Antigiardial Activity of Novel Guanidine Compounds

Abstract

From four focused compound libraries based on the known anticoccidial agent robenidine, 44 compounds total were synthesised and screened for antigiardial activity. All active compounds were counter‐screened for antibiotic and cytotoxic action. Of the analogues examined, 21 displayed IC₅₀<5 μM, seven with IC₅₀<1.0 μM. Most active were 2,2′‐bis{[4‐(trifluoromethoxy)phenyl]methylene}carbonimidic dihydrazide hydrochloride (30), 2,2′‐bis{[4‐(trifluoromethylsulfanyl)phenyl]methylene}carbonimidic dihydrazide hydrochloride (32), and 2,2′‐bis[(2‐bromo‐4,5‐dimethoxyphenyl)methylene]carbonimidic dihydrazide hydrochloride (41) with IC₅₀=0.2 μM. The maximal observed activity was a 5 h IC₅₀ value of 0.2 μM for 41. The clinically used metronidazole was inactive at this timepoint at a concentration of 25 μM. Robenidine off‐target effects at bacteria and cell line toxicity were removed. Analogue 41 was well tolerated in mice treated orally (100 mg/kg). Following 5 h treatment with 41, no Giardia regrowth was noted after 48 h.

Rapid and complete killing of Giardia trophozoites: From four focused compound libraries based on the known anticoccidial agent robenidine, 44 compounds were synthesised and screened for antigiardial activity. All active compounds were counter‐screened for antibiotic and cytotoxic action. Of the analogues examined, 21 displayed IC₅₀ values less than 5 μM, seven with IC₅₀ values less than 1.0 μM.

Introduction

Giardia duodenalis, a bi‐nucleate protozoan pathogen, is the most common enteric human parasitic pathogen known, causing up to 1 billion human infections annually.[ ¹ , ² , ³ , ⁴ ] Infections are most common in developing nations, but are also prevalent in the developed world. Giardia has been added to the World Health Organization neglected diseases initiative. ^[5] Giardia infection occurs via cyst ingestion through a faecal‐oral route or via contaminated food or water.[ ⁵ , ⁶ ] It results in a mal‐absorptive gastrointestinal disease with acute, chronic and at times re‐occurring symptoms including diarrhoea, bloating and abdominal cramping. ^[7] Persistent infection, especially in children and immunocompromised hosts, results in long term effects including malnutrition, developmental delay and failure to thrive syndrome. ^[8]

Drug treatment most commonly uses the nitroimidazoles, nitrothiazole, nitrofuran, acridine, benzimidazole, and aminoglycoside compound classes. ^[8] The most frequently used nitroimidazoles, metronidazole (1) and tinidazole (2), show an 80–90 % success; while albendazole (3), a benzimidazole, has a reported efficacy of 62–95 %. Treatment failures with these drugs are common with side effects including genotoxicity, possible carcinogenesis, nausea, fatigue and general malaise.[ ⁸ , ⁹ , ¹⁰ ] Disturbingly, resistant organisms have been reported for all the commonly used drugs.[ ⁹ , ¹⁰ , ¹¹ ] This combination of limitations highlights a pressing need for new treatments.

Current antigiardial drug discovery efforts have focused in on the small molecule inhibition of a wide range of protein targets. These span pyruvate‐ferredoxin oxidoreductase, nitroreductase 1 and thioredoxin reductase to activate nitroimidazoles;[ ¹² , ¹³ , ¹⁴ ] NADH oxidase for activation of furazolidone; ^[15] auranofin inhibits thioredoxin reductase; ^[15] fumagillin inhibits methionine amino‐peptidase; ^[16] orlistat inhibits gastric lipase; ^[17] proton‐pump inhibitors have been investigated, with omeprazole inhibiting triosephosphate isomerase, a critical enzyme involved in glucose and glycogen metabolism;[ ¹⁸ , ¹⁹ ] disulfiram inhibits acetaldehyde dehydrogenase;[ ²⁰ , ²¹ ] NBDHEX inhibits glutathione S‐transferase. ^[22] Other potential targets include antioxidant and metabolic enzymes[ ²³ , ²⁴ , ²⁵ , ²⁶ , ²⁷ , ²⁸ , ²⁹ ] and protein kinases, with the reduced number of core Giardia kinases suggesting limited redundancy, and thus potentially high efficacy. ^[30]

Herein we report on the discovery and development of novel antigiardial agents based on the anticoccicidal agent, robenidine (4; Figure 1).[ ¹⁰ , ³¹ ]

Figure 1.

Chemical structures of the clinically used metronidazole (1) tinidazole (2), albendazole (3) and lead robenidine (4).

Results and Discussion

Preliminary screening of robenidine (4) against Giardia trophozoites in vitro revealed a 5 h IC₅₀ of 0.9 μM and a 24 h MIC (minimum inhibitory concentration) of 2.8 μM. Clinically used metronidazole (1) was inactive at 25 μM at 5 h, but showed a 24 h IC₅₀ of 3.8 μM (Table 1).[ ³² , ³³ , ³⁴ ]

Table 1.

Inhibition of Giardia duodenalis by Library 1 Robenidine analogues possessing mono‐halogenated aromatic rings (4–13).^[a]

		Antigiardial activity
Compound	R	MIC [μM] (24 h)	IC50 [μM] (24 h)	Antibacterial activity (Y/N)[b]	Toxicity (Y/N) (% growth control)[c]	Selectivity ratio[d]
Metronidazole	–	8.3	3.8	N	N	>6.6
Robenidine 4	4‐Cl	2.8	0.9	Y	Y (40.7±22.6)	>27.8
5	H	–	–	N	–	–
6	4‐Br	8.3	1.7	N	N (97.1±1.6)	>14.7
7	4‐F	12.5	6.52	Y	–	–
8	3‐Cl	6.25	3.5	Y	–	–
9	3‐Br	25	2.9	Y	–	–
10	3‐F	>25	>25	Y	–	–
11	2‐Br	25	0.8	N	Y (79.1±2.3)	>31.3
12	2‐Cl	25	12.37	N	–	–
13	2‐F	>25	>25	N	N[e] (93.9±1.9)	–

[a] Toxicity (CaCo‐2 and Vero cells) and antimicrobial assays performed at 25 μM; MIC: minimum inhibitory concentration; ‘–’ not tested, as inhibitory activity at 25 μM<50 %. [b] MSRA and VRE. [c] Percent CaCo‐2 cell growth (25 μM); compounds are indicated as toxic if >20 % cells are affected. [d] Higher ratios indicate a more selective compound. [e] Toxicity assessed with Vero cells, not CaCo‐2 cells.

While potent, robenidine (4) has off target antibacterial activity and cell line toxicity. ^[35] We sought to address these concerns through the synthesis and subsequent biological screening of focused compound libraries. Robenidine analogues were rapidly accessed through the condensation of a series of benzaldehydes and N,N′‐diaminoguanidine hydrochloride (Scheme 1), afforded rapid access to the desired analogues (Table 1). Retention of the N,N′‐diaminoguanidine core enabled the SAR activity of the pendent aromatic moieties to be examined in this initial study, as the nature of the robenidine antigiardial activity (what protein target) is unknown.

Scheme 1.

Reagents and conditions: (i) aldehyde (see Table 1 for detail), EtOH, reflux.

Limiting Library 1 to mono‐substituted halogenated benzaldehydes afforded nine analogues (5–13) in good to excellent yields. Of these, seven returned antigiardia IC₅₀ values of <25 μM; three of which displayed no antibacterial activity against methicillin‐resistant Staphylococcus aureus (MRSA), vancomycin‐resistant Enterococcus (VRE) or Escherichia coli (6, 11 and 12); with 4‐Br 6 also showing no cytotoxicity against CaCo‐2 or Vero cells at <10 % cell death at 25 μM. No activity against gram negative bacteria was noted (not shown). In this first library, brominated analogues (6, 9 and 11) displayed the highest levels of Giardia activity, but activity was only marginally enhanced relative to the corresponding Cl‐analogues (4, 8 and 12). Isosteric modification to ‘F’ saw moderate levels of activity (with 2‐F 13, inactive IC₅₀>25 μM).

An extension of our structure activity relationship (SAR) evaluation saw the introduction of alkyl substituents with the synthesis (as per Scheme 1) and evaluation of 14–23. The screening data for these Library 2 analogues is presented in Table 2.

Table 2.

Inhibition Giardia duodenalis metabolism by Library 2 Robenidine analogues possessing alkyl and aromatic substituents (14–23).^[a]

		Antigiardial activity
Compound	R	MIC [μM] (24 h)	IC50 [μM] (24 h)	Antibacterial activity (Y/N)[b]	Toxicity (Y/N) (% growth control)[c]	Selectivity ratio[d]
Metronidazole	–	8.3	3.8	N	N	>6.6
14	4‐CH3	25	6.98‐21.6	Y	–	–
15	3‐CH3	6.25	4.9	Y	–	–
16	2‐CH3	25	5.62‐12.98	Y	–	–
17	4‐Ph	2.8	0.4	Y	–	–
18	2‐Ph	>25	2.8	N	Y (75.4±3.4)	>8.9
19	4‐(CH2)2CH3	8.3	0.8	Y	–	–
20	4‐(CH2)3CH3	8.3	1.9	Y	–	‐
21	4‐CH(CH3)2	6.25	1.9	Y	–	–
22	4‐C(CH3)3	8.3	3.2	Y	–	–
23	4‐CCH	25	1.2	Y	–	–

[a] Toxicity (CaCo‐2 and Vero cells) and antimicrobial assays performed at 25 μM; MIC: minimum inhibitory concentration; ‘–’ not tested, as inhibitory activity at 25 μM<50 %. [b] MSRA and VRE. [c] Percent CaCo‐2 cell growth (25 μM); compounds are indicated as toxic if >20 % cells are affected. [d] Higher ratios indicate a more selective compound.

In each instance the alkyl substituted Library 2 analogues displayed Giardia activity <25 μM. The highest levels of activity were observed with the more hydrophobic substituent with 4‐Ph 17 (IC₅₀=0.4 μM) more active than 2‐Ph 18 (IC₅₀=2.8 μM), while all other alkyl substituents (excepting the −CH₃ analogues 14–16) displayed excellent Giardia inhibition at 3.2–0.8 μM (22 and 19 respectively). Acetylenic 23 was well tolerated (IC₅₀=1.2 μM). Given the high level of antigiardial activity, each compound was then evaluated for potential antibiotic activity with only 2‐Ph 18 antibiotic inactive, and thus subjected to toxicity assessment with Vero and CaCo‐2 cells. In this instance, 18 displayed toxicity at the 25 μM concentration evaluated. Despite this finding, our data supports a possible dissection of the biological activity profile of this series of compounds, and we note the favourable antigiardial outcome with 2‐disposed analogues (2‐Br 11 and 2‐Ph 18).

The introduction of a polar or heteroaromatic linked hydrophobic moieties with Library 3 (24–33) was largely detrimental to antigiardial activity with only 4‐OCF₃ 30, 4‐SCH₃ 31, 4‐SCF₃ 32 and 4‐N(CH₃)₂ 33 returning IC₅₀ values<2 μM (0.2, 0.9, 0.2 and 1.1 μM, respectively), but with concomitant activity against gram negative bacteria (Table 3). With the observed antibiotic activity, these analogues were not assessed for their toxicity profile.

Table 3.

Inhibition Giardia duodenalis metabolism by Library 3 Robenidine analogues possessing heteroatom‐substituted aromatic rings (24–33).^[a]

		Antigiardial activity
Compound	R	MLC [μM] (24 h)	IC50 [μM] (24 h)	Antibacterial activity (Y/N)[b]	Toxicity (Y/N) (% growth control)[c]	Selectivity ratio[d]
Metronidazole	–	8.3	3.8	N	N	>6.6
24	4‐OCH3	12.5	5.99	N	Y[e]	>4.2
25	3‐OCH3	>25	>25	Y	–	–
26	2‐OCH3	>25	>25	Y	–	–
27	4‐OH	>25	>25	Y	–	–
28	3‐OH	>25	>25	Y	–	–
29	2‐OH	>25	>25	Y	–	–
30	4‐OCF3	2.8	0.2	Y	–	–
31	4‐SCH3	8.3	0.9	Y	–	–
32	4‐SCF3	2.8	0.2	Y	–	–
33	4‐N(CH3)2	25	1.1	Y	–	–

[a] Toxicity (CaCo‐2 and Vero cells) and antimicrobial assays performed at 25 μM; MLC: minimum lethal concentration; ‘–’ not tested, as inhibitory activity at 25 μM<50 %. [b] MSRA and VRE. [c] Percent CaCo‐2 cell growth (25 μM); compounds are indicated as toxic if >20 % cells are affected. [d] Higher ratios indicate a more selective compound. [e] Toxicity assessed with Vero cells, not CaCo‐2 cells.

Further library development explored the introduction of additional ring substituents with Library 4. The screening outcome for these analogues is presented in Table 4. As speed of action might be anticipated to influence uptake and efficacy of these agents, we explored the time to effective action, noting that essentially 100 % eradication of the Giardia trophozoites was achieved with some analogues in as little as 5 h. As such we determined the efficacy of Library 4 at a 5 h timepoint.

Table 4.

Inhibition of Giardia duodenalis metabolism by Library 4 Robenidine analogues possessing di‐, tri‐ and poly‐substituted aromatic rings (34–47).^[a]

		Antigiardial activity
Compound	R	MLC [μM] (24 h)	IC50 [μM] (5 h)	Antibacterial activity (Y/N)[b]	Toxicity (Y/N) (% growth control)[c]	Selectivity ratio[d]
34	2,4‐Cl	3.13	2.49	N	Y[e] (55.8±4.3)	>10.0
35	3,5‐Cl	25	11.6	N	–	–
36	2,6‐Cl	–	–	N	–	–
37	2,5‐F	–	–	N	–	–
38	3,4‐F	8.3	1.2	Y	–	–
39	2,3,4,5,6‐F	–	–	N	–	–
40	2‐NHCOCH3, 4‐Cl	25	0.8	N	Y (83.1±4.4)	>31.3
41	2‐Br, 4,5‐OCH3	2.8	0.2	N	N	>125
42	3‐Br, 4,5‐OCH3	3.13	2.7	N	Y[e]	>9.3
43	2‐OH, 4‐N(CH3)2	–	–	Y	–	–
44	3,4‐OCH3	>25	24.9	N	–	–
45	3‐NO2, 4‐OH	–	–	N	–	–
46	3‐OH, 4‐OCH3	–	–	N	–	–
47	3‐OCH3, 4‐OH	>25	>25	Y	–	–

[a] Toxicity (CaCo‐2 and Vero cells) and antimicrobial assays performed at 25 μM; MLC: minimum lethal concentration; ‘–’ not tested, as inhibitory activity at 25 μM<50 %. [b] MSRA and VRE. [c] Percent CaCo‐2 cell growth (25 μM); compounds are indicated as toxic if >10 % cells are affected. [d] Higher ratios indicate a more selective compound. [e] Toxicity assessed with Vero cells, not CaCo‐2 cells.

Of the 14 Library 4 analogues developed, seven (34, 35, 38, 40–42 and 44) returned antigiardial IC₅₀ values of ≤25 μM (Table 4 and ESI†). The introduction of a second −Cl moiety (c.f. 34 vs 4) was detrimental to activity, but the 3,4‐difluoro 38 was active (IC₅₀=1.2 μM). However, the equivalent 2,5‐difluoro 37 and pentafluoro 39 were inactive. Other variations in the phenyl ring substituent level and pattern were largely detrimental to activity (ESI†). Of Library 4, only 41 was devoid of antibacterial activity and cell toxicity, and fortuitously this was the most active analogue with a 5 h IC₅₀ of 0.2 μM.

This combined with the lack of antibacterial activity and cell toxicity differentiates 41 from 4 as an antigiardial development lead. To further investigate the validity of this compound series as leads in the development of potential antigiardial treatments, disubstituted 41 was examined in a mouse model of Giardia infection.

Our preliminary animal studies revealed no adverse effects with mice treated per os with 100 mg/kg 41 per day for 3 days, strongly suggesting that 41 is a viable lead development candidate. Moreover, in washout experiments, the removal of 4 and 41 after Giardia death did not result in Giardia regrowth after 48 hour (Figure 2).

Figure 2.

Growth recovery of Giardia duodenalis after exposure to selected analogues for 5 h. ‐λ‐ control (no compound); ‐○‐ metronidazole (1); ‐ν‐ Robenidine (4); ‐υ‐(40); ‐◊‐(6); ‐▵‐ (18); ‐ (11); ‐←‐ (41) Cell numbers were determined 24 and 48 h post compound removal.

Conclusions

From the analogues noted herein, 6, 11, 13, 40 and 41 were identified as potential antigiardial development candidates. Analogue 41 displays the most promising activity and safety profile of the analogues developed herein. All analogues showed the expected high levels of antigiardial activity.

The initial limitations of 4 as an antigiardial agent have been overcome through the development of 41, with a 5‐fold potency increase (5 h IC₅₀=0.2 μM), no antibacterial activity and no observed cytotoxicity in the systems examined. This activity also contrasts the clinically used metronidazole with known antibacterial activity, slow onset of activity (24 h vs 5 h) and 20‐fold lower antigiardial activity. Indeed, 41 shows marked impact on Giardia trophozoites after only 1–2 h (data not shown).

As yet the drug target of these analogues remains undetermined. However, in our antibiotic studies with related compounds we observed that robenidine analogues affected the bacterial cell wall.[ ³¹ , ³² , ³⁶ , ³⁷ , ³⁸ , ³⁹ ] It is possible that a similar mechanism, against the trophozoite cell wall is in play. This would be in keeping with our electron microscopy observations of gross morphological changes to the dorsal cytoplasmic membrane of trophozoites, including membrane rupture. These effects would adversely affect the ability of the Giardia trophozoite to attach to membrane, interrupting a crucial stage in the parasite lifecycle, with the observed parasite killing.[ ³¹ , ³⁶ ]

Conflict of interest

S.W.P. is director of Neoculi Pty. Ltd., who are seeking to develop these analogues for use in at‐risk species. The authors declare no other conflicts of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

A. J. Stevens, R. Abraham, K. A. Young, C. C. Russell, S. N. McCluskey, J. R. Baker, B. Rusdi, S. W. Page, R. O'Handley, M. O'Dea, S. Abraham, A. McCluskey, ChemMedChem 2022, 17, e202200341.

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.