N-Nitrosodimethylamine investigations in Muta™Mouse define point-of-departure values and demonstrate less-than-additive somatic mutant frequency accumulations

Abstract

The N-nitrosamine, N-nitrosodimethylamine (NDMA), is an environmental mutagen and rodent carcinogen. Small levels of NDMA have been identified as an impurity in some commonly used drugs, resulting in several product recalls. In this study, NDMA was evaluated in an OECD TG-488 compliant Muta™Mouse gene mutation assay (28-day oral dosing across seven daily doses of 0.02-4 mg/kg/day) using an integrated design that assessed mutation at the transgenic lacZ locus in various tissues and at the endogenous Pig-a gene-locus, along with micronucleus frequencies in peripheral blood. Liver pathology was determined together with NDMA exposure in blood and liver. The additivity of mutation induction was assessed by including two acute single-dose treatment groups (i.e. 5 and 10 mg/kg dose on Day 1), which represented the same total dose as two of the repeat dose treatment groups. NDMA did not induce statistically significant increases in mean lacZ mutant frequency (MF) in bone marrow, spleen, bladder, or stomach, nor in peripheral blood (Pig-a mutation or micronucleus induction) when tested up to 4 mg/kg/day. There were dose-dependent increases in mean lacZ MF in the liver, lung, and kidney following 28-day repeat dosing or in the liver and kidney after a single dose (10 mg/kg). No observed genotoxic effect levels (NOGEL) were determined for the positive repeat dose-response relationships. Mutagenicity did not exhibit simple additivity in the liver since there was a reduction in MF following NDMA repeat dosing compared with acute dosing for the same total dose. Benchmark dose modelling was used to estimate point of departure doses for NDMA mutagenicity in Muta™Mouse and rank order target organ tissue sensitivity (liver > kidney or lung). The BMD50 value for liver was 0.32 mg/kg/day following repeat dosing (confidence interval 0.21-0.46 mg/kg/day). In addition, liver toxicity was observed at doses of ≥ 1.1 mg/kg/day NDMA and correlated with systemic and target organ exposure. The integration of these results and their implications for risk assessment are discussed.

Keywords: Benchmark dose modelling; Muta™Mouse gene mutation assay; N-Nitrosodimethylamine; N-nitrosamine; point of departure.

Figure 1.

Peripheral blood micronucleus assay dose response relationships following exposure to NDMA. Summary of 14-day treatments in male CD2F1 mice (top row), 28-day repeat dose treatment in Muta™Mouse (middle row) and single bolus doses in Muta™Mouse (bottom row—blue lines). In the plots, dots represent data for individual animals with geometric group means overlaid (stars). Panels (A, D and G) are % CD71 + reticulocytes (RET), panels (B, E and H) are % CD71 + micronucleated reticulocytes (MN-RET), and panels (C, F, and I) are % CD71 + micronucleated normochromic erythrocytes (MN-NCE). There were no statistically significant increases in % MN-RET or % MN-NCE compared with the vehicle controls in any study arm. Red data points are placeholders on the X-axis and represent response information in the concurrent positive controls, i.e. cyclophosphamide in the 14-day treatment and Microflow positive control standard in the 28-day treatment (+ve ctrl(s)). Asterisks represent statistical significance levels relative to concurrent control (Dunnett’s post hoc T-test) at: *P = ≤.05, **P = ≤.01, ***P = ≤.001 levels.

Figure 2.

Peripheral blood Pig-a mutation assay dose response relationships following exposure to NDMA. Summary of 14-day treatments in male CD2F1 mice (top row), 28-day repeat dose treatment in Muta™Mouse (middle row) and single bolus doses in Muta™Mouse (bottom row—blue line). In the plots, dots represent data for individual animals with geometric group means overlaid (stars). Panels (B, E and H) are % CD24- mutant reticulocytes (mut-RET), and panels (C, F, and I) are % CD24- mutant red blood cells (mut-RBC). Panels (G, H and I) show percentage reticulocytes (%RET) included to provide a measure of cytotoxicity. There were no statistically significant increases in % mut-RET or % mut-RBC compared with the vehicle controls in any study arm. Red data points are placeholders on the X-axis and represent response information in the concurrent positive controls (+ve ctrl(s)).

Figure 3.

Mutant frequency (lacZ) dose response relationships following exposure to NDMA. Summary data in various tissues sampled after 28-day repeat dose treatments in Muta™Mouse (left-hand column—panels A, D, I, L, O, R) or single bolus dose treatments (middle column—blue lines: panels B, E, G, J, M, P, S). A cumulative dose comparison of the repeat (black) and single dose (blue) response relationships is shown in the right-hand column (panels C, F, H, K, N, Q, T). In the plots, dots represent data for individual animals with geometric group means overlaid (stars). Compared to the vehicle controls there were no statistically significant increases in lacZ mutant frequencies in bladder, bone marrow, spleen (high single-bolus dose analysed only, i.e. 10 mg/kg NDMA), or stomach. Statistically significant increases (P = ≤.01) in lacZ mutant frequencies were observed in kidney, lung, and liver at the higher doses and in kidney and liver (P = ≤ .05) following the high, single-bolus dose. In the cumulative dose plots, treatment groups with similar cumulative doses (i.e. 28 × 0.19 mg/kg/day vs 5 mg/kg; 28 × 0.36 mg/kg/day vs 10 mg/kg) where statistically compared. These comparisons were statistically significant in kidney (N) and liver (T) but only at the 10 mg/kg cumulative dose. Red circular data points are placeholders on the X-axis and represent response information in the concurrent positive controls (+ve ctrl(s)). Asterisks represent statistical significance levels at: *P = ≤.05, **P = ≤.01, ***P = ≤.001 levels.

Figure 4.

Benchmark dose determination and point-of-departure summaries using Muta™Mouse (lacZ) TGR data. Four-model average benchmark dose (BMD) determination by bootstrap sampling for a critical effect-size (i.e. benchmark response) of 50% for lung, kidney, and liver are shown in panels A, B, and C, respectively. Geometric mean responses for each treatment group are represented by the red triangles. In the plots on the right-side (D–F), dots represent response data for individual animals with geometric group means overlaid as stars for each tissue respectively. The BMD₅₀ values (yellow arrow) and 90% confidence intervals (pale yellow boundary) are presented to illustrate their relative position to the no observed genotoxic effect level (NOGEL) and lowest observed genotoxic effect level (LOGEL) values. Red circular data points are placeholders on the X-axis and represent response information in the concurrent positive controls (+ve ctrl(s)).

Figure 5.

Liver histopathology findings for Muta™Mouse tissues exposed to NDMA. Vehicle controls (Panels A/E are animal 1M1; panels I/L are animal 2M1) and NDMA treatment of 4.0 mg/kg/day (Panels B/F/J/M are animal 702M8) are shown (H&E stain). Note capsular indentations (B: indicated by stars and magnified in C) and reduction in hepatocellular vacuolation in B compared to A (control); also, pigment (F: black arrow and magnified in G) and single-cell necrosis (F: white arrow and magnified in H). Presence of pigment in the centrilobular area using Schmorl’s and Perls’ Prussian blue stain (see J compared with I and magnified in K). Mitotic figure (M: white arrow and magnified in N) and reduction in hepatocellular vacuolation (M compared with L). Inset scales are (C) 100 micron; (G/H/K/N) 50 micron.

Figure 6.

Hepatocyte % Ki67 positive and nuclear size dose response relationships for Muta™Mouse liver exposed to NDMA. Panels A–C show % Ki67-positive hepatocytes and panels D–E hepatocyte nuclear size (µm²). Data in panels A and D (black lines) are for repeat dose treatments with individual animal data shown as dots with geometric group means overlaid using stars. Data in panels B and E (blue lines) are for single bolus doses of NDMA (5 or 10 mg/kg). Panels C and F are composite Figures using cumulative dose on the X-axis to allow comparison of repeat (black) and single-dose (blue) regimens and include statistical comparison for the four-treatment groups with similar cumulative doses (i.e. 28 × 0.19 mg/kg/day vs 5 mg/kg; 28 × 0.36 mg/kg/day vs 10 mg/kg). These comparisons were not statistically significant (P = >0.05). In panels A/B/D/E, asterisks represent statistical significance levels relative to concurrent control at: *P = ≤.05, **P = ≤.01, ***P = ≤.001 levels.

Figure 7.

Compositive blood toxicokinetics in Muta™Mouse exposed to NDMA. The top two panels show area under the blood concentration–time curve (AUC) dose–response relationships. Panel A is for the 28-day repeat-dose arm, where data in green represents Day 1 sampling and data in black represents Day 28 sampling. Panel B shows Day 1 data (blue line) for the bolus-dose treatment arm. The bottom two panels show maximum concentration values (C_max) dose response relationships. Panel C is for the 28-day, repeat-dose arm where data in green represents Day 1 sampling and data in black represents Day 28 sampling. Panel D shows Day 1 data (blue line) for the bolus dose treatment arm.

Figure 8.

Concentrations of NDMA in the livers of Muta™Mouse specimens. Panel A shows NDMA levels in liver as a function of dose and sampling time measured on Day 1 whereas Panel B shows the equivalent data measured on Day 28 (i.e. after 28 days of exposure). The height and colour of the bars indicate the levels of NDMA at the various dose/time points sampled.