By David N. Leff

For openers last year, molecular biologist Robert Vassar and his lab team at Amgen Inc., in Thousand Oaks, Calif., generated a colony of knockout mice. These animals grew up lacking the gene that encodes a protein called BACE1 - "beta-site APP-cleaving enzyme."

APP in turn stands for "amyloid precursor protein," which is the lengthy mother molecule out of which enzymes known as secretases chop the stubby but vicious amyloid beta (A-beta) peptide, which most neuroscientists blame for causing Alzheimer's disease (AD). They base their indictment on senile neuritic plaques, tiny particles that wrap around and degrade the brain neurons responsible for the memory and cognitive functions that wither away in a victim of AD. These plaques are the distinguishing diagnostic mark of the disease.

Vassar's knockouts (KOs) duly matured without making any BICE1 protein, which in turn couldn't shed any plaque-forming A-beta peptide. "First of all," he told BioWorld Today, "we found that these KO mice are normal, which is very good news for drug discovery purposes. We did extensive pathology analysis, and could find no abnormalities in any of the tissues that we measured.

"So we concluded," Vassar continued, "that we can remove the BACE1 gene and activate it, and it does not have any ill effect in the mouse. The implication is that possibly BACE1 inhibitors in human may also be well tolerated - again good news for drug development to treat AD."

Vassar is senior author of a brief communication in the March 2001 issue of Nature Neuroscience. Its title reads: "Mice deficient in BACE1, the Alzheimer's b-secretase, have normal phenotype and abolished b amyloid generation."

A similar, but separate, paper in the same journal bears the title: "BACE1 is the major b-secretase for generation of Ab peptides by neurons." Its senior author is cellular and molecular biologist Philip Wong, at the Johns Hopkins University School of Medicine in Baltimore.

Novel Drug Therapy Ploy Against AD

"That piece of evidence," Wong told BioWorld Today, "really demonstrates for the first time that BACE1 is the principal enzyme involved in cleaving APP at the beta secretase site. So we suggest that this would be a major target for AD therapy.

"Moreover," he went on, "our BACE1 knockout mice do not show any overt phenotype other than not being able to cleave the APP molecule. This suggests that if you were able to devise a compound that would inhibit this enzyme, you might not be fearful of any other side effects."

At Amgen, Vassar and his group went one KO further: "The second knockout," he recounted, "is that we crossbred our beta secretase BACE1 KO mice with transgenic animals that overexpress APP in the brain. We had to do that in order to increase the sensitivity. Endogenous mouse amyloid beta is very difficult to detect, so we needed to increase the A-beta signal. That's why we made a double-KO, the BACE1-minus gene hooked up with the excess-APP transgenics.

"Those double transgenics showed no A-beta in the brain," Vassar recalled, "and that's very good news, because it means that BACE1 is essentially the only beta secretase in the brain, and that BACE2, which is also expressed in the brain, does not contribute to beta secretase cleavage. So it clearly indicates that BACE1 is the drug target, and not BACE 2, or any other enzyme."

Vassar has left the Amgen staff to take up a faculty position April 1 at Northwestern University Medical School in Chicago. However, he continues to collaborate with the company's BACE1 knockout project, which he launched.

"What we want to go on to do," he went on, "is age these mice out. So we have the double transgenic, the BACE1 KOs mated with the APP transgenics. The transgenics on their own develop amyloid plaques in about a year.

"We analyzed the mice at 3 months of age," he related. "They were too young to have any plaques. We would predict that after a year these double transgenics are not going to show any plaques, whereas their littermates that have wild-type-based genes will.

"That would be, I think," he reiterated, "important for AD therapy, suggesting that if you can inhibit amyloid beta production in the human brain then you should be able to halt the development of amyloid plaques - the leading candidate mechanism for pathogenesis in Alzheimer's.

"These mice are approaching 6 months of age or so," Vassar noted, "so they probably have another six months to go. Thus it may be a little while before we have an answer to that basic question."

Vassar made the point that several big pharma and biotech companies are striving to develop therapeutic compounds based on inhibiting APP or its spin-off molecules. "There is evidence from Elan Pharmaceuticals," he observed, "the people who have an A-beta vaccine. Their data is suggestive that amyloid plaques can be reversible in mice. That may or may not be true for such treatment in humans.

"And I know Bristol Myers Squibb," he added, "is in clinical trials, or getting ready for them, with a gamma secretase inhibitor."

A Page From HIV's Cocktail Recipe Book

"Mechanistically," he pointed out, "beta and gamma secretases are very different proteases. The beta secretase BACE1, which is what we cloned, is an asparto protease. That's very similar to proteases like pepsin, cathepsinD and the HIV protease. These are very well known for inhibitors that have been developed.

"I would not go so far as to say our approach is going to be a cure," Vassar observed. "I think ultimately the treatment of AD will require a polypharmacy strategy - a combinatorial drug approach - similar to what we see with HIV patients these days, who receive a cocktail of a couple of protease inhibitors and AZT to inhibit replication of the virus. And in Alzheimer's we may see a similar thing. We may see the immunization approach, which Elan is working on, in combination with nonsteroidal anti-inflammatory drugs like ibuprophin. An anti-inflammatory strategy may work, along with amyloid beta inhibitors such as beta or gamma secretase blockers.

"So probably," Vassar concluded, "we'll see a cocktail approach maybe in 10 years - hopefully sooner." n