Cellular mechanism of action of 2-nitroimidazoles as hypoxia-selective therapeutic agents

Abstract

Solid tumours are often poorly oxygenated, which confers resistance to standard treatment modalities. Targeting hypoxic tumours requires compounds, such as nitroimidazoles (NIs), equipped with the ability to reach and become activated within diffusion limited tumour niches. NIs become selectively entrapped in hypoxic cells through bioreductive activation, and have shown promise as hypoxia directed therapeutics. However, little is known about their mechanism of action, hindering the broader clinical usage of NIs. Iodoazomycin arabinofuranoside (IAZA) and fluoroazomycin arabinofuranoside (FAZA) are clinically validated 2-NI hypoxic radiotracers with excellent tumour uptake properties. Hypoxic cancer cells have also shown preferential susceptibility to IAZA and FAZA treatment, making them ideal candidates for an in-depth study in a therapeutic setting. Using a head and neck cancer model, we show that hypoxic cells display higher sensitivity to IAZA and FAZA, where the drugs alter cell morphology, compromise DNA replication, slow down cell cycle progression and induce replication stress, ultimately leading to cytostasis. Effects of IAZA and FAZA on target cellular macromolecules (DNA, proteins and glutathione) were characterized to uncover potential mechanism(s) of action. Covalent binding of these NIs was only observed to cellular proteins, but not to DNA, under hypoxia. While protein levels remained unaffected, catalytic activities of NI target proteins, such as the glycolytic enzyme glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and the detoxification enzyme glutathione S-transferase (GST) were significantly curtailed in response to drug treatment under hypoxia. Intraperitoneal administration of IAZA was well-tolerated in mice and produced early (but transient) growth inhibition of subcutaneous mouse tumours.

Keywords: Head and neck tumour; Hypoxia; Nitroimidazole; Replication stress.

Fig. 1

Hypoxic cells show preferential sensitivity towards IAZA and FAZA treatment. Chemical formula for IAZA (A) FAZA (B) and N₃-AZA (C). Crystal violet staining assays were performed with FaDu cells under different O₂ levels (20%, 1%, 0.5% or 0.1% O₂). Cells were more sensitive to drug treatment under low O₂ levels (D and E). Colony formation assays with FaDu cells measured clonogenicity in response to drug treatment under normoxia (20% O₂) or different levels of hypoxia (1%, 0.1% or <0.1% O₂). Cells showed the most sensitivity to treatment when cultured under <0.1% O₂ (G and H). IAZA was more toxic than FAZA at the same concentrations. Effects of different levels of O₂ on cell viability (F) and clonogenicity (I) are shown. Data represent mean ± S.E.M. from at least three independent experiments.

Fig. 2

IAZA and FAZA treatment do not induce apoptosis, senescence or ferroptosis under hypoxia, but reduce proliferation in hypoxic/reoxygenated cells. IAZA and FAZA treatment induced morphological changes in a dose dependent manner, but only under hypoxia (A). Compared to vehicle control, hypoxic/reoxygenated drug-treated cells did not show a statistically significant change in apoptotic (B), early apoptotic (C), necrotic (D) or live population (E). IAZA- and FAZA-treated hypoxic/reoxygenated cells stained poorly for beta-galactosidase activity (F). Addition of Ferrostatin-1 did not change the toxicity profile of IAZA (G) or FAZA (H). IAZA- or FAZA-treated hypoxic/reoxygenated cells show slower proliferation rates (I and J); shaded regions show duration of drug exposure. Data represent mean ± S.E.M; scale bar = 20 μm (A) and 100 μm (F).

Fig. 3

Hypoxic cells treated with IAZA and FAZA show slower S phase progression, increased chromatin-bound PCNA staining, reduced DNA synthesis, and signs of replication stress. Flow cytometric analysis for cell cycle phases showed an increase in S phase population in IAZA- and FAZA-treated hypoxic cells. These cells continuously showed slower S phase progression, even when drugs were removed, and hypoxic cells were allowed to reoxygenate (A). Hypoxic cells, when treated with IAZA and FAZA, showed increased chromatin bound PCNA staining, which is significantly higher than vehicle-treated hypoxic cells (B and C); yellow arrow shows mitotic cell (B). Hypoxic cells treated with IAZA and FAZA displayed reduced EdU incorporation (D). Intensity of EdU click-stained micrographs was quantified and plotted as intensity versus %population histograms. Hypoxic exposure by itself increased cell population with low EdU staining; incubation with IAZA and FAZA under hypoxia further increased “low EdU stained” cell fractions (E). Experimental setup for DNA fibre assay (F). IAZA- and FAZA-treated cells showed nascent fork degradation (G) and diminished ability to restart replication (H) in the DNA fibre assay. Representative micrographs are shown; scale bar = 20 μm (B and D). Quantification shows mean ± S.E.M. from three independent experiments. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 4

Analysis of DNA-NI binding characteristics and NI effects on DNA integrity under hypoxia. Trapped in agarose DNA click staining (TARDCS) assay revealed proteins, rather than DNA, form covalent adducts with NIs (A). Quantification of N₃-AZA click staining in TARDCS assays (B). FaDu cells treated with DMSO, IAZA and FAZA under normoxia and hypoxia (<0.1% O₂) were processed for γ-H2AX immunostaining (C). Quantifications of γ-H2AX immunostaining data are represented as histograms (D). A statistically significant increase in percent population with γ-H2AX staining higher than background was seen in hypoxic drug-treated cells (E). Alkaline comet assay showed a significant increase in comet tail moment only in IAZA-treated hypoxic cells (F). A significant increase in H₂O₂ levels was found in IAZA (G) and FAZA (H) treated hypoxic cells upon reoxygenation. Representative micrographs are shown; scale bar = 100 μm (A) and 20 μm (C). Quantification shows mean ± S.E.M. from at least three independent experiments.

Fig. 5

Effects of IAZA treatment on actin and cell adhesion properties. Extracts from FaDu cells treated with IAZA (100 μM or 150 μM) or vehicle control (0.02% DMSO) for 24 h under normoxia or hypoxia (<0.1% O₂) were processed for β-actin immunoblotting; HIF1A was used to show successful induction of hypoxia. No significant difference was observed in total actin levels (A). Hypoxic cells treated with IAZA showed altered actin cytoskeleton with fewer and stunted cell projections (highlighted with yellow arrows) (B). Drug-treated hypoxic cells, upon replating, showed compromised adhesion (C), metabolic capacities (D) and colony forming capacities (E). Representative immunoblots and micrographs are shown; scale bar = 20 μm. Data show the mean from three independent experiments; error bars represent S.E.M. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 6

The effects of IAZA and FAZA treatment on target proteins and glutathione levels under hypoxia. FaDu cells treated with IAZA (A) and FAZA (B) were processed for GAPDH and GSTP1 immunoblotting showing no effects on the protein levels in response to drug treatment; HIF1A indicates successful induction of hypoxia. Drug treatment significantly reduced GAPDH (C and D) and GST (E and F) enzymatic activities only under hypoxia. The cellular GSH pool remained unchanged in response to IAZA (G) and FAZA (H) treatment. Representative Western blot images are shown. Mean ± S.E.M. from three independent experiments are shown.

Fig. 7

In vivo evaluation of IAZA toxicity and tumour growth delay properties. IAZA was injected intraperitoneally in NOD/SCID/IL2R mice at 200, 400 or 600 mg/kg b.w. Changes in body weight were monitored for 14 days post injection (A), after which, mice were sacrificed, and histopathology analysis was performed on organ sections for possible signs of toxicity. Representative histopathology micrographs are shown (B). Tumour growth delay capacities of a single IAZA (400 mg/kg b.w.) administration was monitored in NU/NU nude mice bearing subcutaneous FaDu tumours. IAZA-treated mice showed an initial delay in tumour growth (C), with a small increase in duration of survival (D). Data show mean ± S.E.M.; scale bar = 1 mm (B).