Identification of a Thyroid Hormone Derivative as a Pleiotropic Agent for the Treatment of Alzheimer's Disease

Abstract

The identification of effective pharmacological tools for Alzheimer's disease (AD) represents one of the main challenges for therapeutic discovery. Due to the variety of pathological processes associated with AD, a promising route for pharmacological intervention involves the development of new chemical entities that can restore cellular homeostasis. To investigate this strategy, we designed and synthetized SG2, a compound related to the thyroid hormone thyroxine, that shares a pleiotropic activity with its endogenous parent compound, including autophagic flux promotion, neuroprotection, and metabolic reprogramming. We demonstrate herein that SG2 acts in a pleiotropic manner to induce recovery in a C. elegans model of AD based on the overexpression of Aβ42 and improves learning abilities in the 5XFAD mouse model of AD. Further, in vitro ADME-Tox profiling and toxicological studies in zebrafish confirmed the low toxicity of this compound, which represents a chemical starting point for AD drug development.

Keywords: 5XFAD mice; Alzheimer’s disease; C. elegans; autophagy; drug discovery; polypharmacology; zebrafish.

Figure 1

SG2 recovers fitness in a C. elegans model of AD. SG2 was administered to the worms following two different treatment regimes, namely early treatment at D0 (A) or late treatment at D4 (B). SG2 dosed at 5, 10 and 20 µM showed significant dose-dependent recovery effect, expressed as body bends/min, on AD worms (A,B). Error bars represent the standard error of the mean (SEM). For statistical tests, One-Way ANOVA was used. p ≤ 0.0001 (****).

Figure 2

SG2 does not directly inhibit Aβ42 aggregation in in vitro assays. The aggregation of 2 µM Aβ42 was performed in 20 mM sodium phosphate buffer with 0.2 mM EDTA at pH 8. The amyloid-specific dye thioflavin T (ThT) at 20 µM was used as a fluorescence probe to monitor the progression of the aggregation process as depicted in panel (A). SG2 was added to Aβ42 at different concentrations in presence or in absence of preformed fibrils (seeds), as shown in each panel. In the seeded assays, 2% or 50% of preformed fibrils were added to the solution before the start of the aggregation process. Varying the seeds concentration in the assay solution allows to monitor different steps of the Aβ42 aggregation process. Results are presented as normalized fluorescence intensity, which is indicative of the Aβ42 formation, over time. SG2 has a slight effect on primary enucleation (unseeded assay, panel (B)), while does not affect the secondary enucleation (panel (C)) or the elongation step (panel (D)).

Figure 3

SG2 enhances autophagy in a C. elegans model of AD. (A) GFP structures were counted in 3 different conditions (Fed: 1% DMSO; Starved: 1% DMSO and 6 h of starvation; Treated: 10 µM SG2 in 1% DMSO) of approximately 25–30 worms from two different biological experiments. Bars represent average number of GFP structures normalized to control (Fed) + SEM. For statistical tests, One-Way ANOVA was used p < 0.001 ***. (B) Representative pictures of big muscles cells presenting LGG-1:GFP puncta with white arrows highlighting LGG-1::GFP structures.

Figure 4

SG2 shows a safe toxicological profile in an in vitro panel. A panel of toxicological assays were performed to test (A) SG2 behaviour over cytotoxicity (U2-OS, HEK293, MCF-7, hTERT), (B) epigenetic modulation (HDAC6, HDAC8, SIRT7), and off-target liability (PDE4C1, Aurora B kinase). All assays were performed at 10 µM in triplicate. Raw data were normalized over positive and negative control and reported as percentage of inhibition. In no assay SG2 showed a percentage of inhibition higher than 40%, where in this profiling a value < 50% is generally considered acceptable.

Figure 5

SG2 does not show toxicological effects in zebrafish. (A) SG2 at 5 µM was administered to zebrafish embryos at 4 h post fertilization (hpf); the experimental workflow was created with BioRender.com( accessed on 5 May, 2021). All the experiments reported show non-significant values compared to controls, proving a safe impact of SG2 in zebrafish. (B,C) Mortality rate was calculated until 48 hpf (B) (controls N = 545, vehicle N = 281, SG2 N = 224) and no morphological anomaly was observed at 5 days post fertilization (dpf) (C). (D) Cardiotoxicity was evaluated measuring heart rate at 3 dpf. (E,F) Neurological development was evaluated at 30 hpf with burst activity analysis (E) and at 5 dpf with locomotor measurements considering velocity and distance covered (F) (controls N = 366, vehicle N = 212, SG2 N = 246). All experiments were performed at least in triplicate. Error bars represent the SEM. For statistical tests non-parametric one-tailed Mann-Whitney rank sum test was used.

Figure 6

SG2 improves learning ability in the 5xFAD mouse model of AD. (A) Analysis of MWM learning curve showed no significant differences between experimental groups (repeated measures ANOVA, Time F (3,105) = 21.57, p < 0.0001, Treatment F (3,35) = 0.3141, p = 0.82, Time × Treatment F (9,105) = 1.553, p = 0.14). (B) Learning index analysis demonstrated an improved performance of Tg-SG2 mice compared to Tg-Sal to find the hidden platform. For statistical tests, One-Way ANOVA was used p < 0.02 * (C) The time needed to reach the platform area (C) and the number of entries in the target area (D) were not statistically different between experimental groups. (E) The analysis of search strategy showed an increase in spatially targeted strategies in Tg-SG2 mice compared to Tg-Sal mice (Fisher’s test). Data are shown as mean ± SEM. Experimental groups: Wt-Sal, n = 10; Wt-SG2, n = 10; Tg-Sal, n = 10: Tg-SG2, n = 9.