AWZ1066S, a highly specific anti- Wolbachia drug candidate for a short-course treatment of filariasis

Abstract

Onchocerciasis and lymphatic filariasis are two neglected tropical diseases that together affect ∼157 million people and inflict severe disability. Both diseases are caused by parasitic filarial nematodes with elimination efforts constrained by the lack of a safe drug that can kill the adult filaria (macrofilaricide). Previous proof-of-concept human trials have demonstrated that depleting >90% of the essential nematode endosymbiont bacterium, Wolbachia, using antibiotics, can lead to permanent sterilization of adult female parasites and a safe macrofilaricidal outcome. AWZ1066S is a highly specific anti-Wolbachia candidate selected through a lead optimization program focused on balancing efficacy, safety and drug metabolism/pharmacokinetic (DMPK) features of a thienopyrimidine/quinazoline scaffold derived from phenotypic screening. AWZ1066S shows superior efficacy to existing anti-Wolbachia therapies in validated preclinical models of infection and has DMPK characteristics that are compatible with a short therapeutic regimen of 7 days or less. This candidate molecule is well-positioned for onward development and has the potential to make a significant impact on communities affected by filariasis.

Keywords: anti-Wolbachia; drug discovery; lymphatic filariasis; macrofilaricide; onchocerciasis.

Conflict of interest statement

Conflict of interest statement: An international patent (WO2018134685) that includes chemical compounds described in this manuscript has been filed by the authors (W.D.H., F.B., M.J.T., P.M.O., and S.A.W.). Some authors (F.B., A. Siu, M.S., and F.G.) of this manuscript are full-time employees of Eisai, Inc.

Copyright © 2019 the Author(s). Published by PNAS.

Figures

Fig. 1.

Medicinal chemistry optimization to identify AWZ1066S, including in vitro potency and DMPK properties. (A) Progression of the thienopyrimidine screening hit to the azaquinazoline candidate AWZ1066S through multiparameter optimization to improve potency and DMPK properties. (B) AWZ1066 treatment causes a dose-dependent reduction of Wolbachia from Wolbachia-infected cells. Images are visual representations of cells treated with doxycycline (Top) and AWZ1066 (Bottom) at various concentrations. In the cell-based assay, C6/36 (wAlbB) cells are incubated with compounds and then stained with SYTO11 stain, which stains cell nuclei (large, bright areas; arrowhead) and Wolbachia (small, bright punctate staining; arrows) green. Automated capture and imaging analysis is conducted by Operetta and Harmony software (PerkinElmer) (six fields per well, two wells per concentration), which generates “percentage of infected cells” values (shown in white; typical vehicle-treated cells have 70 to 80% infection, while doxycycline at maximum effect has 10 to 15% infection) that are used to calculate the effective concentration of 50% inhibition of Wolbachia (EC50: doxycycline 20 nM; AWZ1066 2.6 nM). (C) Anti-Wolbachia activity of AWZ1066S in the microfilaria assay (EC50 121 nM) in comparison with the (R)-isomer, AWZ1066R (EC50 408 nM). Doxycycline was used as positive control (EC50 300 nM). Data are expressed as mean ± SD of five replicates for each concentration. (D) In vitro DMPK-related data of AWZ1066S presented in a radar plot. Six axes represent human hepatocyte clearance (µL⋅min−1 1 × 106 cells−1) (H. Hep. CL), human microsome clearance (µL⋅min−1⋅mg−1) (H. Mic. CL), LogD7.4 (LogD), aqueous solubility in pH 7.4 PBS buffer (µM) (Aq. Sol.), permeability in LLC-PK1 cell assay (1 × 106 cm/s) (LLC Papp), and human plasma protein binding (%) (H. PPB). The light blue-shaded area indicates the desired DMPK property target ranges for oral administration.

Fig. 2.

In vivo anti-Wolbachia efficacy of AWZ1066S. (A and B) Anti-Wolbachia efficacy against female adult-stage B. malayi in SCID mice (A) or L. sigmodontis in gerbils (B) after doxycycline (DOX), minocycline (MIN), or AWZ1066S oral treatments at the indicated doses. Tail and whisker plots are minimum, 25th, median, 75th, and maximum Wolbachia wsp or ftsz copy number per female worm in samples of 5 to 24 worms derived from groups of four to nine animals from an individual experiment. Significant differences compared with vehicle/untreated are indicated ****P < 0.0001, **P < 0.01 (Kruskal–Wallis with Dunn’s tests for intergroup Wolbachia variation). bid, twice a day. (C) Longitudinal effects on circulating L. sigmodontis microfilaremias −1 to 18 wk after DOX or AWZ1066S oral treatments at the indicated doses in gerbils. (D) Result of PK/PD Monte Carlo prediction of clinical activity of AWZ1066S when dosed at 600 mg given predicted PK properties in humans and established in vitro PD properties against Wolbachia.

Fig. 3.

Unique mode of action of AWZ1066S was indicated by its fast kill rate and was investigated through a proteomic target identification approach using two photoreactive chemical probes. (A) The Wolbachia depletion rates of AWZ1066S measured for three different treatment lengths (1, 2, and 6 d) in the B. malayi microfilaria time-kill assay. Doxycycline, minocycline, moxifloxacin, and rifampicin were used as positive controls (only doxycycline for day 1). Data are expressed as mean ± SD of five replicates for each time point. (B) Two photoreactive chemical probes used for target identification of AWZ1066S in Wolbachia. Structures (bifunctional, photoreactive, and clickable side chains are highlighted in red) and their anti-Wolbachia activity in the cell assay (n = 3). (C) Fifty-three putative protein targets of photoreactive probes 1 and 2 assigned (as shown in the Venn diagram) based on statistical testing [right-sided t test, permutation-based false discovery rate = 0.001, Artificial within groups variance (S0) = 3] of label-free quantification (LFQ) intensities measured in samples derived from probe-treated cells and control, DMSO-treated cells (n = 3). Protein targets presented in the left-hand block are from W. pipientis and in the right-hand block are from A. albopictus. Numbers within the heatmap legend represent Log2 LFQ t test differences. Protein ID codes above the heatmap given in red indicate 22 proteins identified as common targets of both probes.