Fred E. Cohen

Two different neurodegenerative diseases caused by proteins with similar structures

The downstream prion-like protein (doppel, or Dpl) is a paralog of the cellular prion protein, PrPC. The two proteins have ≈25% sequence identity, but seem to have distinct physiologic roles. Unlike PrPC, Dpl does not support prion replication; instead, overexpression of Dpl in the brain seems to cause a completely different neurodegenerative disease. We report the solution structure of a fragment of recombinant mouse Dpl (residues 26–157) containing a globular domain with three helices and a small amount of β-structure. Overall, the topology of Dpl is very similar to that of PrPC. Significant differences include a marked kink in one of the helices in Dpl, and a different orientation of the two short β-strands. Although the two proteins most likely arose through duplication of a single ancestral gene, the relationship is now so distant that only the structures retain similarity; the functions have diversified along with the sequence.

Time-controlled transcardiac perfusion cross-linking for the study of protein interactions in complex tissues

Because of their sensitivity to solubilizing detergents, membrane protein assemblies are difficult to study. We describe a protocol that covalently conserves protein interactions through time-controlled transcardiac perfusion cross-linking (tcTPC) before disruption of tissue integrity. To validate tcTPC for identifying protein-protein interactions, we established that tcTPC allowed stringent immunoaffinity purification of the γ-secretase complex in high salt concentrations and detergents and was compatible with mass spectrometric identification of cross-linked aph-1, presenilin-1 and nicastrin. We then applied tcTPC to identify more than 20 proteins residing in the vicinity of the cellular prion protein (PrPC), suggesting that PrP is embedded in specialized membrane regions with a subset of molecules that, like PrP, use a glycosylphosphatidylinositol anchor for membrane attachment. Many of these proteins have been implicated in cell adhesion/neuritic outgrowth, and harbor immunoglobulin C2 and fibronectin type III–like motifs.

Tertiary Structure Prediction

The polypeptide chain of globular protein is linear, but the three-dimensional of tertiary structure is quite contorted. This was apparent from the first crystallographic determination of the structure of a protein by Kendrew et al. (1960). The contortion arises because of the need to satisfy a multitude of conflicting interactions: the hydrogen-bonding requirements of buried nitrogen and oxygen atoms, placement of the remaining polar groups near the protein-solvent interface, and internalization of most hydrophobic residues. Anfinsen et al. (1961) demonstarted that the amino acid sequence contained enough information to define a protein tertiary structure. These experiments defined the protein-folding problem: by computation determine the precise tertiary structure of a protein from its amino acid sequence. In the 25 years that have followed, much regularity and order have been recognized in protein tertiary structure (for reviews, see Richards, 1977; Nemethy and Scherega, 1977; Schulz and Schirmer, 1979; Richardson, 1981; Sternberg, 1983; Chothia, 1984). This chapter describes some of the theoritical methods designed to predict protein structure.

Structures of Prion Proteins and Conformational Models for Prion Diseases

Prions are a novel class of “infectious” pathogens distinct from viroids and viruses with respect to both their structure and the neurodegenerative diseases that they cause (PRUSINER 1991). Prion diseases are manifest as sporadic, inherited, and infectious disorders including scrapie, mink encephalopathy, chronic wasting disease, bovine spongiform encephalopathy, feline spongiform encephalopathy, and exotic ungulate encephalopathy of animals (MARSH et al. 1991; WESTAWAY et al. 1994b; WILESMITH and WELLS 1991) as well as kuru, Creutzfeldt-Jakob disease (CJD), Gerstmann-Straussler-Scheinker syndrome, and fatal familial insomnia of humans (GAJDUSEK 1977; HSIAO et al. 1989; Medori et al. 1992). The prion protein (PrP) is the major, if not the only, component of prions (PRUSINER 1991). PrP exists in two isoforms: the normal cellular form (PrPc) and the abnormal disease-related form (PrPSc) (MEYER et al. 1986; OESCH et al. 1985).

Quantitative traits of prion strains are enciphered in the conformation of the prion protein.

Variations in prions, which cause different disease phenotypes, are often referred to as strains. Strains replicate with a high degree of fidelity, which demands a mechanism that can account for this phenomenon. Prion strains differ by qualitative characteristics such as clinical symptoms, brain pathology, topology of accumulated PrP(Sc), and Western blot patterns of glycosylated or deglycosylated PrP(Sc). Since none of these qualitative features can directly explain quantitative strain traits such as incubation time or dose response, we analyzed conformational parameters of PrP(Sc) and the rate of accumulation in different prion strains. Using the conformation-dependent immunoassay (CDI), we were able to discriminate among PrP(Sc) molecules from eight different prion strains propagated in Syrian hamsters. CDI quantifies PrP isoforms by simultaneously following antibody binding to both the denatured and native forms of a protein. In a plot of the ratio of antibody binding to denatured/native PrP graphed as a function of the concentration of PrP(Sc), each strain occupied a unique position, indicating that each strain accumulated different concentrations of particular PrP(Sc) conformers. This conclusion was supported by a unique pattern of equilibrium unfolding of PrP(Sc) found within each strain. By comparing the PrP(Sc) levels before and after limited proteinase K digestion, we found that each strain produces a substantial fraction of protease-sensitive PrP(Sc). We asked whether this fraction of PrP(Sc) might reflect those PrP(Sc) molecules that are most readily cleared by cellular proteases. When the protease-sensitive PrP(Sc) fraction was plotted as a function of the incubation time, a linear relationship was found with an excellent correlation coefficient (r = 0.94). Combined with the data on time courses of prion infection in Tg(MHu2M) and Tg(SHaPrP) mice, the results argue that different incubation times of various prion strains may arise predominantly from distinct rates of PrP(Sc) clearance rather than from different rates of PrP(Sc) formation.

Molecular Biology of Prion Propagation

Prions are novel pathogens and are distinct from both viroids and viruses. They are composed largely, if not entirely, of the scrapie isoform of the prion protein (PrP) designated PrPSc. Prions cause neurodegenerative diseases including scrapie of sheep, mad cow disease, and Creutzfeldt-Jakob disease (CJD) of humans. That prion diseases are manifest as genetic, infectious and sporadic illnesses is unprecedented. The conversion of PrPC into PrPSc is a post-translational process involving a profound conformational change which is the fundamental event underlying the propagation of prions. PrPC contains −40% α-helix and virtually no β-sheet; in contrast, PrPSc has −30% α-helix and −40% β-sheet. These data argue that the conversion of α-helices into β-sheets underlies the formation of PrPSc. Efficient formation of PrPSc requires targeting PrPC by glycosylphosphatidyl inositol (GPI) anchor to a caveolae-like membrane domain (CLD) which is detergent insoluble and enriched for cholesterol and glycosphingolipids. Redirecting PrPC to clathrin-coated pits by creating chimeric PrP molecules with four different C-terminal targeting domains prevented the formation of PrPSc. To determine if these C-terminal transmembrane segments prevented PrPC from refolding into PrPSc by altering the structure of the polypeptide, we fused the 28 amino acid C-termini from the Qa protein. Two C-terminal Qa segments differing by a single residue direct the trans-membrane protein to clathrin coated pits or the BPI form to CLDs. The CLD targeted PrPC was converted into PrPSc while the transmembrane PrPC was not. Transgenic (Tg) mice expressing human (Hu) prion protein (PrP) and chimeric Hu/mouse (Mo) PrP genes were inoculated with brain extracts from humans with inherited or sporadic prion disease.