Polyglutamine fibrillogenesis: 1

N i rine neurodegenerative disof AP protein in Alzheir eases are caused by expanding and a-synuclein in Parki CAG repeats coding for polyas well as other amyloidi glutamine (polyGln) (1-4). (16-18). Even if polyGlr These include Huntington's disease, not the major toxic spec: dentatorubral and pallidoluysian atrotion process appears link phy, several forms of spino-cerebellar esis. Therefore, it is criti ataxia, and spinal and bulbar muscular stand the structure of bc atrophy. Within the central nervous sysmutant polyGln stretche: tem, each disease has a distinctive patDetailed structural infc tern of degeneration, with considerable polyGln has been difficull overlap among the diseases (5, 6). The cause both long and shorl genes containing CAG repeats show no synthetic polyGln peptide homology to each other outside of the uble. Nearly a decade age glutamine repeats, and most are genes of attempted to address this unknown function. Thus, speculation conQ15 peptide flanked by b< cerning pathogenesis has focused on the improve its solubility. He polyGln expansion itself. tide to adopt p-structure, For all of these diseases, there is a an atomic model of poly(: threshold of repeat length that causes consisting of antiparallel I disease. This threshold varies somewhat together by hydrogen bon among the different diseases, but is genmain-chain and side-chair erally in the range of 35-45 consecutive la). This structure, descri glutamines. In all polyGln diseases, the zipper," has been influen age of disease onset is strongly correpolyGln aggregation (14, lated with polyGln length, so that above puter modeling studies ha the threshold, a longer repeat results in additional possible structt an earlier age of onset. panded polyGln, such as I A pathological hallmark of these dis(21), P-hairpins, and high eases is the aggregation of mutant dom coil (22) or 3-sheet , polyGln protein, resulting in the formab-e). Based on x-ray diffr tion of intranuclear inclusion bodies. In tron microscopy data, Pei some of the diseases, inclusions have been group (23, 24) have recen observed in the cytoplasm, dendrites, and polyGln 3-helix model wi axonal processes. The inclusions are genper turn (Fig. lf). erally seen in affected areas of the brain Biophysical analysis of (7, 8), though not limited to those neurecombinant polyGln pe rons most likely to degenerate (9). Thus, stretches containing 5-4 whether inclusions are responsible for glutamines have demons neurotoxicity has been controversial. nomeric polyGln is unsti Some studies have indicated a correlation 27). In contrast, expandc between polyGln-containing inclusions gregates derived from is( and disease progression (10). However, in peptides or from recoml other studies, inclusion formation was discontaining proteins adof sociated from cytotoxicity (11, 12). In fact, ture, as shown by x-ray f inclusion formation may be, in part, a restudies, circular dichroisJ flection of cellular protective mechanisms transform infrared spect (13). Nevertheless, the inclusions are a other methods (20, 25, 2 useful marker for pathology and may prois likely that expanded p vide clues to pathogenesis. quences in the aggregate Aggregation of mutant polyGln pro3-sheets, though detailer teins can be observed biochemically usformation is currently n( ing a filter trap assay (14, 15). AggregaIn a recent issue of PI tion in vitro proceeds by means of a and Wetzel (30) provide nucleation-dependent process and reanalysis to address the q suits in the accumulation of /3-sheet rich polyGln aggregate struct fibrillar structures detected by electron the Wetzel laboratory de microscopy. Thus, the polyGln aggregapolyGln peptides have a, tion pathway appears to resemble that erties similar to exon-1 I