Kufor-Rakeb syndrome (KRS) was originally identified in a Jordanian kindred, but Christian Kubisch, University of Cologne, leading an international team of collaborators from Germany, England, Chile, and Jordan, focused on a large Chilean family to identify the gene: several members of the family are afflicted with a disease that closely resembles KRS. Using mutation screening and linkage analysis, first author Alfredo Ramirez and colleagues eliminated some of the usual PD suspects, including loci coding for parkin, DJ-1, and PINK1, but found that PARK9 was linked to a region between the last two loci. After narrowing down this segment to 3.2 Mb of DNA containing about 40 genes, none of which were obvious functional candidates for KRS, Ramirez and colleagues sequenced all the exons of the heretofore uncharacterized ATP13A2, a gene predicted by expressed-sequence tag and homology data. The authors found two mutations in the ATPase, one each inherited from the father and the mother, neither of whom is affected by the disease.

ATP13A2 codes for a large ATPase with 10 transmembrane domains. The first mutation, a single nucleotide deletion, introduces a premature stop codon that truncates the protein just before the last three transmembrane domains. The second mutation, a guanine-to-adenine transition at a highly conserved splice site, is predicted to reduce splicing efficiency by about 90 percent. In fact, when Ramirez and colleagues analyzed RNA from affected family members, they found that exon 13 was completely skipped, removing a large segment of the third transmembrane domain. “The remaining hydrophobic residues of this domain will not be able to span the membrane, which should lead to distortion of transmembrane topology and also to loss of function,” write the authors. In collaboration with Amir Al-Din at King Hussein Medical Center in Amman, Kubisch and colleagues then analyzed DNA from the original Jordanian kindred and found a 22 base pair duplication in ATP13A2 in all affected family members. This duplication also leads to a frame shift, introducing a stop codon that eliminates the six C-terminal transmembrane domains.

How this ATPase, or lack thereof, brings about KRS is unclear. Ramirez and colleagues found that the protein is normally found in the brain, and by laser microdissection analysis they confirmed that it is expressed in individual dopaminergic neurons of the substantia nigra (SN)—which degenerate in PD—and the associated ventral tegmental area. Quantitative RNA amplification experiments also showed that its expression in the brain is strongest in the SN and weakest in the cerebellum.

Because there are no antibodies available for ATP13A2, the authors studied its cellular localization by expressing epitope-tagged proteins in COS7 cells. The native ATPase turned up in the lysosome, while the truncated mutant proteins seemed to get stuck in the endoplasmic reticulum. But very little of the truncated proteins could be detected, indicating that they are actively degraded. The proteasome machinery most likely takes care of these misfolded proteins because Ramirez found that the proteasome inhibitor MG-132 stabilized the truncated variants. “This may at least partially explain the neurodegeneration in KRS—for example, by proteasomal dysfunction owing to overload with mutant ATP13A2, which in turn might cause toxic aggregation,” write the authors.

But Ramirez and colleagues also accept that lysosomal dysfunction following loss of the ATPase may contribute to the disease, perhaps by compromising lysosomal protein degradation. On this point it is worth noting that lysosomal degradation of α-synuclein, which also causes PD if overexpressed or mutated, may be important in PD pathology (see Webb et al., 2003). Also, mutations in other lysosomal proteins have been found in patients with sporadic PD (Aharon-Peretz et al., 2004; Goker-Alpan et al., 2004). And this leads to another fascinating facet of ATP13A2—when Ramirez and colleagues analyzed SN neurons taken postmortem from sporadic PD patients, they found 10-fold higher levels of the ATPase mRNA, indicating, perhaps, a compensatory upregulation of the protein in surviving neurons.

One issue not addressed in the paper is how closely KRS resembles Parkinson disease. KRS, for example, is characterized by additional symptoms, such as dementia, that are not found in typical PD. This issue is discussed in depth in the following Alzforum comment by Mark Cookson from the National Institute on Aging, Bethesda, Maryland.

As for LRRK2, aka “dardarin,” found at the PARK8 locus (see ARF related news story), Matthew Farrer and colleagues at the Mayo Clinic in Jacksonville and the University of Miami, Florida, report that the most common LRRK2, and PD, mutation, a glycine-to-serine substitution at amino acid 2019 of the kinase, results in remarkably varied symptoms. Writing in the September Archives of Neurology, first author Spiridou Papapetropoulos and colleagues evaluated five patients with the mutation. One had familial PD, one had died at age 68 with no pathological signs of neurodegeneration, and the other three had sporadic PD with varying age of onset (41-79) and pathological features. The findings support earlier data suggesting that this mutation is either not fully penetrant or that the age of onset can be sufficiently late that some carriers fail to develop symptoms in their lifetimes.

In a similar vein, and in the same journal, Lianna Ishihara from the University of Cambridge, England, and international collaborators from Europe, Africa, Japan, and the U.S., report that among 26 patients carrying two copies of the G2019S mutation, symptoms of Parkinson disease are no worse than seen in those carrying only one copy of the mutant gene. Again, the data seem to suggest that there is more to the etiology of PD than simply the right titer of mutant LRRK2.—Tom Fagan.

Reference:
Ramirez A, Heimbach A, Grundemann J, Stiller B, Hampshire D, Cid LP, Goebel I, Mubaidin AF, Wriekat A-L, Roeper J, Al-Din A, Hillmer AM, Karsak M, Liss B, Woods CG, Behrens MI, Kubisch C. Hereditary parkinsonism with dementia is caused by mutation in ATP13A2, encoding alysosomal type 5 P-type ATPase. Nature Genetics. Advanced online publication. Abstract