The second study goes downstream to identify proteases that regulate α-synuclein levels. That work, from Michael Schlossmacher at the Ottawa Health Research Institute in Canada and Jaana Tyynela at Helsinki University in Finland, confirms and extends earlier reports by linking cathepsin D activity to α-synuclein levels in cells and in mouse, sheep, and human brain. Although it is not clear whether the protease directly degrades α-synuclein, or participates in a cascade that activates a different synucleinase enzyme, the results, published February 9 in Molecular Brain, should stimulate further work on cathepsin D as a potential target for decreasing α-synuclein concentrations in PD and other synucleinopathies.

Variants in the gene for UCH-L1, a ubiquitin hydrolase highly expressed in neurons, have been associated with both Parkinson disease and Alzheimer disease, but the in vivo functions of the protein are poorly understood. Previous work by the Lansbury group showed that expression of UCH-L1 could promote α-synuclein accumulation in cells (see ARF related news story). The new study tracks this stabilizing function to a subset of UCH-L1 molecules that are post-translationally modified and membrane-associated (UCH-L1M). First author Zhihua Liu and coworkers find that about one-third of UCH-L1 in brain exists in a farnesylated form, associated with the endoplasmic reticulum membrane. They also show that UCH-L1 lacking the C-terminal farnesylation site loses the ability to localize to membranes and to stabilize α-synuclein.

Therein lies a promising drug target, Lansbury believes. Inhibitors of the farnesyltransferase enzyme have been widely studied and tested in humans as potential cancer therapies because of their effects on the activity of the Ras protein. Lansbury and colleagues show that the farnesyltransferase inhibitor FTI-277 reduces UCH-L1M, as well as levels of α-synuclein and neurotoxicity in SH-SY5Y cells made to overexpress α-synuclein. FTI-277 also prevented the toxicity of the A53T synuclein mutant in mouse primary dopaminergic neurons, and lowered endogenous α-synuclein levels in primary cortical neurons.

“The link between UCH-L1M and accumulation and neurotoxicity of α-synuclein suggests that farnesyltransferase inhibitors that reduce the amount of UCH-L1M may be useful for the treatment of PD and related synucleinopathies,” the authors conclude. Of course, it remains to be seen if the inhibitors can affect synuclein pathology in animals. In an e-mail to ARF, Lansbury wrote that Link Medicine “is interested in the possibility that farnesylated UCH-L1 may be a viable therapeutic target for PD.” He provided no more information, nor does the company’s website give any inkling of news on the progress of this project.

More work needs to be done before scientists fully understand how UCH-L1 acts to modulate α-synuclein degradation. According to data in the paper, UCH-L1M does not affect proteasome function, suggesting that it instead regulates the lysosomal pathway of synuclein degradation first described by Anna-Maria Cuervo and colleagues (see ARF related news story on Cuervo et al., 2004). UCH-L1 probably has other functions, too. Studies in mice showed that overexpression of a soluble form of UCH-L1 protects AD mice from neurotoxicity and memory deficits (Gong et al., 2006; see ARF related conference story). An important avenue for future work will be to figure out the relative importance of UCH-L1M and other forms of the protein for its various functions. Of note, only the neuron-specific homolog UCH-L1 has a farnesylation site, whereas its more widely expressed cousin UCH-L3 does not. “This implies that these relatives have different roles,” Lansbury said.

Once synuclein is in the lysosome, it encounters a host of proteolytic enzymes, but it has not been clear which function as actual synucleinases. The data from Schlossmacher and Tyynela add to a growing body of evidence that the aspartyl protease cathepsin D plays a role in this pathway. Their paper extends two other recent reports that show alterations in α-synuclein stability with overexpression of cathepsin D and in knockout mice (Sevlever et al., 2008; Qiao et al., 2008). In the new work, first author Valerie Cullen and coworkers show that overexpression of cathepsin D in dopaminergic cells lowers α-synuclein levels. In-vivo data from cathepsin D knockout mice shows an accumulation of insoluble α-synuclein and the appearance of intracellular α-synuclein inclusions in brain tissue. A line of cathepsin D-deficient sheep also develop cytoplasmic accumulation of synuclein in neurons, the authors show. In people, a cathepsin D deficit results in fatal lysosomal storage disease, and the researchers find that brain tissue from infants who died from the disorder also contains synuclein inclusions.

To look at the effect of cathepsin D on synuclein toxicity, the researchers collaborated with Mel Feany at Harvard Medical School, who looked at synuclein overexpression in fruit flies lacking cathepsin D. The α-synuclein flies showed degeneration of retinal neurons, which was enhanced in the absence of cathepsin D.

With three groups chiming in, Schlossmacher said the weight of the data points to cathepsin D as a player in synuclein degradation, although it may not be the only or even the major player. “It’s a nice synergy of several teams showing in cellular and in vivo models that it seems to be cathepsin D plays a role in synuclein turnover, but I would not go as far as to say it is the enzyme,” he told ARF. In addition, there is no evidence to point to synuclein as a direct substrate for cathepsin D in vivo, he said. “There’s no proof yet that it’s really the biochemical scissors that chop up the synuclein, but it seems to activate something within the cell that facilitates that degradation.”

The important next steps will involve understanding the pathway, that is, to see whether the cathepsin D acts directly or involves other, effector proteases. Also, it will be important to discern whether cathepsin D can reverse established pathology in animal models of synuclein expression. If that works, Schlossmacher says, the next step would be to start looking for ways to activate cathepsin D or have the brain make more of it, possibly using small molecules. As a precedent for an interest in lysosomal proteases in neurodegeneration, the researchers can look to the cysteine protease cathepsin B, which has received a lot of attention as a regulator of β amyloid levels in AD (see ARF related news story and Sun et al., 2008).—Pat McCaffrey.

References:
Liu Z, Meray RK, Grammatopoulos TN, Fredenburg RA, Cookson MR, Liu Y, Logan T, Lansbury PT. Membrane-associated farnesylated UCH-L1 promotes alpha-synucleinuclein neurotoxity and is a therapeutic target for Parkinson’s disease. PNAS. 2009 Feb 23. Abstract

Cullen V, Lindfors M, Ng J, Paetau A, Swinton E, Kolodziej P, Boston H, Saftig P, Woulfe J, Feany MB, Myllykangas L, Schlossmacher MG, Tyynelä J. Cathepsin D expression level affects alpha-synuclein processing, aggregation, and toxicity in vivo. Mol Brain. 2009 Feb 9;2(1):5. Abstract