Science marches on, trumpeting its new findings, but it's not always a song that everyone is happy to hear.

The enzyme BACE1 cleaves amyloid precursor protein, and therefore has been in the crosshairs of biotech and pharmaceutical companies for some time. Two papers published in 2001 reported BACE1 knockouts were essentially normal, aside from their ability to cleave amyloid precursor protein. (See BioWorld Today, March 20, 2001.)

But two recent papers have assigned a day job to BACE1. Apparently, the protein is important for forming the myelin sheath, which insulates axons and allows high-speed neural conduction. One of the papers suggests that previous reports notwithstanding, targeting the BACE1 for the treatment of Alzheimer's disease might be a risky strategy.

In the Oct. 27, 2006, issue of Science, researchers from Thousand Oaks, Calif.-based Amgen Inc.; Ludwig-Maximilians-University in Munich, the Max-Delbrück-Center for Molecular Medicine in Berlin, and the University of Kiel, all in Germany; and K. U. Leuven in Belgium, showed that BACE1 plays a role in peripheral myelination.

And in the Nov. 12, 2006, advance online issue of Nature Neuroscience, scientists from The Cleveland Clinic in Ohio and Johns Hopkins University in Baltimore extend BACE1's role into myelination of the central nervous system, and show in behavioral experiments that BACE1 knockout mice have motor deficiencies, albeit comparatively mild ones.

"The initial message to the world was that those [BACE1 knockout] mice were healthy and everything is fine," senior author Riqiang Yan, professor of neuroscience at the Cleveland Clinic's Lerner Institute, told BioWorld Today. "But now we see that the problem is not that simple."

Myelin is composed of different cell types in the brain and spinal cord than in the peripheral nervous system. Yan and his colleagues found that both cell types were affected in BACE1 knockouts, leading to thinner myelin sheaths in both the peripheral and central nervous system.

The scientists went on to check out the molecular mechanism, and found that BACE1 knockouts do not cleave neuregulin, a regulator of brain myelination. That, in turn, led to reduced phosphorylation levels of the Akt kinase in the brain. Yan said that while his group has not collected enough peripheral tissue to date to test it, he expects to find the same mechanism in the periphery.

Amgen was more positive about the implications of the findings. Science paper co-author Martin Citron, director of research in neuroscience at Amgen, told BioWorld Today that "in my opinion, the neuregulin/hypomyelination phenotype of the BACE1 knockouts provides us with an idea about the biological function of BACE1 during development. As both publications point out, it is currently not known if this mechanism is active and important in adults. Moreover, it is not known if partial or complete pharmacological inhibition of this mechanism in the adult would impact myelination."

Robert Vassar, who now is a professor at Northwestern University and was at Amgen when he published a 2001 paper on BACE1 knockouts, agreed with Citron's assessment of the significance of BACE1's role in myelination.

Vassar, who co-authored several papers with Citron while at Amgen, praised both papers, specifically the Nature Neuroscience paper's behavioral findings. "I thought this was really an outstanding study," he told BioWorld Today, adding that as BACE1 inhibitors enter the clinic, "physicians should be aware of this - one will need to keep an eye on it" to see whether myelination problems develop.

"But," he added "I do want to point out that this is a developmental defect. These mice grow up without ever having had a BACE1 gene. That's very different from patients who have grown up, and grown old, with BACE1 protein and start taking BACE1 inhibitors in their 70s and 80s." He continues to be optimistic about the therapeutic promise of the BACE1 inhibition approach: "I would rank it up near the top," along with A-beta-immunization and ahead of gamma-secretase inhibitors.

As for why Vassar and his colleagues did not see any neural or behavioral abnormalities in their knockouts, Vassar offers two possibilities. One was that "we never looked at the electron microscope level, where the reduced myelination really comes out," he said. "We might have seen it if we had looked at the EM level."

But it also is possible that there were no such myelination defects even at the electron microscope level in the 2001 study, because Vassar's group used a different mouse model: "When I was at Amgen, when we made our knockout mouse, it was on an outbred background," which makes the mice both more robust and more similar, diversity-wise, to human populations.