacksonville.
In the journal Nature, they report that agents known as gamma-secretase modulators (GSM) work to reduce production of long pieces of the amyloid beta protein (Abeta) that readily stick together and form clumps, and increase production of shorter pieces of the same protein that can inhibit the longer forms from sticking together.
Although the GSM's alter enzyme activity that produces the protein that forms the clumps, now it seems they also act on the amyloid beta itself.
"The action might be analogous to some cholesterol lowering drugs that can lower LDL, the bad cholesterol that sticks to your arteries, and can raise HDL, the good cholesterol that sweeps out LDL," says Prof Golde, who did the work with Dr Thomas Kukar.
"This broadens the notion of what drugs can do, and therefore, has wide reaching implication for future drug discovery for many different disorders," Dr Golde says.
The findings also suggest that GSMs now being tested or in development to treat Alzheimer's may prove to be valuable, the researchers say. One such drug, tarenflurbil (Flurizan), is in Phase III clinical trials, and results from the first, a 1,600-patient US study, are expected this summer.
Results of a phase II study, published online in April in Lancet Neurology, suggest it provides benefit in patients with mild Alzheimer's, Dr Golde says.
Until the Mayo Clinic study, it was not understood that tarenflurbil was a GSM and it was being tested (unsucessfully) to treat prostate cancer.
But then cell and animal studies suggested that tarenflurbil could affect protein clumps linked with Alzheimer's and reduce cognitive deficits in mice with a form of the disease.
"If results from tarenflurbil and other GSM agents are less beneficial than hoped, these findings may help drug designers create newer, more potent drugs," Dr Golde says. "Anytime we gain an increased understanding of the precise molecular action of a drug, that enhances our ability to make better drugs."
By further studies to understand how these agents work to prevent protein build up, the team can then sit down to design more potent versions.
However, Dr Kukar tells the Telegraph: "Next generation GSMs based on this research will likely take years to get in the clinic."
He adds that the find could apply to other diseases marked by protein deposits, notably "human BSE," vCJD.
Study co-authors include researchers from the Swiss Federal Institute of Technology in Lausanne, Switzerland; Harvard Medical School; the Technische Universitaet in Darmstadt, Germany, University College in Dublin, Ireland; and the University of California at San Diego.
Prof Clive Ballard, the director of research at the Alzheimer's Society, said: "This will provide important opportunities to develop clinically effective drugs.
"However, clinical trials are extremely expensive and a large increase in investment in dementia research is drastically needed to turn these exciting scientific discoveries into new treatments."