Application). Novel constrained dipeptide units will be synthesized and oligomerized to test if the resulting peptides will have a high propensity to form extended conformation strands. These putative extended conformation strands should be pre-organized to have highly cooperative H-bonding domains along only one edge of the strand. These investigators will test if these peptides form homodimers via their H-bonding domains in a concentration dependent manner in aqueous buffer. Unlike natural peptides, no further H-bond mediated oligomerization is possible because the constrained dipeptide units lack amide N-H bonds in every other amino acid. The structure (parallel and/or antiparallel B-sheet dimer models) and the aggregation characteristics of these peptide homodimers will be studied to verify these hypotheses. The effect of three different constrained dipeptide units and varying amino acid side chains, peptide sequence, and peptide length on the homodimerization equilibrium will be studied to guide the rational design of extended conformation strands that preferentially bind alternative H-bonding sites. By mimicking the sequences of known amyloid B-peptide{l-40} (A-BETA) fibri/protofibril inhibitors, they will promote binding between these extended conformation strands and the growing edge of the fibrils/protofibrils that supposedly lead to Alzheimer's disease (AD). The investigators hypothesize that these extended conformation strands can cap off growing b-sheet rich fibrils/protofibrils and inhibit or reverse the uncontrolled aggregation of AD in vitro. It is currently disputed whether fibrils or protofibrils or something else are the actual cytotoxic species. Therefore, they will directly measure the cytoprotective activity of extended conformation strands using an established assay in vitro.