The Scientist 16:25, Jan. 21, 2002
By Brendan A. Maher
The all-out assault to impede production of b-amyloid (Ab), the plaque-forming peptide believed by many to cause neurodegenerative Alzheimer's disease (AD), entails a war on two fronts. For those aiming to prevent plaques at their cellular source, the two clear targets are b-secretase and g-secretase, which sequentially cleave amyloid precursor protein (APP) to generate Ab. Some victories are emerging: Small molecules designed to inhibit g-secretase activity are being clinically tested, and the 1999 identification and cloning of b-secretase led researchers to design small molecule inhibitors.
Guilty by Association: While researchers are not yet certain of the causative agent in Alzheimer's, insoluble peptides like Ab aggregate outside the neural cells to form amyloid plaques associated with the disease.
g-secretase, which has made it farthest as a drug target, still has an identity problem. One camp of researchers, including physician Dennis J. Selkoe, professor of neurologic diseases at Harvard Medical School, has put its money on presenilin1 (PS1),1 a molecule long associated with g-secretase activity. It is mutated in one form of early onset familial AD, which accounts for 5% to 10% of Alzheimer's cases. Selkoe says an upcoming paper, scheduled for publication in the Proceedings of the National Academy of Sciences, moves closer to showing that Presenilin-1 is indeed g-secretase.
But not everyone is completely convinced. Raphael Kopan, associate professor of medicine, molecular biology and pharmacology at Washington University in St. Louis, explains, "The absolute proof will, of course, require demonstration that you can isolate presenilin1 with the substrate, play some mood music … and eventually cleavage will occur." Sangram Sisodia, director for the center of neurobiology at the University of Chicago, is even less convinced. "My guess," says Sisodia, "is that [g-secretase] is going to be several different proteases that require presenilin and nicastrin for their function." Yet, in vitro demonstration continues to elude researchers.
Targeting g-secretase activity doesn't necessarily require that the identity be known. Pharmaceutical giant Bristol-Myers Squibb (BMS) is carrying out a Phase I trial of a g-secretase inhibitor, though the company will say little more than that. What does concern researchers is safety and inhibition of other g-secretase functions. And, they have little to go on. Currently, the most prominent known substrate other than Ab is Notch. First discovered in mutant flies with notched wings, Notch is essential to cell fate determination. Like PS1 it is recessive-lethal, but little is known about the role of Notch in adulthood. Notch will be important for self-renewing tissues, predicts Kopan, though he adds, "You could probably inhibit Notch proteolysis by more than 50% and have no effect."
A BMS researcher, attending a symposium hosted by the Center for Neurodegenerative Disease Research at the University of Pennsylvania, said that so far, no deleterious effects on the notch-signaling pathway had been observed in humans. Yet, says Lawrence Goldstein, Howard Hughes Medical Institute Investigator at University of California San Diego, "There's plenty of reason to be worried about * as a target … because it looks as if it does so many other things that could be important to normal viability and health. On the other hand, the disease is so devastating we might be willing to put up with them."
New substrates are on the horizon. At the recent Society for Neuroscience meeting, a group led by Nikolaos Robakis at the Center for Neurobiology, Mount Sinai School of Medicine, New York, showed that N and E cadherins were substrates for g-secretase activity. Published evidence also points to CD442 and ErbB-43 as potential substrates. "I can guarantee that there's going to be a molecule a week from now on," says Sisodia.
Researchers may be able to curb negative side effects by increasing the selectivity for Ab. This presents a formidable challenge, however. Only two kinds of protease inhibitors are used therapeutically: HIV protease inhibitors and angiotensin-converting enzyme inhibitors. "In those cases," says Sisodia, "the enzyme was put in the genome to process one substrate." g-secretase appears to exist for more than one reason, and it's unclear whether APP cleavage is even one of them.
Amyloid Precursor Protein (APP) is first cut by the protease BACE1 to create APPs b and the membrane bound C99. g-secretase makes a second cut within the transmembrane region of C99 releasing Ab. Some believe that presenilin (PS), shown at right, is the g-secretase molecule.
Back to BACE-ics
In 1999, Mark Citron's group at Amgen discovered the beta-site APP cleaving enzyme (BACE).4 It definitively identified the b-secretase enzyme and opened the door for therapeutics. Some labs then independently cloned the gene, and by early 2001, BACE1 knockout mice were created. Researchers found no sign of Ab production in knockouts when observed in Alzheimer's model mice that overproduce APP. Also, they've yet to find a deleterious side effect due to the loss of BACE1 function. But, says Robert Vassar, a former Amgen employee now at Northwestern University Institute for Neuroscience, "We did not challenge [the mice] in any way. One could imagine a phenotype emerging when you subject them to stress, for example."
The healthy knockouts gave b-secretase a big push as a potential target for inhibition, although challenges still exist.
Pharmaceutical companies generally screen huge molecular libraries to see which inhibit Ab production. Most, if not all "hits," says Sisodia, are g-secretase inhibitors. One possible reason is that they are testing small molecules capable of crossing the blood-brain barrier; the rather open catalytic site may require larger inhibitors. Jordan Tang, program head of protein studies research at the University of Oklahoma Health Sciences Center, has worked out the crystal structure of b-secretase, or memapsin 2 as he calls it, and has designed selective inhibitory molecules5 that Tang says were whittled down to about 800 Da. The molecule will need to be small enough, yet selective enough, not to interfere with similar aspartic proteases like BACE2.
But as with g-secretase, substrates other than APP will emerge. A Japanese group recently offered evidence that BACE1 cleaves sialyltransferase, ST6Gal I, found in the golgi. The researchers suggested that inhibition might lead to modest neurologic defects.6
Philip C. Wong, associate professor, Johns Hopkins University School of Medicine, says although his group has identified BACE1 substrates in unpublished experiments, partial reduction of BACE1 activity may not affect those pathways and still reduce Ab plaque deposition. In mouse models for Ab amyloidosis with both APP and PS1 mutations, Wong's group found that plaque burden was dramatically reduced in mice carrying one normal BACE1 gene. "Our data are suggesting now that you only need to reduce BACE by 50%. And that provides the rationale that you may not have to worry about [other substrates]."
1. N.S. Halim, "Elusive b-secretase identified," The Scientist 14:6, June 26, 2000.
2. I. Okamoto et al., "Proteolytic release of CD44 intracellular domain and its role in the CD44 signaling pathway," Journal of Cell Biology, 155:755-62, Nov. 26, 2001.
3. C.-Y. Ni et al., "* secretase cleavage and nuclear localization of ErbB-4 receptor tyrosine kinase," Science, 294:2179-81, Dec 7, 2001.
4. R. Vassar et al., "b-secretase cleavage of Alzheimer's amyloid precursor protein by the transmembrane aspartic protease BACE," Science, 286:735-41, 1999.
5. L. Hong et al., "Structure of the protease domain of memapsin 2 (beta-secretase) complexed with inhibitor," Science, 290:150-3, 2000.
6. S. Kitazume et al., "Alzheimer's b-secretase, b-site amyloid precursor protein-cleaving enzyme, is responsible for cleavage secretion of a Golgi-resident sialyltransferase," Proceedings of the National Academy of Sciences, 98:13554-9 Nov. 20, 2001.
Source: The Scientist 16:25, Jan. 21, 2002