Long interested in the role of presenilins (PS) in calcium signaling, co-corresponding author Kim Green discovered that these membrane proteins activate the sarco-endoplasmic reticulum Ca²⁺-ATPase (SERCA) pump (ARF related news story on Green et al., 2008). Shortly thereafter, Green heard a talk by Ralph Nixon, Nathan Kline Institute for Psychiatric Research, New York, on how PS mutations disrupt lysosomal function in AD patients. That got Green’s mind off calcium long enough to run a few initial experiments hinting at whether wild-type presenilins might also play a role in regulating protein degradation pathways. He and graduate student Kara Neely looked at expression of autophagy-related proteins in embryonic fibroblasts from PS1-deficient, PS2-deficient, and PS double-knockout mice. Comparing them to wild-type mouse cells, “we were shocked to see such huge differences in autophagic markers, particularly LC3-II (a marker for autophagosomes),” Green said. “Protein degradation is so important for aging, and for all these different diseases, that it was something we felt we needed to understand—how presenilins played a role in this.”

That realization launched first author Neely on a series of additional experiments to address the role of PS in autophagy. She looked at autophagy markers in PS-deficient fibroblasts and in human neuroblastoma cells treated with small-interfering RNAs to knock down PS1 or PS2. Her Western blots showed changes, relative to wild-type cells, in several autophagy-related proteins—upregulated LC3-II, reduced levels of phosphorylated mTOR (an autophagy inhibitor) and faster-migrating LAMP2a (a lysosomal protein). Backing the biochemical data, PS double-knockout fibroblasts had more autophagosomes than wild-type cells. Interestingly, the changes seemed independent of PS’s catalytic activity, as treatment with two different γ-secretase inhibitors did not affect levels of key autophagic markers.

Philipp Jaeger, who works with Tony Wyss-Coray at Stanford University in Palo Alto, California, called the study “an exciting step forward”—in part because “presenilins, mainly known for their involvement in amyloid-β pathology of AD now appear onstage again in a γ-secretase-independent pathway.” (See full comment below).

As fate would have it, the current paper was submitted for publication just days before Nixon’s group reported similar findings in Cell (ARF related news story on Lee et al., 2010). Whereas the Irvine researchers studied both presenilins, Nixon and colleagues focused on PS1, identifying its role in glycosylation of a proton pump required for lysosome acidification. For the present paper, reviewers asked Green and colleagues to also examine lysosome acidity in their PS-deficient cells. The scientists stained double-knockout fibroblasts with red LysoTracker dye that enters acidic compartments and fluoresces more in lower pH. The authors found more red-staining puncta per cell in the PS-null cells compared with controls, and took that to indicate “no deficits in lysosomal acidification, despite robust impairments in autophagic degradation,” they wrote.

These findings contrast with Nixon’s study, which found less LysoTracker fluorescence in PS1-null cells, relative to wild-type. (The New York scientists also confirmed the acidification defect with several other methods, including direct pH measurement using a radiometric lysosensor.) The discrepancy between the two studies might stem from the different cell types used (i.e., PS double-knockouts by the Irvine group, PS1-deficient cells by the New York researchers). Nixon’s lab has unpublished data suggesting that PS2 may have “a completely different impact on cells than PS1,” and that the two proteins could compensate for each other in some regard, complicating the final readout, Nixon told ARF. “When you combine the deletions of both in the same cells, you end up with a hybrid where certain aspects of the phenotype are rescued by the PS2 deletion, and others are retained. Both deletions aren’t doing the same thing. Trying to sort out the roles of PS1 or PS2 alone in the double-knockout is just not possible, in our opinion, without confirming what you’re showing in the single-knockout,” Nixon added.

From an overall standpoint, however, the Cell paper and current study “agree almost completely about the basic idea that PS has a fundamental role in autophagy,” Nixon said. “The question then becomes whether there are purely deficits in clearance and lysosomal proteolysis, or whether there are also alterations at the front end of the pathway, i.e., induction,” he said (see Nixon and Yang, 2011, review).

This is where recent work by Wyss-Coray’s lab may come into play. The Stanford researchers found that reduction of the early autophagy protein beclin-1 worsened Aβ deposition and synaptic defects in AD transgenic mice. They also found reduced beclin-1 levels in postmortem brain tissue from AD patients (ARF related news story on Pickford et al., 2008). More recently, Jaeger and colleagues fleshed out this link with cell biology experiments showing that beclin-1 regulates turnover of amyloid precursor protein (ARF related news story on Jaeger et al., 2010).

Neely and colleagues’ current study found lower beclin-1 levels in PS-null cells. On the whole, though, its findings do not tie in with APP processing or tau pathology. Rather, they reveal “the physiological role of PS in an important protein degradation pathway,” Green noted. “Whether this has any implications for sporadic AD we don’t know.”

Besides trying to understand exactly how PS figures in autophagy, the Irvine team is now examining cells with PS mutations associated with the age-related disease frontotemporal dementia to see if autophagy in those cells differs from that of cells with PS mutations that cause familial AD.—Esther Landhuis.

Reference:
Neely KM, Green KN, LaFerla FM. Presenilin is necessary for efficient proteolysis through the autophagy-lysosome system in a γ-secretase-independent manner. J. Neurosci. 23 Feb 2011;31(8):2781-2791. Abstract