Inhibition of heparin-induced tau4RD aggregation by tryptanthrin and analogs, discovered by computationally-guided screening.

Tau4RD (5 μM) aggregation in the presence of 5 μM 3 kDa heparin, with 5 μM of TA, analogs 2–4, or vehicle control. (n = 3, error bands show mean +/-SEM.)

Effects of second and third generation TA analogs on heparin-induced tau4RD aggregation.

A) Structures of second generation TA analogs 5-17, with the table indicating substituents present at specified positions on the quinazoline ring for each analog. B) Tau4RD aggregation under standard conditions in the presence of second generation TA analog (5 μM) or vehicle control. The means (n = 3 or 4 technical replicates) are displayed for each trace. For visual clarity, error bands (mean +/-SEM) are displayed only for the vehicle trace. Other traces showed similar or lower technical variability. C) Structures of third generation analogs 19–30 derived from 16. D) Tau4RD aggregation in the presence of third generation TA analog (5 μM) or vehicle control. Traces and error bands are formatted as in panel B. Compounds 20, 24, and 25 (lower panel) were conducted at a DMSO concentration of 0.25% v/v, while all other experiments were conducted at 0.1% v/v DMSO.

TA analogs inhibit tau4RD aggregation in a substoichiometric, dose-dependent manner.

A) Tau4RD aggregation under standard conditions in the presence of varying concentrations of the third generation compound 27 or vehicle control (n = 3 or 4, mean displayed for all analog traces, S.E.M. error band displayed solely for the vehicle trace). F50 is defined as the fluorescence measured at the transition midpoint of the vehicle trace (t50), whether in the absence [F50(0)] or presence [F50(C)] of varying concentrations of compound. B) Dose-response curves of F50(C) normalized to F50(0) measured on the same plate for compounds 16 and 27, demonstrating nanomolar potency. Best-fit EC50 values are shown, with 95% confidence intervals estimated from resampling shown in parentheses. C) Schematic of resampling protocol used to estimate confidence intervals of best-fit EC50 values.

TA analogs are most active early in the aggregation process.

Tau4RD aggregation with delayed addition of 5 μM TA analog at t = 0, 2, 4, or 8 hours (n = 3, mean +/-SEM). Inhibition was observed when analogs were added early in the aggregation lag phase but not once the elongation phase began. This suggests that TA analogs preferentially inhibit primary nucleation.

Inhibition of seeded tau4RD aggregation by TA analogs.

A) Aggregation of 5 μM tau4RD and 5 μM 3 kDa heparin, seeded with 5% (w/w) pre-formed fibril in the presence of 5 μM TA analog or vehicle control. Traces indicate mean +/-SEM (n = 3 or 4). B) Seeded elongation measured over a range of compound concentrations, with linear fits (bold lines) used to estimate elongation rates. Traces indicate mean +/-SEM (n = 3 or 4). C) kseeded rates from B plotted against compound concentration and fit to Equation 2, with 95% C.I. estimated by resampling.

Proposed mechanism of action of tryptanthrin and its analogs.

Substoichiometric inhibition of early-stage aggregation indicates that TA analogs bind selectively and tightly to tau aggregation nuclei or their close precursors. Substoichiometric inhibition of seeded aggregation indicates that TA analogs can also inhibit subsesquent elongation or secondary nucleation, albeit with lower potency.