For the same reasons, Chinese Hamster Ovary cells were used in parallel with CHO cells transfected with pcDNA5/FRT/PrP plasmid containing the gene for expression of wild type full-length Syrian hamster PrPC. Two assays were used for evaluating the effect of intact or fragmented R- or S-fibrils on cultured cells. The XTT assay assesses cell metabolism by measuring the activity of mitochondrial dehydrogenases, whereas trypan blue selectively stains dead cells. Comparative analysis of cell viability within each cell line and between lines revealed the following relationships. First, the cell lines that expressed low levels of PrPC showed very modest cytotoxic responses. The toxic effects tend to be higher in the lines with high levels of PrPC expression. Second, contrary to the prevailing view, fragmented R-fibrils were found to be substantially less toxic than the intact Rfibrils in all cultured cells. In fact, as judged from both assays, fragmented R-fibrils had minimal effect, if any, as SB431542 compared to the non-treated controls. Again, when treated with intact R-fibrils, the cellular response was weaker in lines with low levels of PrPC expression. Third, the cytotoxic potential of S-fibrils followed the opposite trend : the toxicity of fragmented S-fibrils were similar or more pronounced as compared to intact Sfibrils. These findings revealed that small fibrillar fragments were more toxic than the intact fibrils for one amyloid structure, whereas for the alternative amyloid structure the small fragments were considerably less toxic than the intact fibrils. Interestingly, while fragmented R-fibrils were much smaller than the intact Sfibrils, the former were generally less toxic than the latter. This observation further supports the notion that size difference alone can not explain the toxic potential of amyloid states. The differences in amplitude of response between the two assays were probably due to the fact that the XTT assay probes changes in the cell metabolism that could be considered as an intermediate step in a cell’s response, whereas trypan blue assesses the percentage of cells that were irreversibly damaged and died. Regardless of the level of PrPC expression, all cell lines showed similar rank orders with respect to their susceptibilities to intact or sonicated R- or S-fibrils within each individual cell line. Establishing the relationship between the physical state of a protein and toxicity is essential for developing effective therapeutic strategies against neurodegenerative diseases. In the prevailing opinion, soluble oligomers or small fibrillar fragments are considered to be the most toxic species, whereas formation of large amyloid fibrils and plaques are thought to be a protective process by which cells sequester more dangerous oligomers. The present finding provides new opportunities for reexamining this view. This work warns that without specifying the molecular structures of the protein aggregates, claims regarding the relationships between size and toxicity of amyloid states could be misleading. The current study revealed that for the S-structures produced from full-length recombinant prion protein, small fibrillar fragments were more toxic than the intact fibrils, whereas for the R-structures produced from the same protein, the small fragments were considerably less toxic than the intact fibrils. Remarkably, fragmentation of R-fibrils almost completely abolished their cytotoxic potential. Considering that both R- and Sfibrils are produced within the same amino acid sequence using highly pure rPrP, the differences in cell response to intact versus fragmented fibrils have to be attributed to the distinct molecular structures of the two amyloid states. The R- and Sstructures were analyzed previously using a broad range of biophysical techniques including X-ray diffraction, CD, hydrogendeuterium exchange Raman spectroscopy, FTIR spectroscopy, hydrogen-deuterium exchange monitored by FTIR, proteinase K -digestion assay, binding of a conformation-sensitive fluorescence dye, immunoconformational assay, atomic force microscopy and electron microscopy. The R- and S-fibrils were found to have fundamentally different secondary, tertiary and quaternary structures. While both amyloid states displayed a 4.8 A ° meridional X-ray diffraction typical for amyloid cross-b spines, they showed markedly different equatorial profiles suggesting fundamentally different architectures of the cross b-spine. Using solid state NMR, the cross-b core of R-fibrils was found to consist of in-register, parallel b-sheet structure. No molecular details are currently available from NMR methods about structure of S-fibrils. Nevertheless, together with previous studies this work demonstrates that the relationship between fibril size and their cytotoxic potential is not unidirectional and is controlled by the molecular structures of the amyloid states. The current work demonstrated the remarkable ability of cells to recognize and respond differently to conformationally distinct amyloid states even if they are formed within the same amino acid sequence. As evident from previous studies, not only were the cross b-spine structures markedly different in R- and S-fibrils, but also their surface epitope presentation and PK-resistant regions. For instance, the epitope to R1 antibodies was found to be solvent exposed in S-fibrils, but buried in the fibrillar interior in R-structures. The N-terminal region 23-,50 was found to be PK-resistant in S-structures, but PK-sensitive in R-fibrils. As judged from the epitope presentation and PKresistant profile, R-fibrils resembled the structure of PrPSc more closely than the S-fibrils. Moreover, unlike S-fibrils, R-fibrils were found to be capable of inducing a transmissible form of prion diseases in wild type animals. Unexpectedly, fragmentation of R-amyloids into fibrils of shorter length was found to abolish their cytotoxic potential, an observation that contradicts the currently dominating view.