News Release

Finding an Alzheimer's switch

An unsuspected protein regulates the production of plaque-forming peptides

Peer-Reviewed Publication

DOE/Lawrence Berkeley National Laboratory

BERKELEY, CA -- Researchers at the Department of Energy's Lawrence Berkeley National Laboratory have discovered an unsuspected subunit of the protein complex gamma-secretase, which plays a central role in Alzheimer's disease. The researchers have shown that the newly discovered component, the protein CD147, regulates the production of the toxic peptides that cause amyloid plaques, the brain lesions that are the defining feature of Alzheimer's.

"Alzheimer's is worse than a disease -- it takes the soul of a human being," says Bing Jap of Berkeley Lab's Life Sciences Division, in whose laboratory the new component was identified. "As the population of this country ages, the incidence of Alzheimer's is increasing, at a terrible increase in cost to society. Research leading to prevention or treatment is urgent."

The discovery and role of CD147 as a subunit of gamma-secretase by Jap and his colleagues Shuxia Zhou, Hua Zhou, and Peter Walian is reported in Proceedings of the National Academy of Sciences, in an article now in the online early edition of PNAS at http://www.pnas.org/cgi/content/abstract/0502768102v1?etoc.

How Alzheimer's works:

The most persuasive hypothesis of how Alzheimer's disease invades the brain is the so-called "amyloid beta protein cascade," in which a protein called APP is clipped into shorter pieces by enzymes known as secretases. (APP stands for "amyloid precursor protein"; it is found in many tissues besides brain, but its functions are largely unknown.) If the portion of APP clipped by the beta form of secretase is further clipped by a third form, gamma secretase, the resulting fragments are amyloid beta peptides, A-beta 40 and A-beta 42. A?beta 42 in particular is toxic and causes the formation of amyloid plaques.

Unlike the majority of membrane proteins, gamma-secretase performs its proteolytic function neither inside nor outside the cell; instead, the crucial cut is made within the cell's thin membrane. In fact all the proteins and protein complexes involved -- APP and the alpha, beta, and gamma secretases -- are cell-membrane proteins, which penetrate the walls of the brain's neural cells. Alzheimer's research would greatly benefit if their structures, particularly that of gamma-secretase, could be established at high resolution by x?ray crystallography.

But membrane protein structures are particularly difficult to obtain. "Membrane protein complexes can be very difficult to purify in an intact form," says Jap. "Moreover, it's extremely difficult to get enough pure membrane protein to crystallize." Nevertheless Jap's laboratory has earned a reputation for solving the structures of important membrane proteins.

As the first step to producing enough gamma-secretase to make crystals, Jap asked postdoctoral fellow Shuxia Zhou to lead the effort to characterize the native protein complex. Zhou is a biochemist with M.D. and Ph.D. degrees from Shanghai Medical University; before coming to Berkeley Lab she studied and taught in Shanghai and at Oxford University and Kyoto University.

"Previous experiments establishing the role of gamma-secretase were genetic experiments done by causing its overexpression in cell lines and animal models," Zhou explains. "We wanted to isolate the native form and purify the whole gamma-secretase complex."

An unexpected factor:

Zhou and her colleagues isolated the native complex from cells of the HeLa line and separated its subunits by gel electrophoresis, which pulls the components apart according to their molecular weights.

"There were six strong bands in the gel, five of which we could identify because we expected to find them," says Zhou. The expected bands represented the four known subunits of gamma-secretase, the proteins named Nct (nicastrin), APH-1 (anterior pharynx defective 1), PEN-2 (presenilin enhancer protein 2), and Psn-1 (presinilin 1) -- which is cleaved into two parts in the mature complex, Psn-1 NTF (the N-terminal fragment) and Psn-1 CTF (the C?terminal fragment). In prior Alzheimer's investigations, complexes made up of just these four components were shown to be an enzymatically active form of gamma-secretase, but whether they constituted the native form was not known.

From the evidence of gel electrophoresis, apparently not: "In addition to these five bands we found an extra band," Zhou says. "We didn't know what it was." To find out, she and her colleagues clipped the band from the gel, extracted the protein, and sequenced its amino acids.

The mystery protein turned out to be the membrane protein CD147. CD147 is expressed in many tissues and has many biological functions besides its role in tumor invasion, including reproduction, inflammation, and protein transport and sorting within cells. It also has a role in neural function: when the CD147 gene is deleted in mice, the result is defective nervous system development, loss of working memory, spatial learning deficits, and disorientation -- behaviors remarkably suggestive of Alzheimer's disease.

To investigate CD147's part in the activity of gamma-secretase, the researchers used targeted RNA to silence CD147 in cell cultures. The four previously known components of the gamma-secretase complex, as well as the APP protein on which they operate, were unaffected by this silencing. But when CD147 was silenced, the production of amyloid beta peptides increased markedly.

The researchers established that the native form of gamma-secretase, incorporating CD147, appears in other cell lines, including kidney cells and neuronal cells, and is not unique to HeLa cells (which are derived from cervical cancer). CD147 itself is found in many contexts besides gamma-secretase, but only as a part of gamma-secretase does it regulate the production of A-beta peptides and thus amyloid plaques.

Goals for further research:

Just how does CD147 do what appears to be its normal job of preventing excessive production of A-beta 42 peptides, and what causes it to fail? Zhou says, "We know CD147 is a regulatory subunit of gamma-secretase, but we don't know how it works. As yet we don't know its mode of action with respect to the other members of gamma-secretase and its substrates. Determining this mode of action is a key goal of our future efforts."

About 25 amino-acid residues make up the length of CD147 that crosses through the cell membrane, one of which, glutamic acid, has net electrical charge. Such an unlikely placement for a charged residue suggests that this region of CD147 may seek to align with another protein's oppositely charged region, perhaps that of Psn-1. Disruption of this transmembrane teamwork could lead to increased production of amyloid beta peptides which, in turn, may result in the amyloid beta plaque formation that is a hallmark of Alzheimer's disease.

"The answer to how the components of gamma-secretase components fit together inside the cell membrane has to wait for high-resolution structural work," says Zhou, "and for that we first have to make enough of the native complex to make crystals."

Bing Jap adds, "Determining the atomic structure of the gamma-secretase complex, including CD147, is the next crucial step in understanding the molecular mechanisms by which the substrates are cleaved in various forms -- and the next crucial step to designing Alzheimer's disease therapeutics."

"CD147 is a regulatory subunit of the ã-secretase complex in Alzheimer's disease amyloid â-peptide production," by Shuxia Zhou, Hua Zhou, Peter J. Walian, and Bing K. Jap, appears in the online early edition of the Proceedings of the National Academy of Sciences, http://www.pnas.org/cgi/content/abstract/0502768102v1?etoc.

###

Berkeley Lab is a U.S. Department of Energy national laboratory located in Berkeley, California. It conducts unclassified scientific research and is managed by the University of California. Visit our website at: http://www.lbl.gov.


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.