Identifying new compounds active against Giardia duodenalis

Abstract

Giardia are a genus of enteric pathogens consisting of at least six species (Monis et al. 2009), of which one, Giardia duodenalis, infects humans (Heyworth 2016). In addition to humans, G. duodenalis parasites infect other mammals, which may act as reservoirs for human infection (Traub et al. 2004; Yaoyu and Xiao 2011; Abeywardena et al. 2015; Sroka et al. 2015; Strkolcova et al. 2015; Heyworth 2016). In humans Giardia infection can be asymptomatic, however all infected hosts shed cysts and can transmit parasites (Oliveira-Arbex et al. 2016; Figgatt et al. 2017). Giardia infection can cause giardiasis, a diarrhoeal disease with a variety of clinical manifestations (Wolfe 1992; Homan and Mank 2001; Sahagun et al. 2008; Nielsen et al. 2014). An estimated 180 million symptomatic human Giardia infections occur every year (Kirk et al. 2015), and treatment is reliant on a small number of chemotherapeutic classes, all of which are associated with liabilities. Liabilities include but are not limited to; poor treatment efficacies, long treatment courses and side-effects which impact compliance (Cina et al. 1996; Wright et al. 2003; Escobedo and Cimerman 2007; Nabarro et al. 2015). Growing parasite resistance to the first-line treatment drugs, the 5-nitroimidazoles, is also a concern (Nabarro et al. 2015). The aim of the current study was to pave the way towards improved treatment options for giardiasis by identifying new lead compounds for drug development, and to further examine the activity of these compounds. To achieve this, a new anti-Giardia activity assay to assess parasite growth in micro-titre plates under microaerobic (3 % O2) conditions was developed. This image-based assay uses bright-field microscopy paired with digital phase-contrast microscopy and supervised machine learning software, PhenoLOGIC and Harmony® (Perkin-Elmer, USA) to differentiate and enumerate parasites. Growth assessment does not require cell-staining or a genetically modified parasite line, thus it can assess the growth of any established Giardia line, at multiple time-points which are distinct advantages over other assays currently used in the Giardia field. Importantly, this assay gives 50 % inhibitory (IC50) values for control compounds metronidazole (IC50; 2.7 ± 0.7 µM), albendazole (54 ± 5 nM) and furazolidone (200 ± 90 nM) , against BRIS/91/HEPU/1279 at 48 h consistent with those previously reported by others (Edlind et al. 1990; Cedillo-rivera et al. 2002; Hounkong et al. 2011; Tejman-Yarden et al. 2011). The validated image-based assay was used to screen a sub-set of Compound Australia’s Open-access Scaffold Library for anti-Giardia activity. A total of 2451 compounds (two per scaffold) were screened at 10 µM. Forty-one compounds (1.7 % hit rate) were validated as having anti-Giardia activity (>50 % inhibition at 48 h) in these assays. Secondary testing of hit compounds to determine IC50 values against Giardia and neonatal foreskin fibroblasts (NFF) identified five compounds with IC50 values 10 fold selectivity for parasites over mammalian cells. Rational selection based on selective activity, chemical novelty and chemical liabilities identified seven hit series for further investigation. Compounds within these series (196 total; ~28/series) were then assessed to examine structure activity relationships (SARs) and prioritize hit series for development. Analogues of particularly potent and selective series were also synthesized by collaborators and assessed for anti-parasitic activity and selectivity. The most promising hit was three orders of magnitude more potent than the current first-line anti-Giardia treatment drug, metronidazole (SN00798527; series CL9569; 48 h IC50 9 nM vs. metronidazole 48 h IC50 3 µM), with a selectivity index (SI) of >11,000. Importantly, this activity was maintained against multiple Giardia isolates encompassing both human infecting G. duodenalis assemblages (A and B) and against metronidazole resistant parasites. These data suggest that SN00798527 has a different mode of action to metronidazole and that cross-resistance with the 5-nitroimidazoles is unlikely, and that the molecule is likely to be equally effective against both human infecting genotypes. Preliminary in vivo data suggests that this compound is well tolerated in Swiss mice, with no toxicity seen at oral doses of up to 0.7 mg/kg (10x the calculated therapeutic dose (CTD) for this compound). Preliminary data also suggest that SN00798527 is active in a murine giardiasis model. Neonate Swiss mice dosed daily for three days with orally administered SN00798527 (0.7 mg/kg 10xCTD) harboured significantly reduced parasite loads (73 % reduction in trophozoite load and 99 % reduction in cyst load) compared to untreated control mice. Taken together these data highlight the in vivo potential of series CL9596 and suggest further in vivo trials and mode of action studies are warranted. Lead compounds within two additional compound series (SC003542 and CL9406), SN00776497 and SN00797640 also demonstrated promising in vitro activity (48 h IC50 values 183 and 23 nM and SI of 291-343 and 24-90, respectively) that was consistent against multiple isolates encompassing both human infecting G. duodenalis genotypes and against metronidazole resistant parasites. While the timeframe of this project did not permit the in vivo anti-Giardia activity of these molecules to be evaluated, toxicity studies in Swiss mice also demonstrated these compounds to be safe at doses of up to 2 mg/kg and 5 mg/kg respectively (10x CTD).