Tizoxanide Synthesis Essay

Abstract

Tuberculosis, caused by Mycobacterium tuberculosis infection, is a major cause of morbidity and mortality in the world today. M. tuberculosis hijacks the phagosome-lysosome trafficking pathway to escape clearance from infected macrophages. There is increasing evidence that manipulation of autophagy, a regulated catabolic trafficking pathway, can enhance killing of M. tuberculosis. Therefore, pharmacological agents that induce autophagy could be important in combating tuberculosis. We report that the antiprotozoal drug nitazoxanide and its active metabolite tizoxanide strongly stimulate autophagy and inhibit signaling by mTORC1, a major negative regulator of autophagy. Analysis of 16 nitazoxanide analogues reveals similar strict structural requirements for activity in autophagosome induction, EGFP-LC3 processing and mTORC1 inhibition. Nitazoxanide can inhibit M. tuberculosis proliferation in vitro. Here we show that it inhibits M. tuberculosis proliferation more potently in infected human THP-1 cells and peripheral monocytes. We identify the human quinone oxidoreductase NQO1 as a nitazoxanide target and propose, based on experiments with cells expressing NQO1 or not, that NQO1 inhibition is partly responsible for mTORC1 inhibition and enhanced autophagy. The dual action of nitazoxanide on both the bacterium and the host cell response to infection may lead to improved tuberculosis treatment.

Author Summary

Tuberculosis is responsible for approximately 2 million deaths worldwide each year. Current treatment regimens require administration of multiple drugs over several months and resistance to these drugs is on the rise. Mycobacterium tuberculosis, the causative agent of the disease, can proliferate within host cells. It has been recently observed that autophagy (cellular self-eating) can kill intracellular M. tuberculosis. We report that the antiprotozoal drug nitazoxanide and its metabolite tizoxanide induce autophagy, inhibit signaling by mTORC1, a major negative regulator of autophagy, and prevent M. tuberculosis proliferation in infected macrophages. We show that nitazoxanide exerts at least some of its pharmacological effects by targeting the quinone reductase NQO1. Our results uncover a novel mechanism of action for the drug nitazoxanide, and show that pharmacological modulation of autophagy can suppress intracellular M. tuberculosis proliferation.

Citation: Lam KKY, Zheng X, Forestieri R, Balgi AD, Nodwell M, Vollett S, et al. (2012) Nitazoxanide Stimulates Autophagy and Inhibits mTORC1 Signaling and Intracellular Proliferation of Mycobacterium tuberculosis. PLoS Pathog 8(5): e1002691. https://doi.org/10.1371/journal.ppat.1002691

Editor: Vojo Deretic, University of New Mexico, United States of America

Received: July 25, 2011; Accepted: March 27, 2012; Published: May 10, 2012

Copyright: © 2012 Lam et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Funding: This research was supported in part by funds from the Canadian Breast Cancer Foundation (www.cbcf.org) and the Canadian Institutes for Health Research (www.cihr-irsc.gc.ca), and CIHR operating grant #MOP-106622 (to Y.A-G). X.Z. is a recipient of the University of British Columbia Graduate Entry in Experimental Medicine Award. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Mycobacterium tuberculosis (Mtb) is the bacterial pathogen that causes tuberculosis, a major infectious disease responsible for approximately 2 million deaths worldwide each year [1]. There is a major need for more effective therapy against tuberculosis [2], [3]. Mtb is a highly persistent and successful pathogen in part because of its ability to manipulate intracellular membrane trafficking events in host macrophages [4], [5]. Upon entering the host cell, Mtb resides in single-membraned phagosomes and initiates mechanisms to avoid the innate immune response that can activate macrophages [6]–[9]. A series of fusion events with various endocytic organelles, culminating in fusion with lysosomes, normally converts the phagosome into a phagolysosome that can destroy its microbial contents [7]. Mtb prevents this conversion at an early stage by secreting a protein phosphatase, PtpA, that blocks the acquisition of the vacuolar-type H+-ATPase required for acidification of the lumen [10]–[13], limiting the acquisition of lysosomal hydrolases and depleting the phagosome of phosphatidylinositol 3-phosphate [7], [14], [15].

Autophagy is another intracellular membrane trafficking pathway that can play a role in controlling bacterial infection [16], [17]. In this process, cytoplasmic constituents are sequestered in double-membraned structures called autophagosomes that are subsequently targeted for fusion with lysosomes and are degraded [18]. Under basal conditions this degradative pathway is important for recycling intracellular material and organelles to maintain cellular homeostasis. Experimental induction of autophagy in macrophages by starvation, rapamycin, interferon-γ or its downstream effector LRG-47, toll-like receptor stimulation, ATP stimulation, or by small molecules reduced survival of intracellular Mtb [8], [19]–[23]. This was associated with increased acidification of phagosomes and increased colocalization of lysosomal and autophagosomal markers with Mtb-containing phagosomes [8], [19], [20], suggesting the block to phagosome maturation was overcome and fusion with lysosomal and autophagosomal compartments occurred. Further work has shown that the reduced Mtb survival is associated with delivery to the Mtb compartment of autophagosomal protein cargo that is proteolysed to generate cationic peptides that are toxic to Mtb [24], [25].

Autophagy is in part regulated by the mammalian target of rapamycin complex 1 (mTORC1), a nutrient-, energy- and growth factor-sensing master regulator of cell growth and metabolism [26]. mTORC1 is stimulated by growth factors and nutrients to promote anabolic processes such as translation and protein synthesis. Conversely, nutrient deprivation, cellular stress and the chemical rapamycin inhibit mTORC1, leading to the attenuation of anabolic reactions and the induction of autophagic catabolism as a protective function [27].

The evidence supporting a protective, cell-clearing function for autophagy in Mtb-infected macrophages suggests autophagy and mTORC1 signaling as attractive targets for new treatments for tuberculosis. Few studies have explored the use of approved drugs to manipulate autophagy or mTORC1 to combat Mtb infection. We recently reported results of a screen for chemicals that increase autophagosome formation and identified niclosamide, an approved salicylanilide antihelmintic drug, as a potent stimulator of autophagy and inhibitor of mTORC1 signaling [28]. Although niclosamide is very effective in the intestinal tract, it is not a good candidate for Mtb treatment because of its poor absorption. In the present paper we examine whether nitazoxanide (NTZ, 2-acetyloxy-N-(5-nitro-2-thiazolyl)benzamide, Alinia), a newer antiparasitic drug that was synthesized based on the structure of niclosamide [29] and that shows good gastrointestinal absorption, might also affect autophagy and mTORC1 activity and be a better drug candidate for Mtb treatment.

We report that NTZ stimulates autophagy, inhibits mTORC1 signaling, and inhibits Mtb intracellular proliferation at concentrations found in the blood in humans after administration of a standard oraldose. We show that NTZ inhibits the enzymatic activity of human quinone oxidoreductase NQO1,probably acting upstream of mTORC1 to stimulate autophagy. We also provide insights into the structural requirements of NTZ for activity. Our work further supports a mechanistic link between autophagy and Mtb proliferation, and presents an additional option for the manipulation of autophagy in the treatment of Mtb.

Results

Nitazoxanide modulates autophagy

NTZ (Figure 1A) structurally resembles niclosamide, a drug that stimulates autophagy and inhibits signaling by mTORC1, a negative regulator of autophagy [28]. NTZ is a prodrug that is rapidly hydrolyzed in plasma to the active metabolite tizoxanide (TIZ; Figure 1A) [30]. We therefore examined the effects of both NTZ and TIZ on autophagy. The formation of autophagosomes entails the recruitment of cytosolic Atg8/LC3 to nascent autophagosomes [31]. This process can be monitored quantitatively by automated fluorescence microscopy in MCF-7 cells expressing LC3 fused to EGFP (EGFP-LC3) as a shift from diffuse to punctate cytoplasmic fluorescence [28]. Cells were exposed for 3 h or 24 h to concentrations of NTZ or TIZ ranging from 0.1 to 100 µM. Untreated cells showed mostly diffuse cytosolic EGFP-LC3 staining and a low level of punctate EGFP-LC3, representing basal levels of autophagy (Figure 1B, C). After 3 hincubation, NTZ and TIZ caused a concentration-dependent increase in autophagosomes that was detectable at 3 µM and maximal at 30 µM (Figure 1B and Figure S1). After 24 h incubation, autophagosome accumulation was even more pronounced (Figure 1B and Figure S1).

Figure 1. Induction of autophagosomes by NTZ and TIZ.

(A)Structures of NTZ and TIZ. (B) Punctate EGFP-LC3 levels in MCF-7 cells stably expressing EGFP-LC3. Accumulation was measured quantitatively by automated microscopy in cells incubated for 3 h or 24 h with different concentrations of NTZ or TIZ. (C) Representative images of cells showing EGFP-LC3 (green) and DNA (blue).Scale bar, 10 µm.

https://doi.org/10.1371/journal.ppat.1002691.g001

Autophagosome accumulation observed in a static image may reflect increased autophagic flux but has also been observed in the presence of chemicals that decrease autophagic flux by inhibiting the final stage of autophagy in which autophagosomes fuse with lysosomes and are degraded [32]. To distinguish between these possibilities, the effect of NTZ and TIZ on EGFP-LC3 processing and degradation was examined by immunoblotting. Recruitment of LC3 or EGFP-LC3 to autophagosomes involves its proteolytic cleavage and lipidation, yielding a species with increased electrophoretic mobility termed LC3II/EGFP-LC3II [31]. Upon fusion with lysosomes, the contents of autophagosomes, including sequestered LC3, are degraded by lysosomal hydrolases. However, the free EGFP portion is degraded more slowly than LC3 itself, leading to transient accumulation of a band corresponding to the size of EGFP (∼25 kD) [32].

Untreated cells showed mostly full length EGFP-LC3 and only a small amount of free EGFP (Figure 2A). Incubation with NTZ or TIZ for 4 h (Figure 2A) and 24 h (Figure 2B) caused a concentration-dependent increase in the lipidated product EGFP-LC3II that was visible at 3 µM and strong at ≥10 µM. These results show the drugs increased the processing of EGFP-LC3 that accompanies autophagosome formation. NTZ and TIZ also caused an increase in the free EGFP band (Figure 2A, B) that was abolished in the presence of the vacuolar-type H+ ATPase inhibitor bafilomycin A1 (Figure 2C), a known inhibitor of lysosomal fusion [33], [34]. This shows that NTZ and TIZ can increase autophagic flux. Concentrations of NTZ and TIZ ≥30 µM resulted in decreased free EGFP, despite the presence of an EGFP-LC3II band (Figure 2A,B), raising the possibility that autophagic flux was reduced at high concentrations of NTZ and TIZ.As discussed further below, a decrease could be a consequence of activation of PKB/Akt (Figure 2A, B), which is known to downregulate autophagy [35], [36].

Figure 2. Increased EGFP-LC3 processing and inhibition of mTORC1 signaling by NTZ and TIZ.

Cells were treated with the indicated concentrations of NTZ, TIZ or rapamycin for 4 h (A) or 24 h (B). In panel C, cells were treated with 10 µM NTZ, 10 µM TIZ, 30 nM rapamycin, or DMSO without or with 0.1 µM bafilomycin A1 for 4 h. EGFP-LC3 processing was examined by immunoblotting with antibodies against GFP, mTORC1 activity using antisera against phospho-S6K Thr389 and total S6K, and mTORC2 activity with antisera against phospho-AKT Ser473 and total AKT. Total AKT was also used as a loading control.

https://doi.org/10.1371/journal.ppat.1002691.g002

Nitazoxanide inhibits mTORC1 but not mTORC2 signaling

mTORC1 is a protein kinase complex whose activity may be monitored by measuring phosphorylation of key substrates such as ribosomal S6 kinases (S6Ks) by immunoblotting [37]. Cells maintained in complete cell culture medium containing serum and nutrients showed high levels of Thr389 phosphorylation by mTORC1 and increased electrophoretic mobility of p70S6Kandp85S6K(Figure 2A) and this was completely repressed within 4 h exposure to the mTORC1 inhibitor rapamycin (Figure 2A). Exposure to different concentrations of NTZ or TIZ for 4 h or 24 h resulted in a concentration-dependent decrease inS6K phosphorylation - inhibition was partial at 1 µM, strong at 3 µM and essentially complete at 10 µM(Figure 2A,B). Thus, NTZ and TIZ inhibit mTORC1 signaling.

mTOR is also the catalytic subunit of a second complex termed mTORC2 that phosphorylates PKB/Akt at Ser473[38]. The rapamycin insensitive mTORC2 modulates changes in the cytoskeleton and is largely unaffected by nutrients or energy conditions [39]. To determine whether NTZ and TIZ inhibit mTORC2, the same cell lysates were probed with a PKB/Akt Ser473 phosphospecific antibody. Untreated cells displayed mTORC2 activity that was not inhibited by rapamycin(Figure 2A, B). In fact, rapamycin caused an increase in mTORC2 activity (Figure 2A, B). This is consistent with previous studies showing that mTORC1/S6K1 signaling downregulates PKB/Akt phosphorylation on Ser473 via a feedback loop involving transcriptional inhibition of the insulin receptor substrate IRS-1 gene and degradation of IRS-1 and IRS-2 proteins, and that mTORC1 inhibition prevents the establishment of this feedback loop, thus increasing PKB/Akt Ser473 phosphorylation [40], [41]. NTZ and TIZ also did not decrease PKB/Akt Ser473 phosphorylation but rather increased it in a concentration-dependent manner that paralleled their level of mTORC1 inhibition (Figure 2A, B).The increase in PKB/AKT Ser473 phosphorylation at higher concentrations of NTZ and TIZ was concomitant with the decrease in free EGFP (Figure 2A, B). Since PKB/Akt downregulates autophagy [42], it is possible that at high concentrations of NTZ and TIZ and long incubation times, inhibition of mTORC1 stimulates autophagy while feedback activation of PKB/Akt elicits signals that counteract autophagy.

Inhibition of mTORC1 signaling was detectable after 30 min exposure to 10 µM NTZ or TIZ and maximal after ≥1 h (Figure 3A) and PKB/Akt Ser473 phosphorylation rose detectably at 30 min and increased further at longer exposure times (Figure 3A). To examine whether mTORC1 signaling inhibition was reversible, cells were incubated with 10 µM NTZ or TIZ for 4 h, the drugs were washed away and the cells were incubated in drug-free medium for different times. As a comparison, cells were treated with rapamycin, which inhibits mTORC1 essentially irreversibly. mTORC1 was strongly inhibited after 4 h exposure to either NTZ or TIZ (Figure 3B, 0 h) but its activity increased over time after drug removal, returning to levels seen in untreated cells within 4 h, while mTORC1 activity remained profoundly inhibited upon rapamycin withdrawal (Figure 3B).

Figure 3. Time-course and reversibility of mTORC1 inhibition and EGFP-LC3 processing by NTZ and TIZ.

(A) Cells were incubated with 10 µM NTZ or 10 µM TIZ for the indicated times. (B) Cells were incubated with 10 µM NTZ, 10 µM TIZ or 30 nM rapamycin for 4 h. The drugs were then washed away and cells were incubated in drug-free medium for the indicated times post-washout. Cell lystates were immunoblotted as in Figure 2.

https://doi.org/10.1371/journal.ppat.1002691.g003

Together, these results show that NTZ and TIZ strongly but reversibly inhibit mTORC1 signaling. Since the mTOR catalytic subunit is shared by mTORC1 and mTORC2, the observation that the drugs inhibit mTORC1 while activating mTORC2 is strong evidence that they do not directly inhibit the kinase activity of mTOR itself but likely act on an upstream mTORC1 regulatory pathway.

Structural requirements for stimulation of autophagy and inhibition of mTORC1 by nitazoxanide

NTZ targets in mammals are not well-characterized. To address its mechanism of action in human cells, we first carried out a limited structure-activity relationship study. Sixteen analogues (Figure 4 and Figure S2) were synthesized and tested for effects on autophagosome accumulation, EGFP-LC3 processing and inhibition of mTORC1 activity (Table 1).

NTZ and TIZ(Figure 4) showed similarly strong activity in all three assays (Table 1). Analogue 1 lacking the OH group was completely inactive in all three assays, implying that generation of the OH group by hydrolysis of NTZ is essential for activity. In support of this point, analogue 2 with an ether bond that is not cleavable by esterases was inactive in all three assays (Table 1). In addition, analogue 3 bearing a methylacetate substituent like NTZ but arranged such that cleavage by esterases would generate a carboxylic acid instead of an OH, was also inactive in all assays. Moreover, analogues 4, 5 and 6, bearing more bulky propionate, isobutyrate and pivalate substituents that are cleaved more slowly by esterases, were also active in all three assays. Together, these data show that NTZ is the active prodrug of TIZ and that the OH group is critical for activity.

Methylation of the NH group in the linker region (compound 7, Figure 4) caused complete loss of activity, suggesting that inter- or intramolecular hydrogen bonding by the secondary amine or the generation of an anionic form of NTZ [43] is required for activity. Nitazoxanide also possesses a conspicuous nitro group attached to the 5-membered ring (Figure 4). Analogue 8 lacking this group was completely inactive in all three assays, showing the nitro group is also required for activity. Analogue 9 with an aldehyde substituent was also inactive, indicating that the strongly electron-withdrawing character of the nitro group is more relevant to activity than its steric bulk.

Addition of one or two chlorine substituents to the 6-membered ring of NTZ also eliminated activity. Since niclosamide possesses a Cl substituent para to the OH group, this result indicates that NTZ and niclosamide probably have different cellular targets. Several combinations of substitutions described individually above were also inactive in all assays(Table 1, Figure S1).

Together, these results show that the OH, NH and NO2 groups are all essential for induction of autophagosome accumulation, EGFP-LC3 processing and mTORC1 inhibition, revealing strict structural requirements for activity. The tight correlation between the structural features required for activity in all three assays also strongly implies that these biological responses are linked and probably result from inhibition of a single target rather than multiple independent targets.

Nitazoxanide and tizoxanide inhibit intracellular Mtb proliferation

We next used a luciferase reporter assay to assess the ability of NTZ to inhibit Mtb growth in infected macrophages. Differentiated THP-1 human acute monocytic leukemia cells were infected with a Mtb strain expressing luciferase. After removal of non-internalized bacteria, NTZ was added and Mtb proliferation was determined 24, 48, and 72 h later. NTZ showed concentration-dependent inhibition at all time points, with essentially complete inhibition at 10 µM (Figure 5A). Exposure of Mtb in liquid culture to 10 µM NTZ was able to reduce Mtb growth, albeit less potently compared to the growth in THP-1 cells (Figure S3).

Figure 5. Effect of NTZ on Mtb proliferation and survival of THP-1 cells.

(A) Differentiated THP-1 cells infected with Mtb H37Rv bearing a luciferase-reporting plasmid were treated with various concentrations of NTZ or TIZ for the indicated times. Intracellular Mtb was measured as luciferase activity at 24, 48 and 72 h. (B) Infected THP-1 cells were treated as in (A) but after 24 h treatment the drugs were removed and cells were incubated with medium without drug. Luciferase activity was measured at 24 h, 48 h (24 h post-wash) and 72 h (48 h post-wash). (C) Differentiated THP-1 cells were exposed to various concentrations of NTZ, TIZ or rapamycin for 4 h. Endogenous LC3 processing (using LC3 antibodies) and mTORC1 activity were determined by immunoblotting as in Figure 2.Viability of THP-1 cells treated with drugs for the indicated times was measured with the MTT assay (D) or by propidium iodide (PI) staining (E, F). PI levels were measured quantitatively by automated microscopy, and viable cells was calculated as the percentage of PI-negative cells. (F) After drug removal at 24 h, THP-1 cell survival was measured by PI at 48 h (24 h post-wash) and 72 h (48 h post-wash).

https://doi.org/10.1371/journal.ppat.1002691.g005

To examine the reversibility of the effects of NTZ, infected THP-1 cells were treated with NTZ for 24 h after which time the drug was removed and cells were cultured for up to two additional days. Exposure to 3 µM or 10 µM NTZ significantly reduced Mtb proliferation at 24 h but intracellular Mtb proliferation resumed upon drug removal (Figure 5B). Immunoblotting for endogenous LC3 and for phospho-Thr389S6K indicated that autophagy was induced by NTZ and TIZ in THP-1 cells, and that mTORC1 was inhibited (Figure 5C). Rapamycin, which does not inhibit all functions of mTORC1 and is a weak stimulator of autophagy [44], [45], had no effect on Mtb proliferation in THP-1 cells (Figure 5A, B, and Figure S4) or in liquid culture in the absence of cells (Figure S3),had only a modest effect on the conversion of LC3 to LC3II, while completely inhibiting Thr389S6K phosphorylation (Figure 5C), and did not affect THP-1 cell survival (Figure 5D–F).

We also assessed the effect of NTZ on the viability of differentiated THP-1 cells using three different assays: the MTT assay to measure cell metabolic activity (Figure 5D), propidium iodide (PI) exclusion to measure plasma membrane integrity (Figure 5E),and staining of live cells with Hoechst 33342 to count attached cells. The MTT assay showed a modest time- and concentration-dependent reduction in THP-1 cell metabolic activity (Figure 5D), as did the Hoechst 33342 assay (unpublished data). By contrast, after 72 h of treatment with 10 µM NTZ, large numbers of cells had taken up PI (Figure 5E).Cells treated with NTZ for 24 h and transferred to drug-free medium retained viability (Figure 5F).

The observation that after 72 h 10 µM NTZ had only minor effects in the MTT assay but strongly increased PI uptake in differentiated THP-1 cells was perplexing. Cellular toxicity was unexpected given the knowledge that peak plasma concentrations of approximately 37 µM are achieved after a standard oral dose of 500 mg NTZ [30] and that NTZ has an excellent safety profile, with no significant side effects in patients taking 500 mg NTZ twice daily for 24 weeks [46]. We therefore considered it important to examine the effects of NTZ on human primary peripheral blood mononuclear cells (PBMC), as cells of more direct relevance to Mtb therapy.

PBMC were isolated from blood samples from a number of healthy human donors. Cells were infected with Mtb at a multiplicity of infection (MOI) of 1 or 10 for 24 h, after which non-internalized bacteria were washed away and infected cells were exposed to 3–30 µM NTZ for up to 72 h. The effect of 10 µM NTZ on intracellular Mtb proliferation ranged from 0 to 55% inhibition at 72 h, depending on the donor and the multiplicity of infection (Figure 6A–C). 30 µM NTZ almost completely inhibited Mtb proliferation, starting as early as 24 h (Figure 6B, C).Notably, exposure of PBMC to NTZ at 10 µM or 30 µM for up to 72 h caused no cytotoxicity, with cell survival remaining at 100%, in both the PI uptake and MTT assays (Figure 6D–F). By contrast, rapamycin reduced cell survival to ∼30% (Figure 6D–E).

Figure 6. Effects of NTZ on Mtb proliferation and viability of PBMC.

Peripheral blood mononuclear cells isolated from healthy human subjects and infected with Mtb H37Rv bearing a luciferase-reporting plasmid were treated with various concentrations of NTZ for the indicated times. Intracellular Mtb was measured as luciferase activity. Data presented are: subject 1withMOI 10 (A), subject 2with MOI 1 (B) and subject 2 with MOI 10 (C). Viability normalized to DMSO-treated controls was measured at 48 and 72 h, with PBMC from subject 2 using the MTT assay (D) or by PI staining (E).(F) Viability, normalized to DMSO treated controls, was measured with PBMC from subject 3 at 72 h.

https://doi.org/10.1371/journal.ppat.1002691.g006

These observations show that NTZ can inhibit the intracellular proliferation of Mtb at concentrations that are not toxic to primary human cells. They are consistent with previous observations that rapamycin has toxic effects on primary human cells [47]and that NTZ is safe in humans at a plasma concentration of 30 µM [30], [48].

Nitazoxanide inhibits the human NAD(P)H quinone oxidoreductase NQO1

Nitazoxanide has no known targets in humans. Our results thus far suggest that it may target an upstream mTORC1 regulatory pathway rather than acting directly on mTORC1 or autophagy. It has recently been observed using affinity chromatography that RM4847, a bromo derivative of TIZ, can bind human quinone oxidoreductase NQO1 [49]. We used a modified version of the Prochaska microtiter assay to monitor the effects of NTZ on NQO1 activity [50], [51]. NTZ was able to inhibit purified NQO1 in vitro in a concentration-dependent manner, as did dicoumarol (DIC), a known competitive inhibitor of NQO1 enzymatic activity [52] (Figure 7A). Rapamycin, at a concentration that completely inhibits mTORC1(0.1 µM),did not cause significant NQO1 inhibition, and NQO1 remained 60% active at a very high concentration of rapamycin (1 µM) (Figure 7A). Next, we performed the Prochaska assay on freshly prepared MCF-7 cell lysates. Cellular NQO1 activity was reduced to approximately 14% by 10 µM NTZ (Figure 7B), a concentration that inhibits mTORC1 in cells.

Figure 7. Inhibition of NQO1 by NTZ and TSC2-dependent mTORC1 inhibition.

(A) Relative activity of NQO1 as determined by a modified Prochaska microtiter plate bioassay, in which 0.25 µg/ml pure NQO1 enzyme was treated with different concentrations of NTZ, dicoumarol (DIC), and rapamycin (RAPA) for 1 h. (B) MCF-7 cell lysates were treated with various concentrations of NTZ for 1 h, and relative activity of NQO1 was determined by the Prochaska bioassay. (C) MCF-7 cells or (D) TSC2+/+ and TSC2−/−MEFs were treated with NTZ, DIC, and rapamycin for 4 h and subjected to immunoblotting for total and phosophorylated S6K as previously described.(E) Total NQO1 or (F) total and phosphorylated S6K was assessed by immunoblotting in MCF-7 or HEK 293T cells treated with indicated concentrations of drugs for 8 h. Lines in (F) were used to indicate juxtaposition of noncontinguous lanes from the same gel and image exposure.

https://doi.org/10.1371/journal.ppat.1002691.g007

We also observed that the NQO1 inhibitor DIC can strongly inhibit mTORC1, as assessed by S6K phosphorylation at 4 h (Figure 7C), raising the possibility that NQO1 may be an upstream regulator of mTORC1. The TSC1–TSC2 tuberous sclerosis complex is a major negative upstream regulator of mTORC1 and mouse embryo fibroblasts (MEFs) lacking the TSC2 gene (TSC2−/−) show high mTORC1 activity compared to TSC2+/+MEFs [53]. NTZ and DIC inhibited mTORC1 in TSC2+/+ cells but not in TSC2−/− cells (Figure 7D), indicating that inhibition is dependent on the TSC1–TSC2 complex. We also examined the effects of NTZ on HEK 293T cells, which do not express NQO1 at a detectable level (Figure 7E). Concentrations of NTZ or DIC that significantly inhibited mTORC1 in MCF-7 cells, which express a much higher level of NQO1 protein (Figure 7E),did not exert as great an effect on mTORC1 signaling in HEK 293T cells (Figure 7F). These data are consistent with inhibition of NQO1 leading to downstream mTORC1 inactivation by a pathway involving the TSC1–TSC2 complex. By contrast, rapamycin inhibited mTORC1 equally in both cell lines (Figure 7D–F). The different cellular targets of rapamycin and NTZ may explain why rapamycin was unable to inhibit intracellular Mtb proliferation in infected cells.

Tuberculosis drugs do not interfere with the ability of NTZ and TIZ to induce autophagosome formation

Tuberculosis is always treated with drug combinations. We therefore asked whether NTZ and TIZ retain their ability to induce autophagosomes in the presence of tuberculosis drugs. MCF-7 cells were exposed to ethambutol (EMB), isoniazid (INH), pyrazinamide (PZA), streptomycin (STM) or rifampicin (RMP)aloneor with NTZ or TIZ. None of the drugs, when tested between 0.5 µg/ml and 50 µg/ml(equivalent to 2.45 µM and 244.73 µM EMB; 1.37 µM and 137.14 µM INH; 4.06 µM and 406.13 µM PZA; 0.86 µM and 85.97 µM STM; and 0.61 µM and 60.76 µM RMP), prevented the induction of punctate EGFP-LC3 by NTZ or TIZ at 3 or 10 µM (Figure 8). These results predict that NTZ would retain its ability to affect autophagy when used in combination treatment with tuberculosis drugs.

Figure 8. Effect of NTZ and TIZ on autophagosome formation in the presence of tuberculosis drugs.

MCF-7 cells expressing EGFP-LC3 were exposed to 3 µM or 10 µM NTZ (A) or TIZ (B), without or with 5 µg/ml ethambutol (EMB), pyrazinamide (PZA), isoniazid (INH), streptomycin (STM) or rifampicin(RMP). Representative images of treated cells are shown in (C). Scale bar, 10 µm.

https://doi.org/10.1371/journal.ppat.1002691.g008

Discussion

This study reveals new cellular roles for the antiparasitic drug nitazoxanide and its active metabolite tizoxanide as stimulators of autophagy and inhibitors of mTORC1 signaling, acting at least in part through suppression of the quinone oxidoreductase NQO1. Our work also demonstrates that these drugs inhibit Mtb proliferation in human cells. NTZ is a noncompetitive inhibitor of pyruvate: ferredoxin oxidoreductase (PFOR) [43], an enzyme found in amitochondriate human parasites and most anaerobic bacteria but that is not conserved in mammals. PFOR catalyses the oxidative decarboxylation of pyruvate to acetyl coenzyme A and CO2[43], [54] whereas mammals oxidize pyruvate using pyruvate dehydrogenase, which is not inhibited by NTZ [55]. To our knowledge, NTZ has not previously been shown to have direct targets in human cells or to act on host cell signaling pathways.

The thiazolide RM4847 differs from TIZ in having a Br substituent instead of a nitro group as well as methyl groups attached to the 6-membered ring and to the 5-membered ring. RM4847 coupled to agarose beads has been observed to bind human glutathione-S-transferase, GSTP1 [56] and human NQO1 [49]. We did not observe inhibition of cellular GSTP1 by NTZ (unpublished data) using the 1-chloro-2,4-dinitrobenzene reduction GSTP1 enzymatic activity assay [57], [58]. However, NTZ inhibited cellular NQO1 enzymatic activity in a concentration-dependent manner, strongly reducing NQO1 activity at 10 µM. Moreover, DIC, an NQO1 inhibitor, also inhibited mTORC1 signaling. The reduction of NQO1 enzymatic activity likely contributes to attenuation of mTORC1 signaling, since deletion of the major mTORC1 regulator TSC1–TSC2 rendered cells unresponsive to the effects of NTZ or DIC.NTZ and DIC were also less potent inhibitors of mTORC1 in cells that do not express NQO1, indicating that NQO1 is a relevant cellular target, but probably not the only one. Thus, NTZ may stimulate autophagy and inhibit mTORC1 signaling and intracellular Mtb proliferation at least in part as a result of inhibiting NQO1.

NQO1 can catalyze the reduction of a broad range of reactive substrates including quinones, quinone-imines and nitro-compounds to less reactive and less toxic forms [59]. It has also been shown to be a superoxide scavenger [60] and a “gatekeeper” of the 20S proteasome [61], [62]. We speculate that NQO1 may be linked to autophagy throughp62/sequestosome-1, an adaptor protein that binds LC3 and ubiquitin to facilitate autophagy of polyubiquitinated proteins [63]. This speculation is based on the observations that p62 canactivate the redox-sensing transcription factorNrf2resulting in persistent expression of NQO1 [64] and that knockdown of Nrf2caused a 2-fold increase in autophagy [65]. Our data linking NQO1 activity and mTORC1 signaling are in line with studies showing that NQO1 expression is induced to counter the toxicity of elevated levels of reactive oxygen species (ROS) [66], and that increased oxidative stress from elevated ROS inhibits mTORC1 signaling and induces autophagy through activation of TSC2 [67].Pharmacological inhibition of NQO1 may increase oxidative stress to inhibit mTORC1 and activate autophagy, thereby inhibiting intracellular Mtb proliferation.

NTZ can directly kill both replicating and nonreplicating Mtb in vitro[68]. NTZ kills Mtb in vitro in liquid and solid media at a minimum inhibitory concentration of ∼50 µM after 7 days [68]. This occurs by an uncharacterized mechanism because PFOR is not known to be present in Mtb [68]. This observation raises the question of whether NTZ prevents the replication of Mtb in human cells solely by targeting the bacteria themselves or also by affecting a host cellular process such as autophagy. In our hands, 10 µM NTZ completely inhibited mTORC1 signaling and strongly stimulated autophagy while this concentration directly inhibited Mtb proliferation in liquid culture less robustly (Figure S3). Therefore it can be argued that concentrations of NTZ that directly kill Mtb would also necessarily affect autophagy, a host cellular process implicated in Mtb intracellular proliferation. Additionally, 10 µM NTZ inhibited the intracellular replication of Mtb in THP-1 cells more strongly than it inhibited Mtb directly. The interpretation of this latter result is complicated by the observation that the PI exclusion assay, but not the MTT or Hoechst assays, showed 10 µM NTZ to be significantly toxic to differentiated THP-1 cells at longer exposure times. More importantly, all assays showed NTZ was not toxic to PBMC at the highest concentration and exposure time tested (30 µM, 72 h), consistent with its very safe toxicity profile in humans. Indeed, a peak plasma concentration of 37 µM TIZ was observed after a standard single oral dose of 500 mg NTZ while1 g administered twice daily for one week resulted in peak plasma concentrations reaching 100 µM, and neither was associated with adverse side effects [30]. Therefore, the toxicity of NTZ towards the differentiated THP-1 cell line in culture does not appear to reflect the in vivo situation. NTZ also inhibited the intracellular proliferation of Mtb in PBMC. Significant inhibition required a concentration of 10–30 µM (Figure 6), the range reflecting different sensitivities of PBMC from different donors.

In our experiments, NTZ and rapamycin both efficiently inhibited mTORC1 but NTZ reduced intracellular Mtb proliferation while rapamycin did not. This result is in apparent contrast with a previous report that rapamycin reduced Mtb replication in RAW 264.7 cells [19]. Different experimental conditions may explain this discrepancy, including the use of a much higher rapamycin concentration of 50 µM. Here, we used 30 nM, which was sufficient to completely inhibit the phosphorylation of S6K by mTORC1 in THP-1 cells. A more recent study reported ∼50% reduction in Mtb colony-forming units after 4 h treatment of infected THP-1 cells with 27 µM rapamycin [69].

Anti-tuberculosis drugs are limited in number and efficacy and prone to many side effects and are often inactive against nonreplicating (dormant) Mtb. Furthermore, multi drug-resistant strains are resistant to the first-line drugs rifampicin and isoniazid and extensively drug-resistant strains are also resistant to any fluoroquinolone and any of the second-line injectable drugs, and give very high mortality rates [2], [70]. Even non-resistant strains of Mtb can require up to nine months of therapy with a combination of at least three antibiotics to reduce development of resistance [71], [72]. Therefore new drugs with improved efficacy, shorter treatment time and lower cost are urgently needed.

NTZ is a safe, effective, orally bioavailable and inexpensive drug already approved for human use. It has shown broad-spectrum activity against anaerobic intestinal helminths and protozoans as well as some bacteria, including Clostridium difficile[73]–[79] and is approved for the treatment of enteritis caused by Cryptosporidium parvum and Giardia intestinalis[29]. The length of NTZ treatment is typically 3–14 days, but it has been used for up to 4 years in AIDS-related cryptosporidiosis patients without significant adverse effects [77], [80] and resistance to NTZ has not been reported. NTZ inhibited mTORC1, modulated autophagy, inhibited NQO1 and inhibited Mtb proliferation in PBMC at concentrations well within tolerated plasma concentrations. The autophagy-stimulating effect of NTZ was not impaired in the presence of the anti-tuberculosis drugs ethambutol, isoniazid, pyrazinamide, streptomycin, or rifampicin. Drugs such as NTZ that both directly inhibit the proliferation of pathogenic microorganism and stimulate host cellular defense mechanisms such as autophagy may provide new opportunities to combat intracellular pathogens like Mtb.

Materials and Methods

Ethics statement

Peripheral blood was collected from normal human subjects, in accordance with the ethics approval guidelines of the University of British Columbia Research Ethics Board and Tri-Council Policy Statement for Ethical Conduct for Research Involving Humans, for secondary use of anonymous human blood. Written informed consent was obtained from all study participants.

Reagents

Cell culture reagents were purchased from Invitrogen, unless stated otherwise. General laboratory chemicals were purchased from Sigma-Aldrich, Fisher Scientific and BDH Inc. Nitazoxanide (N490100) and tizoxanide (T450100) were purchased from Toronto Research Chemicals Inc, rapamycin (553210) from Calbiochem, and Hoechst 33342 (H3570) from Invitrogen. Dicoumarol (Sigma M1390) was prepared fresh in DMSO before each experiment. Anti-phospho Thr389 S6K (#9205), anti- phospho Ser473 Akt (#9271), anti-Akt (#9272), and anti-NQO1 (#3187) antibodies were from Cell Signaling Technology. Anti-S6K C-18 (#230) antibody was purchased from Santa Cruz Biotechnology, and anti-GFP antibody (#1814460) was from Roche. Antibody to endogenous LC3 was purchased from Nanotools (0260-100). The M. tuberculosis H37Rv strain carrying pMV361-c1-lucplasmid was used for macrophage infection studies. pMV361-c1-luc carries a 1.65 kb fragment of the firefly luciferase gene under constitutive expression of a mycobacterial HSP60 promoter [81].

Cell culture procedures

MCF-7 cells stably transfected with pEGFP-LC3 have been described previously [28]. These cells were maintained in RPMI-1640 medium (Invitrogen 72400) supplemented with 400 µg/mL G418 (Promega #V7982) and 10% (v/v) fetal bovine serum (FBS) with 2 g/L sodium bicarbonate and 1 mM HEPES. THP-1 cells were maintained in RPMI-1640 supplemented with 10% FBS and 1% glutamine.TSC2−/−/p53−/− and TSC2+/+/p53−/− mouse embryo fibroblasts (MEF) were a generous gift of Dr. David Kwiatkowski [82]. MEF and HEK 293T cells were maintained in high glucose Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% FBS.

Automated assay for autophagosome induction

MCF-7 cells stably expressing EGFP-LC3 were seeded at 20,000 cells/well in PerkinElmer View 96-well plates. Eighteen hours after seeding, chemicals were added to each well and plates were incubated at 37°C for the indicated times. The medium was removed and cells were fixed with 3% (v/v) paraformaldehyde containing 500 ng/mL Hoechst 33342 for 15 min at room temperature. Fixed cells were washed once with PBS containing 1 mM MgCl2 and 0.1 mM CaCl2 and stored in the same medium at 4°C overnight. Plates were scanned using a Cellomics™Arrayscan VTI automated fluorescence imager. Cells were photographed and quantified using the compartment analysis algorithm as previously described [28].

Cell lysis and protein quantification

Cells were harvested in 20 mM Tris-HCl pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% (v/v) Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM β-glycerophosphate supplemented with fresh 1 mM Na3VO4, 1 mM dithiothreitol, and 1× complete protease inhibitor cocktail (Roche, #1169748001). Supernatants were collected after centrifugation at 18,000×g for 15 min at 4°C and quantified using the BCA protein assay kit (Pierce #23227).

SDS-PAGE and immunoblotting

Electrophoresis and immunoblotting conditions for EGFP-LC3 processing, S6K and AKT in MCF-7, THP-1 and HEK 293T cells were exactly as previously described [28]. MEF were seeded in 6-well plates at 400,000 cells/well (TSC2+/+) or 200,000 cells/well (TSC2+/+) in normal DMEM medium, cultured overnight and treated similarly.NQO1 levels were assessed by resolving protein samples on a 10% acrylamide gel and subjected to electroblotting onto nitrocellulose, then blocked with 5% (w/v) fat-free milk and incubated with NQO1 antisera. For immunoblotting of endogenous LC3, proteins were transferred onto polyvinylidene fluoride membrane with membrane with 0.2 µm pores, which were then fixed with 0.2% glutaraldehyde in PBS with 0.02% Tween-20 for 20 min at room temperature, prior to blocking with 5% fat-free milk.

Synthesis of nitazoxanide analogues

Analogues 1, 10, 11 were prepared by coupling between benzoyl chloride and heteroaromatic primary amines. Analogues 2, 12, 13 were prepared by coupling between 2-ethoxybenzoyl chloride and heteroaromatic primary amines. Analogues 4, 5, 6, 8, 9, 14, 15, 16 were prepared by coupling between salicyloyl chloride esters and heteroaromatic primary amines. Analog 3 was prepared by coupling between 2-(methoxycarbonyl)benzoyl chloride and 2-amino-5-nitrothiazole. Analogue 7 was prepared by methylation of nitazoxanide using the methylating agent iodomethane and K2CO3. Full details are given in the Supporting Information.

THP-1 differentiation, Mtb co-infection and luciferase assay

THP-1 cells were differentiated for 24 h with 40 µM PMA in RPMI-1640 medium supplemented with 10% FBS and 1% glutamine. For macrophage co-infection, cell suspensions were differentiated in 96-well plates and incubated at 37°C overnight. The cells were washed 3 times with 100 µl RPMI medium before infection. Mtb cultures, grown in Middlebrook 7H9 supplemented with oleic acid-albumin-dextrose-catalase (OADC) and 0.05% Tween 80, were washed with 7H9 medium and the bacteria were opsonized by resuspension in RPMI-1640 medium containing 10% human serum for 30 min at 37°C. Donor-specific serum was used for the infection of PBMC. The opsonized bacteria were resuspended in RPMI-1640 and added to differentiated THP-1 cells or PBMC at a multiplicity of infection (MOI) of 1 or 10 for 3.5 h at 37°C in 5% CO2. The cells were washed gently 3 times with RPMI-1640 to remove non-internalized bacteria and then incubated in 100 µl of RPMI-1640 containing 1% glutamate and 10% FBS. After 24 h, the medium was aspirated and replaced with medium containing test compounds. After 24, 48 or 72 h, the medium was aspirated and 100 µl Bright-Glo reagent (Promega TM052) was added. After 10 min incubation, luciferase activity was measured with aTropix TR7171 luminometer (Applied Biosystems).

Isolation of human PBMC

Peripheral blood was collected from normal human subjects, in accordance with the ethics approval guidelines as stated above. Blood was collected in BD Vacutainer Plus plastic plasma tubes coated with sodium heparin (158 USP units) (BD Diagnostics #367874). Blood from a single donor was diluted 1∶1 with PBS, layered over Ficoll-Paque PLUS (Stemcell Technologies #07957), and centrifuged at 900 g for 30 min according to manufacturer's instructions. White blood cells were isolated from buffy coats after centrifugation, and their viability was assessed by Trypan blue to be >98%. 5×106white blood cells were plated directly in 96-well assay plates in RPMI-1640 supplemented with 5% donor-specific human serum. Human serum was prepared from the plasma portion of the Ficoll gradient, heat-inactivated at 55°C for 30 min, centrifuged at 3,000×g for 20 min and filter sterilized before use. PBMC were allowed to adhere to plates for 20–24 h, after which cells were washed 2–3 times with supplemented RPMI to remove non-monocytic cells. Adherent cells were subsequently infected or treated with drugs as described above.

Viability assays

The viability of PBMC or differentiated THP-1 cells infected with M. tuberculosis in 96-well plates was measured at 24, 48 and 72 h using the (3-[4,5-Dimethylthiazol-2-yl]-2,5-diphenyl-tetrazolium bromide (MTT) assay (M2128, Sigma) as described [83]. For propidium iodide (PI) and nuclear staining assays, 100 µl of pre-warmed RPMI-1640 containing propidium iodide (Sigma-Aldrich P-4170) and Hoechst 33342was added directly to wells at indicated timepoints, to final concentrations of0.25 µg/ml and 0.5 µg/ml, respectively. Plates were incubated at 37°for 30 min, and scanned using the CellomicsArrayscan VTI automated fluorescence imager. The cells were imaged with a 20× objective with the Hoechst and TRITC (Ch2, red) channel, and data was collected from at least 800 cells per well. The target activation algorithm was applied to obtain a nuclear mask in the Hoechst channel and corresponding cytoplasmic mask applied in the TRITC channel to quantitate cell number and PI positive cells, respectively. The untreated wells were identified as reference wells during the scan. PI-positive cells show up as bright fluorescent objects in Ch2, with a higher intensity than live cells. The data output “% responder_AvgIntenCh2” is the percentage of cells with average intensity in cytoplasmic mask Ch2which had higher average intensity than that of untreated cells. Cell viability was calculated as the proportion of cells that were propidium iodide-negative.

NQO1 enzymatic activity assay

NQO1 activity was measured using a modified version of the Prochaska microtiter plate bioassay, which measures enzymatic activity based on the menadione-mediated reduction of MTT [50], [51].0.25 µg/ml of purified NQO1 enzyme (human DT Diaphorase, Sigma D1315) in dH2O, or fresh lysates from MCF-7 cells were used in each reaction in a 96-well plate format. MCF-7 cells were grown overnight in RPMI in 6-well plates, then washed and resuspended in cold PBS. 9×106

Abstract

The activities of the N-(nitrothiazolyl) salicylamide nitazoxanide and its metabolite tizoxanide were compared with metronidazole in vitro in microplates against six axenic isolates of Giardia intestinalis. Tizoxanide was eight times more active than metronidazole against metronidazole-susceptible isolates and twice as active against a resistant isolate. In 10 axenic isolates of Entamoeba histolytica, while tizoxanide was almost twice as active as metronidazole against more susceptible isolates, it was more than twice as active against less susceptible isolates. Fourteen metronidazole-susceptible isolates of Trichomonas vaginalis were 1.5 times more susceptible to tizoxanide, which was nearly five times as active against resistant isolates. Two highly metronidazole-resistant isolates retained complete susceptibility to tizoxanide, and one moderately resistant isolate showed reduced susceptibility. In all three organisms, nitazoxanide results paralleled those of tizoxanide. Analogues lacking the reducible nitro-group had similar low activities against susceptible G. intestinalis, E. histolytica and T. vaginalis, indicating that nitro-reduction and free radical production was a probable mode of action. Nitazoxanide and its metabolite tizoxanide are more active in vitro than metronidazole against G. intestinalis, E. histolytica and T. vaginalis. Although, like metronidazole, they depend on the presence of a nitro-group for activity, they retain some activity against metronidazole-resistant strains, particularly of T. vaginalis. The results suggest that resistance mechanisms for metronidazole can be bypassed by nitazoxanide and tizoxanide.

Introduction

Metronidazole, a 5-nitroimidazole drug, is the treatment of choice for intestinal disease caused by the flagellate Giardia intestinalis and the amoeba Entamoeba histolytica and for vaginitis caused by the flagellate Trichomonas vaginalis.1 Although failure following metronidazole therapy has been reported in some cases of liver abscess due to E. histolytica,2 resistance has not yet been proven in strains isolated from patients. In contrast, proven cases of metronidazole resistance in G. intestinalis and T. vaginalis have been documented.3,4 Should resistance to metronidazole or its analogues become widespread, new active compounds will be needed. This is especially relevant for T. vaginalis, which may be associated with increased HIV transmission.5 The objective of this work was to evaluate in vitro the activities of an N-(nitrothiazolyl) salicylamide, nitazoxanide6 [2-acetyloxy-N-(5-nitro 2-thiazolyl) benzamide] (Figure), and its metabolites,7,8 in comparison with metronidazole against isolates of G. intestinalis, E. histolytica and T. vaginalis, and to make a preliminary assessment of its mode of action.

Nitazoxanide was designed for broad-spectrum antiparasitic activity and low toxicity. It is active against the coccidian protozoan Cryptosporidium parvum in cell culture and in mice and piglets,9–11 the microsporans Encephalitozoon intestinalis and Vittaforma corneae in cell culture,12 the helminths Syphacia obvelata and Hymenolepis nana in mice13 and Taenia pisiformis, Uncinaria stenocephala and Trichuris vulpis in dogs.14In vitro studies have also shown a broad range of antimicrobial activity against facultative or obligate anaerobic and microaerophilic Gram-positive and Gram-negative bacteria15 including Helicobacter pylori.16 Clinical trials carried out in North and South America, Western Europe, the Middle East and Africa in more than 1500 patients have reported the effect of nitazoxanide on infections with C. parvum,17,18 the microsporan Enterocytozoon bieneusi,19E. histolytica, G. intestinalis, nematodes such as Ascaris lumbricoides, Ancylostoma duodenale and Trichuris trichiura,20 the cestodes Taenia saginata and H. nana21 and the liver trematode Fasciola hepatica.22

Following a single oral dose of [14C]nitazoxanide 500 mg in humans, 32% of the radioactivity was recovered in the urine and 66% in the faeces, indicating significant absorption from the gastrointestinal tract. The parent drug, nitazoxanide, is not found in blood, urine or faeces, but is presumably present to some extent in the contents of the intestine. The two major metabolites are produced by hydrolysis and glucurono conjugation; desacetyl-nitazoxanide (tizoxanide) is found in faeces, plasma and urine and tizoxanide glucuronide in plasma, urine and bile. The maximum concentration of tizoxanide in plasma following an oral dose of [14C]nitazoxanide 500 mg was c. 2 mg/L (6.5 μM), and its half-life in plasma was c. 1–2 h. Minor metabolites include salicyluric acid, tizoxanide sulphate and traces of hydroxytizoxanide in urine. Salicylate is found in faeces.8 Tizoxanide and to some extent nitazoxanide probably account for activity in the intestine, while in other locations tizoxanide may be the most important agent. However, high concentrations of tizoxanide glucuronide are excreted in bile,8 and the action of this compound may be manifested on liver trematodes such as F. hepatica.

Materials and methods

Sources of strains of protozoa

Strains of G. intestinalis (VNB1, VNB2, VNB5, EBC and EBE) were isolated from patientsapos; faecal samples in London.23 The last two strains were used with the agreement of Prof. John P. Ackers [London School of Hygiene and Tropical Medicine (LSHTM)]. The strains were isolated from diagnostic faecal samples from cases of giardiasis returning from the Indian subcontinent, subsequently successfully treated with metronidazole. All of these strains were determined by us to be susceptible to metronidazole in vitro. A further isolate of G. intestinalis (JKH-1) was isolated from a chronic human infection refractory to metronidazole by Dr Tim Paget (University of Hull), and was kindly supplied by him.

Two E. histolytica isolates, HM1:IMSS (Mexico) and NIH200 (Korea), had been held in culture and cryopreservation in the LSHTM laboratory since 1976, when they were supplied to us by the late Dr R. A. Neal (Wellcome Research Laboratories, Beckenham, UK). Isolates IULA: 1092:1 and IULA:0593:2 isolated from amoebic dysentery patients in Venezuela were kindly supplied by Professor John P. Ackers (LSHTM). Other isolates were purchased from the American Type Culture Collection (ATCC). The history and relevant references for the E. histolytica strains used are in Table 1.

A set of recent isolates of T. vaginalis from cases of vaginitis in Peru before treatment with nitazoxanide was received by air in ‘InPouch’ cultures. After establishment in continuous culture they were cryopreserved in our laboratory (isolates with ‘PER’ prefix). The history of the standard strains of T. vaginalis used is in Table 2.

Antimicrobial agents

The drugs used in the study were obtained as dry powders. Metronidazole was purchased from Sigma–Aldrich (Poole, UK), and nitazoxanide, tizoxanide, tizoxanide glucuronide, denitronitazoxanide and denitrotizoxanide were provided by Romark Laboratories, L.C. (Tampa, FL, USA).

In vitro studies

Stabilates were recovered from cryopreservation and cultured using Diamond' TYI-S-33 complete medium.34 Inactivated adult bovine serum was used at 10% and the vitamin–Tween 80 mixture at 3%. The complete medium was supplemented with bile 0.5 g/L (Sigma) for G. intestinalis.35 Parasites were routinely grown in 12 mL flat-sided plastic test tubes (Nunc) at 35°C. Medium was removed from cultures in the log phase of growth showing a monolayer of trophozoites (observed under the inverted microscope); they were then exposed to 10 mL fresh chilled culture medium at 0°C on an ice/water bath for 10 min and resuspended. For subculturing, 0.2–0.4 mL of this suspension was transferred to fresh tubes containing culture medium.

For the in vitro drug test the method described by Cedeño & Krogstad25 for E. histolytica was modified as follows. Stocks of metronidazole, dissolved in water, and tizoxanide, nitazoxanide, tizoxanide glucuronide, denitronitazoxanide and denitrotizoxanide, dissolved in DMSO, were prepared and stored at 4°C. Dilutions were made in serum-free culture medium and sterilized through a 0.2 μm filter. Final dilutions were made immediately before the test in complete medium to give twice the top concentration to be used for the in vitro testing. Two hundred microlitres of this was transferred to duplicate drug wells (B1–H1, in a flat-bottomed 96-well microtitre test plate) (Corning, New York, USA). Row A, wells 7–12 and rows B–H wells 2–12 received 100 μL of complete medium. Two-fold serial dilutions were made from B1–H1 to B12–H12. In row A, wells 1–12 were drug free, and 200 μL of complete culture medium was dispensed into wells 1–6 to serve as uninfected controls, while wells 7–12 were inoculated with 100 μL of parasite suspension as infected controls. Final concentrations of DMSO required in the tests for nitazoxanide and tizoxanide at IC50 values were 0.01% (v/v) or lower, and inactive in the assay systems. DMSO alone began to have inhibitory effects at 0.1% (v/v).

Parasite preparation

Healthy growing cultures in the log phase were selected for in vitro drug testing. Medium was carefully decanted, after gentle centrifugation where necessary, and replaced with fresh complete medium. Parasites were resuspended in 5 mL of medium. Twenty microlitres of suspension were fixed by adding 1 μL of 2% formalin and counts were made using a haemocytometer. The number of organisms per well varied in preliminary experiments with E. histolytica, the most difficult of the organisms to grow consistently, and we found that 50000 organisms/mL were needed in order to obtain optimal reproducibility with the range of strains tested. This is similar to the concentration of 30000/mL found to be optimal by Upcroft et al.4 for Entamoeba, and in contrast to 6000/mL as used by Cedeño & Krogstad25 in their work on strain HK9 of E. histolytica.

Dilutions were made in fresh tubes, with complete medium, to give a density of 100 000 organisms/mL and this was transferred in 100 μL volumes from a sterile vessel, frequently agitated, to each of wells 7–12 of row A (infected control) and to the test wells of rows B1–H1 to B12–H12. Each well (test and infected control) finally contained 10000 organisms in a final volume of 200 μL of drug–medium–parasite mixture (50 000 organisms/mL) and this system was applied to all three parasites. The final concentration of organisms in the Giardia and Trichomonas cultures is lower than that used by Upcroft et al.4

Incubation

The culture plates were placed in an airtight modular incubator chamber (Billups-Rothenberg, CA, USA) that had been swabbed clean with 70% ethanol and humidified with damp tissue treated with cupric sulphate solution (Sigma) to avoid bacterial and fungal contamination. The chamber was gassed for 5 min with a filter-sterilized gas mixture of 3% O2, 4% CO2 and 93% N2 (a gas phase suitable for microaerophilic organisms and for ‘aerobic’ tests on T. vaginalis).36 Incubation was at 35°C. After 24 h, plates were viewed on an inverted microscope to monitor parasite growth. Wells were dosed with 5 μL of methyl [3H]thymidine solution (Amersham, UK) to a final concentration of 0.2 μCi/well. Culture plates were returned to the chamber, gassed as above and incubated at 35°C for a further 24 h. Methyl-[3H]thymidine was obtained as 1 mL aqueous solution with a specific activity of 51 Ci/mmol and total activity of 1 mCi. This was transferred to a sterile Universal tube and vial rinsings added in a total volume of 25 mL serum-free culture medium (40 μCi/mL). This was filter-sterilized through a 0.2 μm filter and kept at −20°C until needed.

Assessment of drug activity

After addition of the radiolabelled thymidine and 24 h of incubation, plates were chilled at 4°C in an ice/water-bath. Parasites were harvested with a cell harvester (Skatron Inc., Liev, Norway) on to glass fibre filter paper (ICN Biomedicals Inc., CA, USA).37,38 After drying, the filter discs were punched out into scintillation vials. Three millilitres of scintillation fluid (Ecoscint; National Diagnostics, UK) were added and radioactivity counted (3 min per vial) in a Tricarb liquid scintillation spectrometer (Packard, Meriden, USA). The disintegrations per minute (dpm), representing the incorporation of methyl-[3H]thymidine by surviving parasites, were recorded. Counts from the uninfected control served as 100% inhibition and those from the infected control represented 0% inhibition. The dose–response was analysed using a non-linear regression (Levenberg-Marquardt algorithm) (XLFit programme #1.02, an add-in for Microsoft Excel). The dpm for each well was converted to percentage inhibition, which was plotted as a function of the logarithm of drug concentration. IC50s were obtained from the sigmoid curves.

All experiments were performed at least three times in duplicate, and the mean values with standard deviations are given in Tables 3–6. Multiplication factors to convert mg/L to μM are: metronidazole = 5.84; tizoxanide = 3.77; nitazoxanide = 3.254; tizoxanide glucuronide = 2.266; denitronitazoxanide = 3.81; denitrotizoxanide = 4.54.

Results

The results are expressed as molar drug concentrations that were required to obtain a reduction of 50% in tritiated thymidine incorporation into the culture (IC50) as compared with untreated controls.

Six isolates of G. intestinalis (Table 3)

One clearly resistant isolate was three times less susceptible to metronidazole than the mean for the susceptible isolates, and this differential was 13 times for tizoxanide, the active metabolite of nitazoxanide. Tizoxanide was eight times as active against susceptible isolates as metronidazole and twice as active against the resistant isolate. Nitazoxanide results paralleled those of tizoxanide, but for the metabolite tizoxanide glucuronide, which was about one-third as active as metronidazole, there was no detectable difference in activity against the resistant strain (JKH-1). The activity of tizoxanide glucuronide in G. intestinalis VNB1 did not differ from that of denitrotizoxanide (Table 4). Denitrotizoxanide was 48 times less active than tizoxanide, while denitronitazoxanide was 17 times less active than nitazoxanide (see Table 5).

Ten isolates of E. histolytica (see Table 4)

The mean metronidazole IC50 value for the most susceptible isolates of E. histolytica was 18.47 μM, and >30 μM was chosen as the cut-off for ‘resistance’. On the basis of literature values this was set at about three times the value for the most susceptible strain.31 Susceptibility to metronidazole showed a seven-fold range from 9.5 to 65.9 μM. On a molar basis, tizoxanide was 1.42 times as active against susceptible isolates as metronidazole, but was 2.44 times as active as that drug against the resistant isolates. Nitazoxanide results paralleled those of tizoxanide, but for tizoxanide glucuronide, which was about half as active as metronidazole on susceptible isolates, there was no detectable difference in activity against the resistant strains. Activity of tizoxanide glucuronide against E. histolytica HMI:IMSS was only marginally higher than that of denitrotizoxanide. Denitrotizoxanide was more than four times less active than tizoxanide, while denitronitazoxanide was 1.9 times less active than nitazoxanide (see Table 5).

Seventeen isolates of T. vaginalis (see Table 6)

The isolates of T. vaginalis used showed a 10-fold range of susceptibility to metronidazole from 0.6 to 60 μM. The cut-off for resistance in our study, following the susceptibility designations of the standard strains, was taken as an IC50 > 20 μM. These strains were still susceptible to tizoxanide. Against metronidazole-resistant isolates, tizoxanide was 4.88 times as active as metronidazole, and 1.39 times as active against susceptible isolates. Nitazoxanide results paralleled those of tizoxanide, but for tizoxanide glucuronide, which was about 0.33 times as active as metronidazole against the susceptible isolates, activity against the resistant strains was 1.93 times higher than metronidazole. Denitrotizoxanide was five times less active against susceptible T. vaginalis PER 014/CGF than tizoxanide, while denitronitazoxanide was 10 times less active than nitazoxanide (Table 4). Nitazoxanide results paralleled those of tizoxanide.

Discussion

Comparison of different studies is particularly difficult because different methods are used for activity measurement. Some workers use minimal lethal concentration (MLC), some MIC and some IC50. The ratio between mean susceptible and resistant values may afford a fair comparison between studies. Upcroft & Upcroft39 found that the mean MIC value for their strains of metronidazoleresistant G. intestinalis was 7.9 times higher than for susceptible strains (6.3 and 50 μM), compared with an IC50 ratio of 3.1 (4.93 and 15.42 μM) for our five susceptible and one resistant strains. For T. vaginalis there is the further complication of aerobic and anaerobic methods of measurement. Our comparison is with the former, which appears more clinically relevant. In metronidazole-susceptible clinical cases of trichomoniasis the isolates showed a geometrical mean aerobic MLC of 24.1 mg/L compared with 195.5 mg/L from drug refractory cases.33 This is a ratio of eight, which compares with our ratio of mean IC50 values for susceptible and resistant isolates of 6.22 (5.92 and 38.81 μM). The corresponding range of T. vaginalis aerobic MIC values found recently39 was 25–200 μM (also a ratio of eight). Among the standard strains of E. histolytica examined we found that the metronidazole IC50 ratio of mean ‘resistant’ to mean ‘susceptible’ was 2.79 (18.47 and 51.59 μM) while Aguirre-Cruz et al.31 reported a ratio of 3.3 of the IC50 for susceptible strains HK9 and HM1 to that of the less susceptible HM3 (1.75 and 5.84 μM). Recently a two-fold range in MIC values for laboratory-passaged E. histolytica strains (12.5–25 μM) was reported by Upcroft & Upcroft.39 A higher inoculum may be a possible reason why our IC50 values for HK9 clone 2 are much higher than those reported for uncloned strain HK9 by Cedeño & Krogstad.25 This clone has not, to our knowledge, been tested previously.

In G. intestinalis, tizoxanide was eight times more active than metronidazole against metronidazole-susceptible isolates and twice as active against a resistant isolate. In 10 axenic isolates of E. histolytica tested in the same way, tizoxanide was almost twice as active as metronidazole against more susceptible isolates and was more than twice as active against less susceptible isolates.

In T. vaginalis tizoxanide was 1.5 times as active as metronidazole against susceptible isolates but was nearly five times as active against resistant isolates.

In G. intestinalis, E. histolytica and T. vaginalis the glucuronide metabolite of nitazoxanide was 0.3, 0.63 and 0.33 times as active, respectively, as metronidazole in susceptible strains, and was 1.1, 1.59 and 1.9 times as active as metronidazole against resistant strains. The activity of tizoxanide glucuronide is appreciably less than that of tizoxanide and nitazoxanide, and little difference could be found between the resistant and the susceptible strains (Table 5).

This suggests that the glucuronide may not readily enter the cell to be activated by intracellular reduction. The reductions in tizoxanide activity brought about by removal of the nitro-group and by glucuronidation were remarkably similar within species (Table 5). The measured IC50 concentrations across the species (for denitrotizoxanide, 29.0, 37.0, 32.6 μM; for tizoxanide glucuronide, 33.2, 25.9, 19.6 μM) are remarkably similar considering the differing IC50 values for the parent tizoxanide. These observations support the idea that the glucuronide metabolite may not undergo nitro-reduction in these organisms and indicates that intracellular nitro-group reduction is necessary for activity of tizoxanide. The removal of the nitro-group renders reductive activation impossible, but may also have an impact on other properties of the drug, for example the lipophilic character is increased, and the yellow colour is lost, the latter feature indicating a decrease in the degree of conjugation of the ring system. It is however probable that, as with the 5-nitroimidazoles, the reducibility of the nitro-group is the most important feature involved.40

Nitazoxanide is active against a wider range of organisms than metronidazole. In addition, metronidazole-resistant strains of H. pylori were susceptible.16 Since resistance to metronidazole in G. intestinalis and T. vaginalis apparently depends on decreased ability to activate the drug,41 evidence of some cross-resistance and relative inactivity of denitro-derivatives presented in this paper helps to confirm the probable mode of action. Higher activity than metronidazole seen for nitazoxanide and tizoxanide in relatively metronidazole-refractory strains of all the three organisms studied indicates that resistance mechanisms to metronidazole may be to a variable extent bypassed by nitazoxanide and tizoxanide. This feature may relate to retention of the ability of the metronidazole-resistant organisms to reduce the drug at a less negative redox potential than that required for metronidazole. The redox potential reported for one electron reduction of another 5-nitrothiazole was −390 mV, much less negative than −486 mV seen in the 5-nitroimidazole group of metronidazole.42 The data we have so far on resistant strains are insufficient to draw firm conclusions on the mechanisms of resistance and need further expansion.

Although the effect of tizoxanide and related agents on C. parvum originally seemed to be anomalous, there is now evidence for an unusual cytosolic pyruvate oxidoreductase resembling pyruvate ferredoxin oxidoreductase,43 which may well be involved in tizoxanide nitro-reduction.

Table 1.

Standard strains of E. histolytica used

Strain From Date isolated Disease Origin Metronidazole (IC50 μM) Ref. 
50007 DKB ATCC 1924 dysentery UK – 24 
50542 HK9 clone-2 ATCC 1951 dysentery Korea (uncloned) 2.0 25 
30190 HB301:NIH ATCC 1960 dysentery Burma – 26 
50481 SD157 ATCC 1993 colitis USA – 27 
IULA:0593:2 JPA 1995 dysentery Venezuela – 28 
30886 Rahman ATCC 1964 colitis UK – 29 
30887 H303:NIH ATCC 1972 dysentery Vietnam? – 30 
(ATCC 30459) RAN 1967 dysentery Mexico 1.75 31 
HM1:IMSS 
IULA:1092:1 JPA 1995 dysentery Venezuela – 28 
(ATCC 30458) RAN 1948 dysentery seaman 1.87 32 
NIH 200 
Strain From Date isolated Disease Origin Metronidazole (IC50 μM) Ref. 
50007 DKB ATCC 1924 dysentery UK – 24 
50542 HK9 clone-2 ATCC 1951 dysentery Korea (uncloned) 2.0 25 
30190 HB301:NIH ATCC 1960 dysentery Burma – 26 
50481 SD157 ATCC 1993 colitis USA – 27 
IULA:0593:2 JPA 1995 dysentery Venezuela – 28 
30886 Rahman ATCC 1964 colitis UK – 29 
30887 H303:NIH ATCC 1972 dysentery Vietnam? – 30 
(ATCC 30459) RAN 1967 dysentery Mexico 1.75 31 
HM1:IMSS 
IULA:1092:1 JPA 1995 dysentery Venezuela – 28 
(ATCC 30458) RAN 1948 dysentery seaman 1.87 32 
NIH 200 

View Large

Table 2.

Standard strains of T. vaginalis used

Strain From Date isolated Origin R/S Ref. 
50143 (CDC085) ATCC 1980 USA 33 
50142 (RU393) ATCC 1983 USA 33 
FB2911 JPA 1997 UK – 
1910-SK JPA c. 1980 UK – 
UCH-1 JPA not known UK – 
50144 (CDC337) ATCC 1983 USA 33 
Strain From Date isolated Origin R/S Ref. 
50143 (CDC085) ATCC 1980 USA 33 
50142 (RU393) ATCC 1983 USA 33 
FB2911 JPA 1997 UK – 
1910-SK JPA c. 1980 UK – 
UCH-1 JPA not known UK – 
50144 (CDC337) ATCC 1983 USA 33 

View Large

Table 3.

IC50 values of nitazoxanide, tizoxanide and tizoxanide glucuronide compared with metronidazole for G. intestinalisa

Strain MTZ (μM) ± s.d.NTZ (μM) ± s.d.TIZ (μM) ± s.d.TIZG (μM) ± s.d.
JKH-115.42 ± 4.37.81 ± 1.7 8.18 ± 1.414.00 ± 1.3
EBE  5.95 ± 5.8 2.25 ± 2.5 0.641 ± 0.5  4.28 ± 1.2 
VNB1  5.72 ± 3.4 1.17 ± 1.1 0.603 ± 0.2  33.2 ± 0.9 
VNB5  5.20 ± 3.5 1.69 ± 0.9 0.641 ± 0.4 14.66 ± 5.4 
EBC  4.55 ± 2.2 1.07 ± 0.7 0.716 ± 0.3 11.87 ± 0.6 
VNB2  3.21 ± 1.2 1.20 ± 0.9 0.528 ± 0.4 15.22 ± 1.9 
Mean  6.68 ± 4.4 2.53 ± 2.6  1.89 ± 3.1 15.54 ± 9.5 
Mean (S)  4.93 ± 1.1 1.48 ± 0.49  0.63 ± 0.07 15.85 ± 10.6 
t-test versus S P = 0.0009 P = 0.0009 NS 
MTZ ratio (S)  3.3  7.8  0.3 
MTZ ratio (R)  2.0  1.9  1.1 
MTZ ratio (total)  2.6  3.5  0.4 
R/S  3.1  5.3  13.0  0.9 
Strain MTZ (μM) ± s.d.NTZ (μM) ± s.d.TIZ (μM) ± s.d.TIZG (μM) ± s.d.
JKH-115.42 ± 4.37.81 ± 1.7 8.18 ± 1.414.00 ± 1.3
EBE  5.95 ± 5.8 2.25 ± 2.5 0.641 ± 0.5  4.28 ± 1.2 
VNB1  5.72 ± 3.4 1.17 ± 1.1 0.603 ± 0.2  33.2 ± 0.9 
VNB5  5.20 ± 3.5 1.69 ± 0.9 0.641 ± 0.4 14.66 ± 5.4 
EBC  4.55 ± 2.2 1.07 ± 0.7 0.716 ± 0.3 11.87 ± 0.6 
VNB2  3.21 ± 1.2 1.20 ± 0.9 0.528 ± 0.4 15.22 ± 1.9 
Mean  6.68 ± 4.4 2.53 ± 2.6  1.89 ± 3.1 15.54 ± 9.5 
Mean (S)  4.93 ± 1.1 1.48 ± 0.49  0.63 ± 0.07 15.85 ± 10.6 
t-test versus S P = 0.0009 P = 0.0009 NS 
MTZ ratio (S)  3.3  7.8  0.3 
MTZ ratio (R)  2.0  1.9  1.1 
MTZ ratio (total)  2.6  3.5  0.4 
R/S  3.1  5.3  13.0  0.9 

View Large

Table 4.

IC50 values of nitazoxanide, tizoxanide and tizoxanide glucuronide compared with metronidazole for E. histolyticaa

Strain MTZ (μM) ± s.d. NTZ (μM) ± s.d. TIZ (μM) ± s.d. TIZG (μM) ± s.d. 
ATCC DKB65.88 ± 14.6017.34 ± 3.2528.09 ± 1.5123.87 ± 11.69
ATCC HK9-cl246.55 ± 4.0918.03 ± 0.1021.30 ± 2.6423.90 ± 3.77
ATCC HB30142.34 ± 12.8523.01 ± 2.2814.10 ± 4.9049.46 ± 1.89
ATCC SD175 29.78 ± 4.09 15.29 ± 2.60 10.56 ± 4.15 20.81 ± 4.90 
IULA:0593:2 28.50 ± 9.93  6.05 ± 1.30 18.02 ± 4.90 22.13 ± 4.90 
ATTC Rahman 26.46 ± 1.75 15.33 ± 0.65 22.70 ± 2.64 59.00 ± 6.79 
H303 14.13 ± 6.42 10.54 ± 1.95  7.54 ± 1.51  8.03 ± 0.75 
HM1:IMSS 11.21 ± 4.09  7.78 ± 0.98  9.46 ± 0.75 43.02 ± 7.92 
IULA:1092:1  9.69 ± 0.48  9.47 ± 1.63 16.89 ± 0.74 29.71 ± 12.82 
NIH 200  9.52 ± 6.42  6.38 ± 0.98  5.62 ± 0.75 12.18 ± 4.15 
Mean 28.41 ± 18.68 12.92 ± 5.70 15.43 ± 7.27 29.21 ± 16.35 
Mean (S) 18.47 ± 9.32 10.12 ± 3.88 12.97 ± 6.29 29.21 ± 16.35 
Mean (R) 51.59 ± 12.55 19.46 ± 3.09 21.16 ± 7.00 32.41 ± 14.77 
t-test versus S (PNS NS NS 
t-test versus R (P0.04 0.03 NS 
MTZ ratio (S) 1.83 1.42 0.63 
MTZ ratio (R) 2.65 2.44 1.59 
MTZ ratio (total) 2.20 1.80 1.00 
R/S 2.79 1.92 1.63 1.10 
t-test: S versus R (P0.01 0.005 NS NS 
Strain MTZ (μM) ± s.d. NTZ (μM) ± s.d. TIZ (μM) ± s.d. TIZG (μM) ± s.d. 
ATCC DKB65.88 ± 14.6017.34 ± 3.2528.09 ± 1.5123.87 ± 11.69
ATCC HK9-cl246.55 ± 4.0918.03 ± 0.1021.30 ± 2.6423.90 ± 3.77
ATCC HB30142.34 ± 12.8523.01 ± 2.2814.10 ± 4.9049.46 ± 1.89
ATCC SD175 29.78 ± 4.09 15.29 ± 2.60 10.56 ± 4.15 20.81 ± 4.90 
IULA:0593:2 28.50 ± 9.93  6.05 ± 1.30 18.02 ± 4.90 22.13 ± 4.90 
ATTC Rahman 26.46 ± 1.75 15.33 ± 0.65 22.70 ± 2.64 59.00 ± 6.79 
H303 14.13 ± 6.42 10.54 ± 1.95  7.54 ± 1.51  8.03 ± 0.75 
HM1:IMSS 11.21 ± 4.09  7.78 ± 0.98  9.46 ± 0.75 43.02 ± 7.92 
IULA:1092:1  9.69 ± 0.48  9.47 ± 1.63 16.89 ± 0.74 29.71 ± 12.82 
NIH 200  9.52 ± 6.42 

0 Replies to “Tizoxanide Synthesis Essay”

Lascia un Commento

L'indirizzo email non verrà pubblicato. I campi obbligatori sono contrassegnati *