Hepatotoxicity in mice of a novel anti-parasite drug candidate hydroxymethylnitrofurazone: a comparison with Benznidazole

Abstract

Background: Treatment of Chagas disease, caused by Trypanosoma cruzi, relies on nifurtimox and benznidazole (BZL), which present side effects in adult patients, and natural resistance in some parasite strains. Hydroxymethylnitrofurazone (NFOH) is a new drug candidate with demonstrated trypanocidal activity; however, its safety is not known.

Methods: HepG2 cells dose response to NFOH and BZL (5-100 µM) was assessed by measurement of ROS, DNA damage and survival. Swiss mice were treated with NFOH or BZL for short-term (ST, 21 d) or long-term (LT, 60 d) periods. Sera levels of cellular injury markers, liver inflammatory and oxidative stress, and fibrotic remodeling were monitored.

Results: HepG2 cells exhibited mild stress, evidenced by increased ROS and DNA damage, in response to NFOH, while BZL at 100 µM concentration induced >33% cell death in 24 h. In mice, NFOH ST treatment resulted in mild-to-no increase in the liver injury biomarkers (GOT, GPT), and liver levels of inflammatory (myeloperoxidase, TNF-α), oxidative (lipid peroxides) and nitrosative (3-nitrotyrosine) stress. These stress responses in NFOH LT treated mice were normalized to control levels. BZL-treated mice exhibited a >5-fold increase in GOT, GPT and TNF-α (LT) and a 20-40% increase in liver levels of MPO activity (ST and LT) in comparison with NFOH-treated mice. The liver inflammatory infiltrate was noted in the order of BZL>vehicle≥NFOH and BZL>NFOH≥vehicle, respectively, after ST and LT treatments. Liver fibrotic remodeling, identified after ST treatment, was in the order of BZL>vehicle>NFOH; lipid deposits, indicative of mitochondrial dysfunction and in the order of NFOH>vehicle>BZL were evidenced after LT treatment.

Conclusions: NFOH induces mild ST hepatotoxicity that is normalized during LT treatment in mice. Our results suggest that additional studies to determine the efficacy and toxicity of NFOH are warranted.

Figure 1. Hepatocytes dose response to NFOH and BZL.

HepG2 cells (7.5×10⁴/well) were incubated in presence of 5–100 µM NFOH or BZL for 24 h. Cells incubated with vehicle solution (10% DMSO) were used as controls. Shown are flow cytometry evaluations of (A) CellROX Green oxidation by intracellular ROS and (B) 8-oxo-dG levels as an indicator of DNA damage. The mean fluorescence intensity (MFI) of CellROX (A.a) and 8-oxo-dG (B.a) as an indicator of the extent of ROS production and DNA damage, respectively; and the mean percentage of CellROX⁺ (A.b) and 8-oxo-dG⁺ (B.b) cells was determined. (C) HepG2 cells were incubated with 5–100 µM NFOH or BZL as above. Cell survival was determined at 24 h (C.a) and 48 h (C.b) post-treatment by an AlamarBlue assay. Data are presented as mean ± S.D. and derived from two independent experiments (triplicate observations per treatment per experiment). Significance is shown as *p<0.05, **p<0.01, ***p<0.001 (vehicle control-versus-treated).

Figure 2. (A) Experimental design for the toxicity studies in mice.

Mice received an oral dose of 150 mg/kg/day of NFOH, BZL or the vehicle (9% NaCl - 5%Tween-80) once a day for 6 days per week. Mice were treated for 21 or 60 days to determine the toxicity of short-term (ST) and long-term (LT) treatment, respectively. (B–D) Liver toxicity of anti-parasite drugs. Glutamate oxaloacetate transaminase (B&D) and glutamate pyruvate transaminase (C&E) activities in mice sera (B&C) and liver homogenates (D&E) after ST and LT treatment with NFOH and BZL were determined as an indicator of liver injury. Mice treated with vehicle only were used as controls. Data in figures 2–7 are presented as mean ± S.D. (n≥9 mice per group, triplicate observations per mouse). Significance is shown as *p<0.05, **p<0.01, ***p<0.001 (vehicle control-versus-treated).

Figure 3. NFOH causes short-term phagocytic activation and oxidative damage.

Mice were treated with NFOH and BZL as in Fig. 1, and liver tissue homogenates were prepared. (A) Myeloperoxidase activity was measured as a biomarker of phagocyte activation. (B) Lipid hydroperoxides levels were evaluated as a marker of oxidative stress. (C&D) Liver homogenates were subjected to Western blotting (C) with anti-3-nitrotyrosine antibody (Control: anti-β-actin antibody). Densitometry quantitation of 3-NT band intensity, normalized to β-actin, is presented in panel D.

Figure 4. Liver inflammatory cytokine response to treatment with anti-parasite drugs.

Liver tissues were harvested after ST and LT treatment with NFOH, BZL, and vehicle, as detailed in Fig. 1. (A) Shown are the relative changes in gene expression for TNF-α, determined by real time RT-PCR. Results were normalized to GAPDH mRNA. (B) TNF-α levels in tissue homogenates were determined by an ELISA.

Figure 5. NFOH does not cause chronic liver inflammation.

Paraffin-embedded liver tissue sections (5 µm) were submitted to H&E staining (blue: nuclear, pink: muscle/cytoplasm). (A) Shown are the representative images (magnification: 40×) of liver tissue-sections from mice treated with vehicle (a&b), NFOH (c&d), and BZL (e&f) for short-term (a,c,e) and long-term (b,d,f). (B) Inflammatory infiltrate in liver tissue sections (3-sections/mouse, n>3/group) was scored as described in Materials and Methods.

Figure 6. Liver remodeling in response to NFOH treatment.

Mice were treated with vehicle solution (a&d), NFOH (b&e) or BZL (c&f) for a short-term (a,b,c) and long-term (d,e,f). (A) Representative micrographs of liver tissue-sections stained with Masson's Trichome (magnification: 400×, scale bar: 100 µm) are shown. (B) Tissue sections (3-sections/mouse, n>3/group) were scored for fibrosis as described in Materials and Methods.

Figure 7. NFOH does not cause metabolic stress.

Mice were treated for ST (a,b,c) and LT (d,e,f) with vehicle solution (a&d), NFOH (b&e) or BZL (c&f). (A) Liver deposition of lipids, an indicator of metabolic dysfunction, was determined by Oil Red O staining. Shown are the representative micrographs (magnification: 600×). (B) Tissue sections (3-sections/mouse, n>3/group) were scored for lipid deposition, as described in Materials and Methods.