“Science is about observation and about not really accepting your dogma to the degree that it forces you to ignore what is right in front of your own eyes,” said Raphael Kopan, senior author on the Journal of Neuroscience paper. “That is how progress is made.”

No Nicastrin? No Problem
For Guojun Zhao, first author on the Kopan group paper, progress started with an unexpected, unexplainable band on a Western blot. It was the right size for the product of γ-secretase cleavage of Notch, but the cell line he was using lacked nicastrin, thought to be required for the enzyme’s activity.

In addition to Notch receptors, γ-secretase cleaves the amyloid precursor protein (APP), producing toxic Aβ fragments. The enzyme complex contains four proteins: presenilin, nicastrin, Pen2, and Aph1. Presenilin contains the catalytic site. Pen2 and Aph1 are of uncertain, but essential function. And everybody knows that nicastrin provides the substrate recognition site (see ARF related news story on AbstractShah et al., 2005).

But lately, there has been some controversy over nicastrin’s precise contribution. In 2008, researchers from Bart De Strooper’s laboratory at KU Leuven in Belgium reported that nicastrin contributes to maturation of the complex, but not its activity (Chávez-Gutiérrez et al., 2008). Zhao chose to pursue that pesky band.

He added γ-secretase inhibitors to his cells. “And, lo and behold, the band went away,” Kopan said. To further test the enzyme’s activity, Zhao transfected embryonic fibroblasts from a nicastrin-knockout mouse with a γ-secretase substrate. When he treated the cells with protease inhibitors, the cleavage product band intensified. The results suggest that γ-secretase minus nicastrin is active, but the enzyme or its cleavage product are normally degraded by the proteasome. Using protease inhibitors to reveal the cleavage product was the “key trick,” said Michael Wolfe of Harvard Medical School, who was not involved in the research. Without those inhibitors, the cleavage product would be easy to miss. The researchers confirmed their results in a second nicastrin-negative cell line, and with APP instead of Notch as a substrate. Without nicastrin, the cells were able to cleave an APP substrate and release Aβ40 into the media.

“We thought the complex required all four components, and that was that, so I was dubious,” Wolfe said. He was surprised to find the paper convincing, he told ARF. It is difficult to reconcile the current work with past papers suggesting nicastrin is crucial for substrate recognition, but Wolfe suggested nicastrin might assist in substrate recognition without being absolutely necessary. “We need to revise our thinking about how the enzyme is assembled,” Kopan said. The authors suggest that nicastrin acts to stabilize the enzyme. “It is still essential, but not for the function we thought,” Kopan said. He suggested that working with a three-part complex might simplify biochemical and structural studies. Nicastrin is the largest component of the complex, and he suggested crystallizing three proteins might be easier than four, for example.

P25 Not More Potent
Another cleavage product relevant to neurodegenerative disease is p25. Its parent is the longer p35, a membrane-bound regulatory subunit of Cdk5. Cdk5/p35 is an essential neural kinase. When Cdk5 instead hooks up with p25, a cytoplasmic proteolysis product, the two wreak havoc, phosphorylating tau, causing DNA damage, and promoting neurodegeneration. The p25 form accumulates in the brains of people with AD and, because it is not readily degraded, promotes constitutive activity of Cdk5 (Patrick et al., 1999). Cdk5/p25 has also been linked to Parkinson disease (Smith et al., 2003) and amyotrophic lateral sclerosis (Nguyen et al., 2001).

“There is a common belief in the field that Cdk5/p25 is actually hyperactive,” said John Lew, senior author on the PNAS study with first author Dylan Peterson and colleagues. A 2002 in vitro analysis by a group in Tokyo, Japan, appeared to support this theory; the researchers found differential enzyme kinetics between the p25 and p35 forms (Hashiguchi et al., 2002).

However, Peterson and Lew doubted those results. When they repeated in vitro analysis of the enzyme kinetics, collecting more extensive data, they found that p25 and p35 were statistically the same on every parameter. “This is done by direct comparison of the two enzymes, purified, with nothing else interfering,” Lew said. Kinetic parameters and phosphorylation sites, determined by nuclear magnetic resonance, matched with both histone H1, Cdk5’s best-known substrate, and tau. Why the difference from the 2002 results? “There is really no good explanation,” Lew said. “I think it is just a matter of different hands.” He noted that he cannot rule out cellular factors, not present in the in vitro experiments, that could influence p25 activity.

Researchers must now think of another reason that p25 causes trouble, Lew said. “P25 is a much more stable protein, with a longer half-life,” noted Li-Huei Tsai of MIT, but Lew doubts protein levels are the answer, and said p25 and p35 steady-state levels are similar. Localization could also make a difference: While p35 resides in the cell membrane, p25 is free in the cytoplasm, where it is likely to encounter different substrates. One important substrate, Tsai suggested, is histone deacetylase 1 (HDAC1). Cdk5/p25 inhibits HDAC1, leading to DNA breaks, cell cycle disruption, and neurodegeneration (see ARF related news story on Kim et al., 2008).

“In a way, I am not surprised at all,” Tsai said. “This happens all the time in science.” Together, these two papers show the importance of keeping an open mind. “When nature throws hints in your path, you should pay attention,” Kopan said.—Amber Dance.

References:
Zhao G, Liu Z, Ilagan MX, Kopan R. Gamma-secretase composed of PS1/Pen2/Aph1a can cleave Notch and amyloid precursor protein in the absence of nicastrin. J Neurosci. 2010 Feb 3;30(5):1648-56. Abstract

Peterson DW, Ando DM, Taketa DA, Zhou H, Dahlquist FW, Lew J. No difference in kinetics of tau or histone phosphorylation by CDK5/p25 versus CDK5/p35 in vitro. Proc Natl Acad Sci U S A. 2010 Feb 16;107(7):2884-9. Abstract