The talks at this meeting fell into these categories:

Fibril formation
Tauopathies, α-synucleinopaties
α-synuclein function
Parkin, synphilin, ubiquitination

Hilal Lashuel of Brigham and Women's’ Hospital in Boston, Massachusetts, talked about the biophysics of fibril formation. A cascade of folded protein to protofibril (oligomers) to protofilaments (thin) to fibrils (cross-β-sheet amyloid) is postulated, and toxicity could increase along this cascade. The working hypothesis is that the protofibrils, rather than the fibrils, are the toxic species of amyloids in general. Synthetic α-synuclein (αSYN) protofibrils display some heterogeneity.

Investigation of the various protofibrillar species shows that some of them resemble pore-like tubes that allow passage of small molecules (<0.8kD; e.g., protons, calcium, dopamine). This annular/tubular conformation is more prominently observed with the A53T or the A30P mutations of αSYN. Dopamine and other catechols stabilize αSYN protofibrils, while βSYN prevents αSYN protofibril and fibril formation. Apparently, also, other amyloid proteins yield such "channels": examples include Aβ (arctic AβPP mutation ->[E22G]Aβ), IAPP (type II diabetes), serum amyloid A (uniform hexameric channels), prion protein, and [A4V]SOD1 (see ARF related news story); Lansbury interview; Wang et al., 2002; Lashuel et al., 2002; Lashuel et al., 2002b). Remarkably, when wild-type and arctic Aβ are mixed, the amount of protofibrils increases.

Virginia Lee, University of Pennsylvania School of Medicine, Philadelphia, started with a summary of the three classes of FTDP-17 tau mutations:

• mutations impairing tau function (reduced MT binding) G272V
• mutations promoting tau aggregation P301L
• mutations altering exon 10 splicing N272N

Both tau and α-syn are abundant, small, highly soluble, amphipatic neuronal proteins, with a long half-life and the potential to become pathologically phosphorylated. Both aggregate into 10-20nm filaments, generate true amyloid and can be ubiquitinated, nitrated and phosphorylated. Remarkably, both co-occur in LBVAD, AD, familial AD, Down’s syndrome, but also in diseases with scant Aβ pathology: in DLB and PD there is partial overlap of immunofluorescence within cell bodies and neurites. In FTPD-17, there is co-localization of tau and α-syn in glial cytoplasmic inclusions of MSA. α-syn and all six isoforms of tau synergistically enhance each other’s fibrillization in vitro, as detected by K114 fluorometry. Interestingly, homopolymers are formed, so there is synergistic formation of separate α-syn and tau fibrils. In (PrP)-[A53T] α-syn mice, there is tau pathology in brainstem and spinal cord. Bigenic mice expressing tau and α-syn under control of the oligodendrocyte-specific CNP promoter have enhanced fibrillization of tau and α-synin white matter. Also, in an A53T α-syn mutant patient, a lot of tau pathology occurs even in regions where little α-syn is observed.

John Trojanowski, University of Pennsylvania School of Medicine, Philadelphia, discussed the possible functions of α-syn, which include:

• vesicle binding
• regulation of the number of synaptic vesicles
• synaptic transmission
• chaperoning
• phospholipase D2 inhibition

Epitope mapping with many α-syn monoclonal antibodies revealed the presence of full-length α-syn in all α-synopathy lesions. High molecular weight α-syn species can be extracted from patient brains. In glial cytoplasmic inclusions of MSA, C-terminal α-syn epitopes are immunodominant, whereas in all the other lesions the epitopes are equally represented. α-syn in lesions is ubiquitinated, nitrated, and phosphorylated. Oxidized α-syn antibodies reveal striatal pathology. HSP70 rescues dopaminergic neurons from degeneration in the α-syn fly model (see ARF related news story). Indeed, α-syn lesions contain HSP70 and HSP40.

Mice highly overexpressing h[A53T] α-syn driven by the PrP promoter develop progressive locomotor deterioration. Age of onset is between nine and 16 months, and the phenotype is lethal within a month of outbreak. The sites of α-synopathy are outside of the dopaminergic system. Insoluble α-syn filaments accumulate in the transgenic mice. For more on α-syn mouse models, see ARF related news story. Finally, posttraumatic recovery is impaired in a neurofilament-inclusion-bearing mouse model, because of increased necrotic neuron loss. Perhaps the induction of iNOS in the traumatic area contributes to this sensitization.

Philipp Kahle, Ludwig Maximilians University of Munich, centered his presentation around a recently recognized pathological feature of α-syn, namely proteinase K (PK) resistance (Neumann et al., 2002). A parallel study was undertaken with postmortem brains that were split, one hemisphere frozen and the other fixed in formalin. PK resistance of α-syn was assessed in biochemically isolated fibrils and in situ on digested paraffin-embedded tissue blots. This revealed classical α-synopathy, as well as previously underappreciated pathology. Generally, the regional distribution of PK-resistant α-syn correlated with the stage of α-synopathy, suggesting that Lewy pathology constitutes a primary lesion in the classical disease course, which spreads from the medulla to the brainstem, and then to the limbic and neocortical systems. Interestingly, FAD patients displayed strong α-synopathy in the amygdala, indicating that under certain circumstances, and in the limbic system specifically, α-syn may fibrillize secondary to AD plaque deposition. Kahle went on to point out that the A30P mutation of α-syn also causes Lewy pathology and locomotor deterioration in a transgenic mouse model. As in human patients, PK-resistant and Ser-129-hyperphosphorylated α-syn fibrils developed in phenotypic transgenic mice. Additional markers of α-synopathy in the A30P α-syn mice (Thy1 promoter) included silver-positive dystrophic neurites, thioflavin S-positive Lewy bodies and Lewy neurites, electron-dense fibrils clogging swollen neurites, oxidative protein modifications, and gliosis, and ubiquitination of the lesions. Interestingly, the sites of α-syn fibril formation were in the spinal cord, brainstem, and deep mesencephalic nuclei, but the nigrostriatal dopamine system remained unaffected.

In a second talk, Iwatsubo reported on the purification of 3.5 million Lewy bodies (LB) to generate the LB-specific monoclonal antibody LB509. MALDI-TOF analysis of LB-derived α-syn revealed Ser-129 phosphorylation, whereas soluble α-syn was not phosphorylated. Also, mono- and diubiquitinated α-syn was identified. These are being purified by size-exclusion HPLC. Iwatsubo used the Mec-7 promotor to drive α-syn expression touch neurons of C. elegans, but even aged animals of this strain did not hyperphosphorylate α-syn. The touch response is impaired in transgenic worms expressing the PD-causing A30P mutant of α-syn, but less so in worms expressing the other mutant A53T; almost no deficiency was seen with wild-type α-syn. Using the dat1 promoter, α-syn was expressed in dopaminergic neurons. In these cells, α-syn sometimes becomes hyperphosphorylated. The function of dopaminergic neurons in live C. elegans can be evaluated by measuring a decrease in body bending and movement in food-rich areas. Indeed, A53T and A30P α-syn expressed under control of the dat1 promoter did decrease the worm’s typical motor behavior in food-rich areas; adding L-DOPA rescues this phenotype. Baumeister cautioned that the arrest of body bending and movement in food-rich areas by α-syn overexpression in C. elegans may be unspecific, since β-syn expression, for example, produces the same phenotype.

Ryosuke Takahashi of the RIKEN Brain Science Institute in Saitama, Japan, spoke about the autosomal recessive form of juvenile parkinsonism caused by mutations in parkin. Parkin has E3 ubiquitin ligase activity and mutant parkin shows enzymatic deficiency in vitro. One of the substrates of parkin is Pael (Parkin-associated endothelin receptor-like) receptor (see ARF related news story). This protein is almost exclusively expressed in the dopaminergic neurons of the substantia nigra. In the rest of the CNS, Pael-R is mostly localized in oligodendrocytes. Parkin regulates Pael-R turnover in the context of endoplasmic reticulum-associated degradation. Proteasome inhibition by six-hour lactacystin treatment causes Pael-R accumulation in the ER of SH-SY5Y cells; longer treatment leads to Pael-R aggresome formation. Unfolded (more specifically, insoluble) Pael-R accumulates in AR-JP (but not idiopathic PD) brain, but no Pael-R aggregates are stained in patient brain. Transgenic Drosophila expressing Pael-R show selective degeneration of DA neurons (Yang et al., 2003).

Other parkin-binding proteins are HSP70 and CHIP (C-terminus of Hsc70-interacting protein). CHIP contains a C-terminal TPR domain (Hsp70/Hsp40 binding) and a U box (E2 binding). It is a cofactor of the parkin complex that enhances parkin’s E3 activity, particularly when CHIP and Hsp70 are upregulated by ER stress (mediated by tunicamycin). Hsp70/Hsp40 induction peaks six hours after tunicamycin addition (initially preventing Pael-R aggregation), and then CHIP gradually increases for 24 hours, helping parkin to degrade the accumulating, misfolded Pael-R.

Christopher Ross of Johns Hopkins University School of Medicine, Baltimore, Maryland, reviewed the aggregation-promoting properties of the α-syn-interacting protein synphilin (SPH-1), and demonstrated that parkin interacts with synphilin-1 (see ARF related news story). Parkin ubiquitinates synphilin-1 in the cytosolic inclusions formed in cells cotransfected with SPH-1 and α-syn. α-syn was not a substrate for parkin, but co-localized with SPH-1 and parkin in the inclusions. Interestingly, parkin-mediated ubiquitination did not reduce the half-life of SPH-1, because it catalyzed poly-ubiquitination via K63 rather than K48. Unlike K48-linked polyubiquitin chains that direct the attached proteins to proteasomal degradation, K63-linked ubiquitination influences subcellular localization or signaling of the target proteins. It is of prime importance to clarify the biology and potential relevance to Parkinson’s of K63-linked polyubiquitination catalyzed by parkin.