Currently, definitive AD diagnosis is made post-mortem by the appearance of extracellular amyloid-beta peptide plaques and intracellular tau tangles. These lesions are proteinaceous aggregates in the form of in-register cross-beta-sheet fibrils. For pre-mortem diagnosis, sensitivity and specificity are 80% and 70%, respectively. CSF biomarkers and radiopharmaceutical imaging could improve AD diagnostic accuracy. The lumbar puncture procedure needed for CSF sampling is more painful and risky than whole brain PET imaging. Although imaging of both amyloid-beta and tau lesions could provide diagnostic information, tau-selective imaging agents should be superior to amyloid-beta-selective or pan-binding agents because the appearance of tangles in certain brain regions precedes plaque formation by decades, tangle presence correlates better than plaques with neuronal loss, and tangles develop in a more stereotypical spatial pattern over time than do plaques. Lewy bodies, seen in PD and LBD, are composed mainly of alpha-synuclein protein, and could interfere with tau lesion imaging. Therefore, compounds are needed which bind tau, but not amyloid-beta or alpha-synuclein filaments.
ThS and ThT are well-characterized probes of cross-beta-sheet structures and are commonly used in displacement assays to identify fibril-binding compounds. ThS and ThT displacement from recombinant tau filaments and PHFs was similar for several compound classes. Multimodal compound binding to 0N4R tau fibrils was observed with ThT displacement by BTA-2. Certain BTAs and oxindoles displaced ThT from 2N4R tau, 0N4R tau, and alpha-synuclein fibrils more potently than from 1N3R tau, amyloid-beta1-40, and amyloid-beta1-42 filaments, despite frequent incomplete ThT displacement by BTAs at high compound concentration. However, a few potent (AC50 < 10 nM) BTA compounds displaced ThT from 2N4R and 0N4R tau isoforms ~10-fold better than from alpha-synuclein. BTA compound potency was proportional to percentage of observed ThT displacement for all fibril types.
Binding potency of BTAs and oxindoles was increased by maximizing the number of intervening aromatic electrons between the electron donating and electron acceptor groups (polymethine system), with mono- and dimethyl amines being the strongest donors. For oxindoles, the presence of a chlorine or bromine atom at position 5 of the oxindole ring and/or having a double ring system for the aryl group increased potency. Increased compound width, such as “E-form” oxindoles and BTA 23, seemed to increase compound selectivity for 2N4R and 0N4R tau fibrils over amyloid-beta fibrils. I attribute this selectivity to double glycine channels found in tau fibrils that are not present in amyloid-beta filaments.
Future BTAs and oxindoles should be wide, have large aryl groups (such as double ring systems), and contain mono- or di-methyl amino donor groups in positions that maximize the number of polarizable electrons. QSAR methods may uncover other important compound design factors. Additional future work for BTA and oxindole compound development that must precede imaging agent trials in humans includes displacement of radiolabeled probes in a filter-based assay, staining post-mortem AD patient brain tissue sections with radiolabeled BTA and oxindole compounds, identifying appropriate 18F labeling methods, and testing of 18F-labeled imaging agents in AD animal models for appropriate PK/PD properties and PET efficacy.