N -Methyl- N -Nitrosourea (MNU): A Positive Control Chemical for p53+/− Mouse Carcinogenicity Studies

Abstract

This study evaluated the effects of a single intraperitoneal injection of N-methyl-N-nitrosourea (MNU) in citrate buffer (pH 4.5) at a dose of 75 mg/kg in thirty male and thirty female p53+/− mice followed by a six-month observation period. Fifteen control mice per sex received a single intraperitoneal injection of citrate buffer. Fifty-six of sixty mice treated with MNU died or were sacrificed before the end of the observation period. Twenty-four males and twenty-seven females treated with MNU developed malignant lymphoma of the thymus; of these, twenty-three males and twenty-seven females had corresponding enlargement or masses in the thymus at necropsy. Lymphoblasts in thymic lymphomas stained positively for mouse CD3 antigen, indicating a T-cell lineage. One control female mouse had malignant lymphoma of the spleen that did not involve the thymus. Nine males and five females treated with MNU had adenomas or adenocarcinomas of the small intestine, whereas no intestinal neoplasms were observed in control mice. These findings support the use of a single dose of MNU as a positive control chemical in six-month p53+/− mouse carcinogenicity studies and suggest that examination of the thymus alone is sufficient to evaluate the validity of the model system.

Introduction

A six-month B6.129-Trp53^tmlBrdN5 heterozygous (p53+/−) mouse study is an acceptable alternative to a two-year mouse carcinogenicity study to support registration of pharmaceutical products that demonstrate positive results in at least one mutagenicity or clastogenicity assay (CPMP Safety Working Party 2004; ICH 1998; Jacobson-Kram et al. 2004; MacDonald et al. 2004). Regulatory agencies currently expect that alternative mouse carcinogenicity studies using p53+/−, Tg rasH2, or Tg.AC mice will include a positive control group (CPMP Safety Working Party 2004; MacDonald et al. 2004). The most commonly used and accepted positive control chemical for the p53+/− mouse carcinogenicity assay is p-cresidine (Delker et al. 2000; Floyd et al. 2002). Daily treatment of p53+/− mice with p-cresidine in the diet or by gavage produces hyperplasia, squamous metaplasia, and neoplasia of the transitional epithelium of the urinary bladder (Floyd et al. 2002; Sagartz et al. 1998), and transitional papillomas and transitional cell carcinomas usually occur in less than half of the mice following six months of treatment. In some studies, p-cresidine failed to produce statistically significant increases in the incidence of transitional cell neoplasia in one or both sexes (Jacobson-Kram et al. 2004). The U.S. FDA and others have suggested that alternative positive control chemicals for p53+/− mouse carcinogenicity studies would be desirable (Jacobson-Kram et al. 2004; MacDonald et al. 2004).

N-methyl-N-nitrosourea (MNU), benzene, and phenolphthalein have been suggested as better positive control agents than p-cresidine for p53+/− mouse carcinogenicity studies (Floyd et al. 2002; Hoivik et al. 2005). MNU is the accepted positive control chemical for the six-month rasH2 mouse alternative carcinogenicity model (Morton et al. 2002; Takaoka et al. 2003; Urano et al. 2001; Usui et al. 2001). In rasH2 mice, a single injection of MNU produces a high incidence of thymic malignant lymphoma with frequent involvement of spleen, lymph nodes, liver, and other parenchymal organs (Takaoka et al. 2003; Urano et al. 2001). Advantages to the use of MNU include: (1) most mice have treatment-related neoplasia; (2) few mice die from toxicity unrelated to neoplasia; (3) technical staff are exposed to very little carcinogen, since a single injection or oral gavage administration produces almost no aerosols and mice excrete MNU and metabolites for only a few days after treatment. If MNU consistently produces high incidences of neoplasia in p53+/− mice, MNU would be preferable to p-cresidine as a positive control chemical in these mice. This study was conducted to evaluate the effects of a single intraperitoneal injection of MNU at a dose of 75 mg/kg in p53+/− mice.

Materials And Methods

Male and female B6.129-Trp53^tmlBrdN5 heterozygous (p53+/−) mice were obtained from Taconic (Germantown, NY, USA). All experimental procedures were described in an animal use protocol approved by the institutional animal care and use committee. Mice were housed in individual polycarbonate cages with autoclaved hardwood chip bedding (Certified Sani-Chips, PJ Murphy Forest Products Corp., Montville, NJ, USA) supplemented with Nestlets (Ancare Corp., Bellmore, NY, USA) in compliance with all applicable regulations and the Guide for the Care and Use of Laboratory Animals (National Research Council 1996) in an Association for Accreditation and Assessment of Laboratory Animal Care (AAALAC) International-accredited animal care facility. The animals were fed Certified Rodent Diet 5002 meal (PMI Nutrition International, St. Louis, MO, USA) and provided with water ad libitum. After excluding animals with unacceptable pretest findings, animals were randomly assigned to treatment groups based on the most recent pretest body weight to achieve balance with respect to pretest body weights.

Thirty male and thirty female p53+/− mice were given 75 mg/kg MNU in citrate buffered saline adjusted to pH 4.5 as a single intraperitoneal injection on Day 1 followed by a six-month observation period. All mice were treated within sixty minutes of mixing the MNU solution. Fifteen additional control mice of each sex were given one intraperitoneal injection of citrate buffered saline. All mice were observed twice daily, and clinical signs were recorded once daily. Mice were palpated for masses once before treatment. Beginning three months after treatment, mice were palpated weekly until the end of the study. Body weights and food consumption were recorded at least once before treatment and weekly following treatment.

Animals found dead were refrigerated and necropsied at the earliest possible time. Animals killed at an unscheduled interval or at scheduled necropsy were anesthetized with isoflurane, killed by exsanguination from the abdominal vessels, and then necropsied. Tissues listed below were collected from all animals, preserved in 10% neutral buffered formalin, processed routinely, stained with hematoxylin-phloxine-eosin, and examined microscopically. Tissues examined included: stomach, duodenum, jejunum, ileum, gut-associated lymphoid tissue, cecum, colon, mesenteric lymph node, thymus, spleen, salivary gland, mandibular lymph node, Harderian gland, pancreas, urinary bladder, heart, aorta, skeletal muscle (biceps femoris), tongue, lung, liver, gallbladder, kidney, ureter, testis, epididymis, prostate, seminal vesicle, ovary, oviduct, uterus, cervix, vagina, thyroid gland, parathyroid gland, trachea, esophagus, adrenal gland, pituitary, spinal cord (thoracolumbar), brain, peripheral nerve (sciatic nerve), eye, optic nerve, skin and adnexa, mammary gland, sternum, bone marrow, stifle joint, larynx, nasal turbinate, gross lesions, rectum, preputial gland, clitoral gland, and Zymbal’s gland with external ear.

Immunohistochemical stains were applied to formalin-fixed, paraffin-embedded sections of thymus from the first five animals per sex with thymic malignant lymphoma (excluding animals that were found dead) in the MNU-treated group and to thymic sections from one control male and two control females. Antibodies directed against CD3 (T-lymphocyte marker), CD45R/B220 (a specific B-lymphocyte marker for mice), and F4/80 (macrophage marker) were used to classify the lineage of the neoplastic cells in the thymus. Sections for CD3 staining were pretreated by steaming at 96°C in Citra buffer (Biogenix, San Ramon, CA, USA) at pH 6 for twenty minutes. Thymic sections were incubated with anti-CD3 antibody (clone SP7, Thermo, Fremont, CA, USA) at a 1:750 dilution for sixty minutes at room temperature. CD3 immunoreactivity was detected using a biotinylated goat anti-rabbit secondary antibody followed by an avidin-biotin-horseradish peroxidase complex and visualized with diaminobenzidine. Sections for CD45R/B220 staining were pretreated by steaming at 96°C in Citra buffer at pH 6 for twenty minutes. Thymic sections were incubated with anti-CD45R/B220 antibody (clone RA3–6B2, BD Biosciences, San Jose, CA, USA) at a dilution of 1:1500 for sixty minutes at room temperature. CD45R/B220 immunoreactivity was detected using a biotinylated rabbit anti-rat secondary antibody followed by an avidin-biotin-horseradish peroxidase complex and visualized with diaminobenzidine. Sections for F4/80 staining were pretreated with pepsin (Dako, Carpinteria, CA, USA) for ten minutes at 37°C. Thymic sections were incubated with anti-F4/80 antibody (clone C1:A3–1, Serotec, Raleigh, NC, USA) at a dilution of 1:100 for sixty minutes at room temperature. F4/80 immunoreactivity was detected using a biotinylated rabbit anti-rat secondary antibody followed by an avidin-biotin-horseradish peroxidase complex and visualized with diaminobenzidine. All immunohistochemical sections were counterstained with Mayer’s hematoxylin (Dako, Carpinteria, CA, USA), dehydrated in graded concentrations of ethanol, and coverslipped routinely using permanent mounting medium.

Following completion of the microscopic tissue evaluation by the study pathologist, a second pathologist performed a peer review evaluation.

The Fisher least significant difference test was used to analyze body weight and food consumption. The Fisher exact test was used to analyze tumor incidence data. Statistical significance was evaluated at the p ≤ .05 level. Survival adjustments were not performed.

Results

Body Weight, Food Consumption, Clinical Signs, and Mortality

Survival, mean body weights and mean food consumption at representative time points, and clinical signs are presented in Table 1. Mean body weights were consistently lower in treated mice and mean food consumption was lower in mice treated with MNU than in controls at almost all time points. Treatment-related clinical signs of decreased activity, hunched posture, thin appearance, labored breathing, and rapid breathing were observed primarily or exclusively in mice with advanced neoplasia near the time of death.

Of the thirty mice of each sex treated with MNU, twenty-eight males and twenty-eight females died or were killed in extremis before the scheduled sacrifice date. Death was attributed to malignant lymphoma in twenty-three MNU-treated males and twenty-seven MNU-treated females. The first mouse to die of malignant lymphoma died fifty-six days after MNU injection. Death was attributed to adenocarcinoma of the jejunum in two males, and sarcoma of the prostate was the cause of death in one additional male. The cause of death was not determined for two males and two females that did not survive until scheduled sacrifice. All control animals survived until scheduled sacrifice.

Necropsy Findings

The incidences of treatment-related necropsy and microscopic findings are presented in Table 2. Microscopic findings of malignant lymphoma were associated with thymic enlargement, thymic mass/growth/nodule, thoracic mass/growth/nodule, splenic enlargement, lymph node enlargement, liver mass, liver enlargement, liver focus (focal or multifocal discoloration), kidney enlargement, and kidney discoloration. One jejunal mass observed at necropsy was diagnosed as intestinal adenocarcinoma. An abdominal mass in a control mouse was microscopically diagnosed as fat necrosis.

Microscopic Findings

Malignant lymphoma and intestinal adenoma and adenocarcinoma were the major neoplasms attributed to MNU treatment. Malignant lymphoma was associated with abnormalities of the thymus at necropsy in all affected mice except for one male mouse with microscopic malignant lymphoma that had a normal-appearing thymus at necropsy. Microscopic findings of malignant lymphoma included effacement of thymic corticomedullary architecture by diffuse sheets of lymphoblasts with large euchromatic nuclei; moderate to high numbers of mitotic figures; infiltration of lymphoblasts through the thymic capsule and into surrounding soft tissues including sternal musculature; and the presence of clusters or sheets of lymphoblasts in lymph nodes, spleen, liver, lung, kidney, bone marrow, eye, and other visceral organs (Figure 1). When malignant lymphocytes were present in sternal bone marrow, there was often loss of trabecular and cortical bone in the sternebrae. In sections of malignant lymphoma from five male and five female mice stained for CD3 antigen, neoplastic cells in all sections stained positively, indicating that the neoplasms were of T-cell origin (Figure 2). Neoplastic cells were negative for the B-lymphocyte marker CD45R (B220) and the macrophage marker F4/80. The incidences of malignant lymphoma in males and females were statistically significantly increased over control incidences (p < .01).

Adenomas and/or adenocarcinomas of the small intestine were observed in nine males and five females treated with MNU (Figures 3 and 4). Adenomas and adenocarcinomas contained glandular, crypt-like structures lined by one or more layers of columnar epithelial cells. In adenomas, the closely packed crypts were lined by large basophilic columnar epithelial cells that were sometimes pseudostratified. Adenomas were expansile, contained within the mucosa, and small. Adenocarcinomas had more pleomorphic, atypical, and sometimes flattened epithelial cells that formed irregular glandular structures in the mucosa and in deeper layers of the intestinal wall. The incidence of adenomas and adenocarcinomas combined in the small intestine in treated males was statistically significantly greater than in control males (p = .014). Although there were no intestinal neoplasms in control mice, the increased incidence of intestinal neoplasms in females was not statistically significant (p = .109). Intestinal epithelial neoplasms occurred in mice with concurrent focal or multifocal small intestinal glandular (crypt) hyperplasia.

Other neoplasms that occurred in single male animals treated with MNU included sarcoma of the prostate with lymph node metastases, sarcoma of the skin and adnexa, adenoma of the seminal vesicle, and sarcoma of the urinary bladder.

Two neoplasms unrelated to MNU treatment were found in control mice. Malignant lymphoma of the spleen was observed in one control female. This animal had no evidence of lymphoma in the thymus. Another control female had osteosarcoma in the sternum.

Treatment-related non-neoplastic microscopic findings were observed in the gastrointestinal tract, eye, sternum, spleen, lung, and heart. Hyperplasia of the epithelium of the forestomach and glandular stomach, hyperkeratosis of the forestomach, and epithelial hyperplasia of the small intestine were observed in two to twelve of sixty mice treated with MNU. Retinal degeneration with loss of the photoreceptor, outer nuclear, outer plexiform, and in some cases inner nuclear and inner plexiform layers was observed in most mice treated with MNU (Figure 5). In some mice with retinal degeneration, deeply basophilic round bodies were observed in the inner nuclear layer and between the inner nuclear layer and retinal pigment epithelium. The retinal degeneration seen in this study has been described previously in mice treated with MNU (Yuge et al. 1996). In the spleen, increased hematopoiesis involving predominantly the erythroid series was observed in almost all mice with lymphoma. Fibrinocellular inflammation of the pericardium and visceral pleura was observed in some mice with lymphoma, presumably related to compression and friction associated with extensive thymic enlargement.

Discussion

Most mice treated with MNU were euthanatized in moribund condition before the end of the study, with 50% survival reached approximately 100 days after treatment. Malignant lymphoma was the cause of death in most mice. MNU produced malignant lymphoma in 80% of males and 90% of females in this study. Every case of malignant lymphoma in mice treated with MNU involved the thymus. All but one mouse with malignant lymphoma had thymic findings at necropsy. In some mice, malignant lymphoma had spread to multiple organs including spleen, lymph nodes, liver, kidney, lung, pancreas, and eye. All control mice survived until the end of the study. Only one control mouse had malignant lymphoma of the spleen, and the thymus in this control animal was normal.

Others have reported a similarly high incidence of thymic malignant lymphoma in p53+/− mice treated with a single dose of MNU. In a study of eight- to nine-week-old p53+/− mice administered a single dose of 90 mg/kg MNU by gavage, ten of fifteen males and fourteen of fifteen females developed malignant lymphoma of the thymus by the end of the thirteen-week observation period, whereas no control p53+/− mice had malignant lymphoma (Hoivik et al. 2005). The incidence of lymphoma in vehicle control male and female p53+/− mice approximately eight months of age is approximately 2% (Mahler et al. 1998; Storer et al. 2001). The highest reported incidences of malignant lymphoma in single studies with negative control p53+/− mice are three of fifteen (20%) in females and two of fifteen (13%) in males (Storer et al. 2001). The consistently high incidence of thymic malignant lymphoma in p53+/− mice treated with MNU compared to the relatively low incidence of malignant lymphoma in vehicle control p53+/−mice demonstrates that MNU would be a consistent and reliable positive control agent for six-month p53+/− mouse carcinogenicity studies. Microscopic examination of the thymus alone would be sufficient to demonstrate susceptibility of the p53+/− mice to MNU.

The consistently high incidence of malignant lymphoma in p53+/− mice treated with MNU and the relatively low incidence of spontaneous malignant lymphoma in p53+/− mice suggests that a limited number of MNU-treated positive control mice would be needed to demonstrate a statistically significant increase in malignant lymphoma in mice treated with MNU. A power analysis (one-tailed Fisher exact test) confirmed that a study design including ten animals per sex in both negative and positive control groups would be sufficient to determine a statistically significant difference between the spontaneous incidence of lymphoma (~2% in males and females) and any incidence >50% in mice treated with MNU with statistical power of >83% (data not shown). Study designs with more animals in negative or positive control groups or a higher underlying true incidence of lymphoma would yield even greater power.

MNU has several advantages over p-cresidine as a positive control agent for p53+/− mouse studies. First, based on our data and published information, MNU produces a higher incidence of neoplasia than p-cresidine, whereas the response of p53+/− mice to p-cresidine is quite variable (Jacobson-Kram et al. 2004; Storer et al. 2001). Second, MNU is given once by intraperitoneal injection rather than daily in diet or by gavage, therefore use of MNU greatly reduces worker exposure to known carcinogens and the precautions needed to minimize human exposure. Third, the high overall incidence of malignant lymphoma (~85%) in mice treated with MNU and the low incidence of malignant lymphoma in control mice indicate that microscopic examination could be limited to the thymus of mice treated with MNU to demonstrate a positive neoplastic response. Finally, ten mice/sex should be sufficient to demonstrate a positive response to MNU if the incidence of lymphoma in mice treated with MNU is >50% (power >83%, p < .05).