Fragment-based virtual screening identifies a first-in-class preclinical drug candidate for Huntington's disease

Abstract

Currently, there are no therapies available to modify the disease progression of Huntington's disease (HD). Recent clinical trial failures of antisense oligonucleotide candidates in HD have demonstrated the need for new therapeutic approaches. Here, we developed a novel in-silico fragment scanning approach across the surface of mutant huntingtin (mHTT) polyQ and predicted four hit compounds. Two rounds of compound analoging using a strategy of testing structurally similar compounds in an affinity assay rapidly identified GLYN122. In vitro, GLYN122 directly binds and reduces mHTT and induces autophagy in neurons. In vivo, our results confirm that GLYN122 can reduce mHTT in the cortex and striatum of the R/2 mouse model of Huntington's disease and subsequently improve motor symptoms. Thus, the in-vivo pharmacology profile of GLYN122 is a potential new preclinical candidate for the treatment of HD.

Figure 1

In the first step, we conducted an in-silico-based drug screening with a deep artificial neuronal network. Depicted are the selection of 67 small molecules docked on the surface of the polyQ tract peptide. 40 compounds were tested for their actual binding affinity to mutant huntingtin (mHTT) with Biacore, a system using surface plasmon resonance. Next, compounds were tested in an in-vitro mHTT assay and selected based on the ability of the compounds to lower mHTT and overcome an artificial blood–brain barrier in the PAMPA BBB permeability assay. Finally, we demonstrated in the R6/2 Huntington disease (HD) mouse model that the preclinical candidate GLYN122 can improve motor symptoms as measured with rotarod. Created with BioRender.com.

Figure 2

(A) Chemical structures of the 4 hits identified from the Biacore screening and (B) selection of the preclinical candidate (GLYN122) from the analogue expansion of one of the hit compounds (GLYN026) derived from two adjacent chemical fragments obtained from virtual protein surface scanning.

Figure 3

The concentration of GLYN122 that penetrated the blood–brain barrier of male CD1 mice (n = 3) treated with GLYN122 at a dose of 33.3 mg/kg body weight. Dashed line: brain, straight line: plasma.

Figure 4

The effects of chronic administration of GLYN122 (33 mg/kg) on rotarod latency of female R6/2 mice at 4 to 10 weeks of age. Data are presented as percentage from 4-week pre-treatment baseline (mean + SEM) (WT Vehicle, n = 5; R6/2 Vehicle, n = 5; R6/2 GLYN122 33 mg/kg, n = 5). R6/2 GLYN122 33 mg/kg vs. R6/2 Vehicle. * p ≤ 0.01, ** p ≤ 0.001.

Figure 5

Confocal fluorescence images of the striatum region from WT vehicle (5A,5D), TG vehicle (5B,5E), and TG GLYN122 (5C,5F) samples labeled with EM48 mHTT antibody and detected with a fluorophore in the 488 nm channel. Grayscale images were pseudocolored with LUT “green” (top row) and “Green Fire Blue” (bottom row) in Fiji to visualize the gradient in signal intensities. For the latter LUT, increasing signal intensities are represented from blue to green to white. Overall, the strongest mHTT signal was visible in TG vehicles, while both WT vehicles and TG GLYN122 displayed decreased signals. Scale bars = 10 µm. WT wildtype, veh vehicle, TG transgenic. LUT Lookup Table.

Figure 6

Diffuse mHTT reduction in the cortex (6A,6B) and striatum (6C,6D). Compared to a vehicle (veh), administration of GLYN122 reduced diffuse mHTT in the cortex by 40% in the nucleus (6A) and by 45% in the cytoplasm of treated TG mice (* p = 0.01 and p = 0.02, respectively). Likewise, GLYN122 reduced diffuse mHTT in the striatum by 40% in the nucleus (6C) and 43% in the cytoplasm (6D) compared to vehicle (veh) of treated TG mice (p = 0.14 and p = 0.06, respectively). The points on the graph represent different samples of the same animals.

Figure 7

Correlation between rotarod and diffuse mHTT in R6/2 mice treated with GLYN122. Pearson r: − 0.8, p = 0.01, n = 9 (all transgenic female mice), confidence interval of r =  − 0.9561 to − 0.2896 for nuclear diffuse mHTT. TG, transgenic R6/2 mice; mHTT mutated huntingtin.