Transmissible spongiform encephalopathies (TSEs), such as scrapie in sheep and Creutzfeld–Jakob disease (CJD) in humans, occur as numerous different strains in each species. The existence of these stable variants of the disease is one of the main unresolved issues for Stanley Prusiner's prion hypothesis, or 'protein-only hypothesis', of TSEs. This hypothesis proposes that the infectious agent in TSEs is a rogue form (PrpSc) of a cell's normal prion protein (PrpC), possibly along with a putative non-nucleic- acid cofactor.

Strains are a problem for Prusiner's hypothesis because, assuming that each strain represents a different conformational version of the prion protein, it is unclear how the protein can adopt, under physiological conditions, as many different conformations as there are strains of TSEs. Elsewhere in this issue (Nature Cell Biology 1, 55–59; 1999), John Wadsworth and colleagues report an intriguing link between metal-ion occupancy of the prion protein and its biochemical strain characteristics.

Recent research has focused on the copper-binding capabilities of the amino- terminal domain of Prp C as a clue to understanding the protein's normal cellular function. Wadsworth and colleagues look instead at metal-ion occupancy in Prp Sc isolated from two biochemically distinct strains of sporadic CJD. They use metal-chelating agents to disrupt binding of PrpSc to Cu2+ and Zn2+; this treatment slightly alters the electrophoretic mobility of the cleavage products of PrpSc after limited proteolysis of the prion protein using proteinase K — one of the biochemical indicators of strain in TSEs.

It would be tempting to claim that this observation points to a molecular explanation of TSE strains, with metal-ion occupancy determining strain specificity, but this would be premature. For a start, Wadsworth and colleagues found that the effect on PrpSc biochemistry of altering metal-ion occupancy extends to only two of the four biochemically defined strains of CJD, and so does not seem to offer a general explanation for strain variation. It also remains to be shown whether or not these altered strain characteristics can be transmitted experimentally to animals in inocula pretreated with chelators. But this provocative link between the copper-binding capabilities of prion protein and strain variations in PrpSc points research in a new direction.