In vivo rodent erythrocyte micronucleus assay. II. Some aspects of protocol design including repeated treatments, integration with toxicity testing, and automated scoring

Three-color labeling method for flow cytometric measurement of cytogenetic damage in rodent and human blood

Experiments described herein were designed to evaluate the performance characteristics of a flow cytometry-based system that scores the incidence of peripheral blood micronucleated reticulocytes (MN-RETs). These procedures represent the continued refinement of a previously reported anti-CD71-based method (Dertinger et al. [1996]: Mutat Res 371:283-292), with the following modifications: incorporation of a third fluorescent label to exclude platelets from the MN-RET region, and use of a CD71-associated fluorescence thresholding technique to increase data acquisition rates. Mouse, rat, and human blood samples were analyzed using both the previously described two-color procedure (anti-CD71-FITC and propidium iodide) and a newly developed three-color technique (which adds an antiplatelet-PE antibody). The rodent specimens were also evaluated by standard microscopy procedures (acridine orange staining). Mouse blood was collected via heart puncture of vehicle- and 5-fluorouracil-treated CD-1 mice; blood samples from saline-treated Sprague-Dawley rats were collected from the tail vein and via heart puncture. Rodent blood samples were analyzed by both the two- and three-color methods. Human blood specimens, obtained via arm venipuncture from cancer patients undergoing radiation therapy, were analyzed for MN-RETs using the two-color method. Subsequently, blood samples from a single chemotherapy patient were analyzed by both the two- and three-color methods. Finally, the chemotherapy patient blood samples and blood samples from 15 healthy volunteers were evaluated at very high densities in conjunction with a CD71-associated fluorescence thresholding technique. Results of these investigations showed that data from mouse blood analyzed by the two- and three-color procedures correlated well with microscopy data (r values = 0.917 and 0.937 for the two- and three-color methods, respectively); all three methods confirmed the genotoxicity of 5-FU. Data from rat tail vein samples showed improved reproducibility with the three-color technique, but no significant difference between the two techniques was seen with the heart puncture specimens. Human blood analyzed according to the two-color procedure produced unreliable results, as platelets and platelet aggregates impacted the rare MN-RET scoring region. The three-color technique effectively overcame this problem and produced reproducible measurements that fell within expected ranges. For human blood analyses, the high cell density/CD71-thresholding technique provided significant improvements over the low-density technique, as it allowed data acquisition to occur approximately six times faster with no loss of sensitivity.

Micronucleated CD71-positive reticulocytes: a blood-based endpoint of cytogenetic damage in humans

The frequency of micronuclei (also known as Howell-Jolly bodies) in peripheral blood erythrocytes of humans is extremely low due to the efficiency with which the spleen sequesters and destroys these aberrant cells. In the past, this has precluded erythrocyte-based analyses from effectively measuring chromosome damage. In this report, we describe a high-throughput, single-laser flow cytometric system for scoring the incidence of micronucleated reticulocytes (MN-RET) in human blood. Differential staining of these cells was accomplished by combining the immunochemical reagent anti-CD71-FITC with a nucleic acid dye (propidium iodide plus RNase). The immunochemical reagent anti-CD42b-PE was also incorporated into the procedure in order to exclude platelets which can interfere with analysis. This analytical system was evaluated with blood samples from ten healthy volunteers, one splenectomized subject, as well as samples collected from nine cancer patients before and over the course of radio- or chemotherapy. The mean frequency of MN-RET observed for the healthy subjects was 0.09%. This value is nearly two orders of magnitude higher than frequencies observed in mature erythrocytes, and is approximately half the MN-RET frequency observed for the splenectomized subject (0.20%). This suggests that the spleen's effect on micronucleated cell incidence can be minimized by restricting analyses to the youngest (CD71-positive) fraction of reticulocytes. Furthermore, MN-RET frequencies were significantly elevated in patients undergoing cancer therapy. Collectively, these data establish that micronuclei can be quantified in human peripheral blood reticulocytes with a single-laser flow cytometer, and that these measurements reflect the level of chromosome damage which has occurred in red marrow space.

Comparative scoring of micronucleated reticulocytes in rat peripheral blood by flow cytometry and microscopy

A flow cytometric technique for scoring the incidence of micronucleated reticulocytes in rat peripheral blood was compared to a standard microscopy-based procedure. For these studies, groups of five male Sprague-Dawley rats were treated with vehicle or a broad range of chemical genotoxicants: 6-thioguanine, N-methyl-N'-nitro-N-nitrosoguanidine, vincristine, methylaziridine, acetaldehyde, methyl methanesulfonate, benzene, monocrotaline, and azathioprine. Animals were treated once a day for up to 2 days, and peripheral blood was collected between 24 and 48 h after the final administration. These samples were processed for flow cytometric scoring and microscopy-based analysis using supravital acridine orange staining, and the percentage of reticulocytes and micronucleated reticulocytes was determined for each sample. The resulting data demonstrate good agreement between these scoring methodologies, although careful execution of the flow cytometric method was found to enhance the micronucleus assay by reducing both scoring time and scoring error. These data add further support to the premise that the peripheral blood compartment of rats can be used effectively to detect genotoxicant-induced micronuclei.

Urban air pollution induces micronuclei in peripheral erythrocytes of mice in vivo

In this study, we explored the role of chronic exposure to urban air pollution in causing DNA damage (micronuclei frequency in peripheral erythrocytes) in rodents in vivo. Mice (n=20) were exposed to the urban atmosphere of São Paulo for 120 days (February to June 1999) and compared to animals (n=20) maintained in the countryside (Atibaia) for the same period. Daily levels of inhalable particles (PM10), CO, NO(2), and SO(2), were available for São Paulo. Occasional measurements of CO and O(3) were made in Atibaia, showing negligible levels of pollution in the area. The frequency of micronuclei (repeated-measures ANOVA) increased with aging, the highest values obtained for the 90th day of experiment (P<0.001). The exposure to urban air pollution elicited a significant (P=0.016) increase of micronuclei frequency, with no significant interaction with time of study. Associations (Spearman's correlation) between pollution levels of the week that precede blood sampling and micronuclei counts were observed in São Paulo. The associations between micronuclei counts and air pollution were particularly strong for pollutants associated with automotive emissions, such as CO (P=0.037), NO(2) (P<0.001), and PM10 (P<0.001). Our results support the concept that urban levels of air pollution may cause somatic mutations.

Etoposide and merbarone are clastogenic and aneugenic in the mouse bone marrow micronucleus test complemented by fluorescence in situ hybridization with the mouse minor satellite DNA probe

The topoisomerase II (topo II) inhibitors etoposide (VP-16) and merbarone (MER) were investigated with the in vivo micronucleus test (MN test) combined with fluorescence in situ hybridization (FISH) using the mouse minor satellite DNA probe to discriminate MN of clastogenic and aneugenic origin. All experiments were performed with male (102/ElxC3H/El) F1 mice bred in the mouse colony of the GSF Research Center. The sample size per experimental group was five animals and 2,000 polychromatic erythrocytes (PCE) were scored per animal from coded slides in the conventional MN test. A separate set of coded slides was used for the FISH analysis. All treatments consisted of single intraperitoneal injections. Colchicine (COL, 3 mg/kg) and mitomycin (MMC, 1 mg/kg) were used as a positive control aneugen and clastogen, respectively, and these compounds produced the expected responses. A dose of 1 mg/kg VP-16 induced 3.44% MNPCE (compared to the concurrent solvent control of 0.37%, P < 0.001) and of these 39.9% (1.4% MNPCE) showed one or more fluorescent signals. MER (7.5-60 mg/kg) increased the MNPCE frequencies in a dose-dependent manner, with 15 mg/kg being the lowest positive dose. At the highest dose of 60 mg/kg of MER, a total of 4.26% MNPCE were found (compared to 0.31% in the concurrent solvent control, P < 0.001) and of these 46.2% (2.0% MNPCE) contained one or more fluorescent signals. The data demonstrate that VP-16 and MER induced both clastogenic and aneugenic events despite their different modes of topo II inhibition.

ECVAM and pharmaceuticals

In the pharmaceutical industry, toxicology testing is normally done by preclinical scientists during the Development phase. In the last decade, the implementation of high-throughput screens during the Discovery phase has resulted in an ever-increasing number of lead candidates to be selected for drug development. The low throughput of the conventional safety tests is a bottleneck in the drug-development process. The pharmaceutical industry needs new techniques, down-scaled tests and in vitro alternative test models to determine the absorption, distribution, metabolism, and excretion (ADME) and toxicology profiles of compounds in the late-Discovery phase and/or early in the Development phase. Medium-throughput ADME and toxicity tests will enhance the selection of safer new chemical entities for animals and/or humans. Consequently, this testing strategy will not only reduce the use of resources and the overall development time, but will also result in a substantial decrease in animal use.

Evaluation of genotoxic damage of cadmium chloride in peripheral blood of suckling Wistar rats

The aim of this study was to evaluate possible genotoxic damage of cadmium chloride exposure in suckling rats by means of the comet assay and the in vivo micronucleus test of rat blood lymphocytes, because no information is available on the genotoxic effect of cadmium in rats at this early age. Pups were receiving cadmium (as CdCl(2).H(2)O) orally in fractions of 0.5 mg for 9 days, totalling 4.5 mg Cd kg(-1) body wt, or were given a single subcutaneous injection of 0.5 mg Cd kg(-1) body wt. Some pups in both exposed groups were receiving calcium supplement (CaHPO(4).2H(2)O) in feed to reduce the body load of cadmium. Control pups did not receive either cadmium or calcium supplement. Cadmium in the carcass and organs was measured by atomic absorption spectrometry. The results showed that the cadmium body burden was significantly lower when the animals were receiving calcium supplements along with oral cadmium. The results of the micronucleus and comet assays showed significant differences between the control and exposed groups, regardless of the route of cadmium administration. The only statistically significant difference between the two exposed groups (oral cadmium and oral cadmium + calcium supplements) was in the number of micronuclei. The results of the comet assay showed that tail length differed statistically only between the control and all exposed groups, regardless of the route of cadmium administration. It can be concluded that the applied cadmium doses caused detectable genome damage but it was lower in calcium-treated pups receiving cadmium orally.

Collaborative validation study of the in vivo micronucleus test using mouse colonic epithelial cells

The in vivo micronucleus test using mouse colonic epithelial cells was evaluated as the 11th collaborative study organized by the Collaborative Study Group on the micronucleus test (CSGMT) with three model chemicals that were known to induce chromosome damage in mouse colonic cells. Five laboratories participated in this validation study. All three model chemicals, i.e. 1,2-dimethylhydrazine dihydrochloride (1,2-DMH), N-methyl-N-nitrosourea (MNU), and mitomycin C (MMC), induced micronucleated colonic epithelial cells in a 4-day exposure protocol in all participating laboratories. We confirmed that the present single cell suspension method could be used to detect the model chemicals as micronucleus inducers in mouse colonic epithelial cells. Advantages of this method are that experiments are easy to perform and that intact cells can be analyzed. The present study suggested that the colon micronucleus assay proposed here is useful for mechanistic studies of colon carcinogenesis.

Enumeration of micronucleated CD71-positive human reticulocytes with a single-laser flow cytometer

The extreme rarity of micronucleated reticulocytes (RETs) in the peripheral blood of non-splenectomized humans has precluded facile enumeration of these cells, as well as evaluation of this endpoint as an index of cytogenetic damage. In this report, we describe a high-throughput, single-laser flow cytometric system for scoring the incidence of micronuclei (MN) in newly formed human RETs. The procedure is based on an immunochemical reagent that differentially labels the most immature fraction of RETs from mature erythrocytes based on the expression level of the transferrin receptor (also known as CD71). The resolution of four erythrocyte populations (young RETs and mature erythrocytes, with and without MN) was achieved for human blood cells treated with phycoerythrin-conjugated anti-CD71, RNase, and either SYTOX Green or SYBR Green I nucleic acid dyes. Anti-glycophorin A labeling of erythroid cells (CyChrome conjugate) was also incorporated into the staining procedure to ensure that debris or other potential artifacts did not adversely impact the analyses. Instrument calibration procedures utilizing malaria-infected rodent erythrocytes were also developed, and are described. Using this analytical system, blood samples from 10 healthy non-splenectomized human volunteers were analyzed for micronucleus frequencies with a single-laser flow cytometer. Average micronucleus frequencies in the mature and most immature fraction of RETs were 0.016 and 0.19%, respectively. Blood samples from three healthy splenectomized volunteers were also evaluated. As expected, these samples exhibited higher micronucleus frequencies in the mature subset of erythrocytes (range 0.03-0.18%). The resulting data suggest that MN can be quantified in human erythrocyte populations with a single-laser flow cytometer, and that the frequency of MN cells in the youngest reticulocyte population approaches values expected in the absence of splenic selection against MN-erythrocytes. This high throughput system is potentially important for evaluating the value of the micronucleated reticulocyte endpoint as an index of chromosome breakage and/or chromosome segregational abnormalities in human populations.

Further evaluation of an in vivo micronucleus test on rat and mouse skin: results with five skin carcinogens

In a previous paper, we presented a practical in vivo micronucleus (MN) test that used rat skin as the target organ. To evaluate the test, as well as to determine the reproducibility and applicability of the method to mice, we used it to test the effect of five skin carcinogens (N-ethyl-N'-nitro-N-nitrosoguanidine (ENNG), N-methyl-N'-nitro-N-nitrosoguanidine (MNNG), 4-nitroquinoline 1-oxide (4NQO), 7,12-dimethylbenz[a]anthracene (DMBA), and benzo[a]pyrene (B[a]P)) on rat and mouse skin. All five compounds significantly and dose-dependently increased the MN frequencies in the basal cells of the chemical-treated skin. These results indicated the reproducibility of the test results and also the applicability of the test to mice as well as rats.

Dietary folate deficiency enhances induction of micronuclei by arsenic in mice

Folate deficiency increases background levels of DNA damage and can enhance the genotoxicity of chemical agents. Arsenic, a known human carcinogen present in drinking water supplies around the world, induces chromosomal and DNA damage. The effect of dietary folate deficiency on arsenic genotoxicity was evaluated using a mouse peripheral blood micronucleus (MN) assay. In duplicate experiments, male C57Bl/6J mice were fed folate-deficient or folate-sufficient diets for 7 weeks. During week 7, mice on each diet were given four consecutive daily doses of sodium arsenite (0, 2.5, 5, or 10 mg/kg) via oral gavage. Over the course of the study the folate-deficient diet produced an approximate 60% depletion of red blood cell folate. Folate deficiency by itself was associated with small but significant increases in MN in normochromatic erythrocytes (NCEs) and polychromatic erythrocytes (PCEs). Arsenic exposure was associated with significant increases in MN-PCEs in both folate-deficient and folate-sufficient mice. MN-PCE frequencies at the 10 mg/kg dose of arsenic were increased 4.5-fold over vehicle control in folate-deficient mice and 2.1-fold over control in folate-sufficient mice. At the 5 and 10 mg/kg doses of arsenic, MN-PCE levels were significantly higher (1.3-fold and 2.4-fold, respectively) in folate-deficient mice compared to folate-sufficient mice. Very few MN from either control or treated animals in either experiment exhibited kinetochore immunostaining, suggesting that the MN were derived from chromosome breakage rather than from whole chromosome loss. These results indicate that folate deficiency enhances arsenic-induced clastogenesis at doses of 5 mg/kg and higher.

Genotoxicity biomonitoring in regions exposed to vehicle emissions using the comet assay and the micronucleus test in native rodent Ctenomys minutus

Exposure to motor vehicle emissions represents an important concern for possible long-term health effects. The present report describes: 1) the application and verification of the alkaline comet assay in Ctenomys minutus to detect the possible genotoxicity of automobile emissions; 2) a comparison of the comet assay results with peripheral blood micronucleus (MN) assay results performed in the same animals; and 3) the identification of agents involved in the responses and in the seasonal variation of the effects. Ctenomys minutus (Octodontidae-Rodentia) were captured in two different fields from both sides of RS/030, a highway on the coastal plain of the Brazilian state of Rio Grande do Sul. Reference animals were obtained from a nearby field that was about 3 km distant from any road. By the end of this study, 123 rodents (73 females and 50 males) were live-trapped. Our results indicate that there was an increase in cells with DNA damage for C. minutus environmentally exposed to automobile emissions, as demonstrated by the alkaline comet assay, but there was no increase in micronucleated cells. The alkaline comet assay showed age and gender differences in the response. The comet assay results suggest that adult females are the principal population affected by air pollutants from vehicle emissions. Chemical data were also collected from areas exposed to automobile exhaust and these indicated that elevated levels of hydrocarbons, metals, and NO(2) were associated with the elevated levels of damaged cells observed in the wild rodent C. minutus. Our results agree with previous data on engine and fuel components, where weak increases in damage for native rodents exposed to emissions have been observed. Other larger, controlled studies are needed to better understand how the metabolism of C. minutus affects its response to emission exposure.

Micronucleus formation in mouse bone marrow cells in vivo in response to trans, trans-muconaldehyde

Benzene is a human leukemogen, and a bone marrow toxin and carcinogen in experimental animals. The reactive intermediates involved in benzene toxicity and their mechanism(s) of action have not been clearly delineated. We have investigated the clastogenic and cytotoxic effects of trans,trans-muconaldehyde (MUC), a reactive ring-opened benzene metabolite, in mouse bone marrow cells in vivo. Micronucleus formation was significantly increased when CD-1 mice were treated ip with MUC at 4 and 6 mg/kg/day for two days. These results suggest that leukemogenesis by benzene may be attributable, in part, to MUC-related clastogenic and cytotoxic effects in the bone marrow cells.

Flow cytometric enumeration of micronucleated reticulocytes: high transferability among 14 laboratories

This laboratory previously described a single-laser flow cytometric method, which effectively resolves micronucleated erythrocyte populations in rodent peripheral blood samples. Even so, the rarity and variable size of micronuclei make it difficult to configure instrument settings consistently and define analysis regions rationally to enumerate the cell populations of interest. Murine erythrocytes from animals infected with the malaria parasite Plasmodium berghei contain a high prevalence of erythrocytes with a uniform DNA content. This biological model for micronucleated erythrocytes offers a means by which the micronucleus analysis regions can be rationally defined, and a means for controlling interexperimental variation. The experiments described herein were performed to extend these studies by testing whether malaria-infected erythrocytes could also be used to enhance the transferability of the method, as well as control intra- and interlaboratory variation. For these studies, blood samples from mice infected with malaria, or treated with vehicle or the clastogen methyl methanesulfonate, were fixed and shipped to collaborating laboratories for analysis. After configuring instrumentation parameters and guiding the position of analysis regions with the malaria-infected blood samples, micronucleated reticulocyte frequencies were measured (20,000 reticulocytes per sample). To evaluate both intra- and interlaboratory variation, five replicates were analyzed per day, and these analyses were repeated on up to five separate days. The data of 14 laboratories presented herein indicate that transferability of this flow cytometric technique is high when instrumentation is guided by the biological standard Plasmodium berghei.

In vivo cytogenetics: mammalian germ cells

This chapter summarizes the most relevant methodologies available for evaluation of cytogenetic damage induced in vivo in mammalian germ cells. Protocols are provided for the following endpoints: numerical and structural chromosome aberrations in secondary oocytes or first-cleavage zygotes, reciprocal translocations in primary spermatocytes, chromosome counting in secondary spermatocytes, numerical and structural chromosome aberrations, and sister chromatid exchanges (SCE) in spermatogonia, micronuclei in early spermatids, aneuploidy in mature sperm. The significance of each methodology is discussed. The contribution of novel molecular cytogenetic approaches to the detection of chromosome damage in rodent germ cells is also considered.

In vivo rodent micronucleus assay: protocol, conduct and data interpretation

In vivo rodent micronucleus assay has been widely used to detect genotoxicity. Evaluation of micronucleus induction is the primary in vivo test in a battery of genotoxicity tests and is recommended by the regulatory agencies around the globe to be conducted as part of product safety assessment. The assay, when performed appropriately, detects both clastogenicity and aneugenicity. Methods for performing micronucleus evaluation have evolved since its initial description in the 1970s. In recent years, the focus has been directed toward improving micronucleus detection with high efficiency by proposing data-based recommendations to the standard initial protocol design. Such improvements include, e.g., the use of appropriate harvest time(s), inclusion of one or both sexes, number of doses tested, limit dose, integrating micronucleus assessment into the routine toxicology studies, use of fluorescent staining, automation of micronucleus detection and assessment of micronuclei in multiple tissues. This protocol paper describes: the mechanism of micronucleus formation, a generalized protocol for manual detection, enumeration of micronuclei, and data interpretation in light of published information thus far, on the regulatory aspects of this assay. Certain recent protocol issues that are practical in nature are equally valid in relation to standard manual method and provide robust database, which are also included for consideration. It is expected that such improvements of the protocol will continue to drive the utility of this assay in the product safety assessment.

Historical vehicle and positive control micronucleus data in mice and rats

The rodent bone marrow micronucleus (MN) assay has been widely used as part of an in vivo genotoxicity test battery in product safety evaluation. In this assay, the historical vehicle and positive control data form an important component in the assay performance and data interpretation. Also, in light of minimizing animal use in research and still obtain required data from a study, the routine use of positive control in every MN assay has been questioned by the scientific community, especially in laboratories which have demonstrated assay reproducibility and conduct studies under Good Laboratory Practice regulations. In this paper, mouse and rat vehicle and positive control MN data, collected manually, are described as a reference for a period of 12 years (1987-1998) in our laboratory. The vehicles generally included a variety of aqueous solutions and suspensions and cyclophosphamide dosed intraperitoneally at 20mg/kg (rats) or 40 mg/kg (mice) served as positive control, in all studies. Based on combined sex data (430 animals), for CD(1) mice, the vehicle control MN polychromatic erythrocyte (PCE) range was 0.9-3.1 with a mean of 1.75 per 1000 PCE and the positive control range (220 animals) was 8.8-42.1 with a mean of 23.1 MNPCE per 1000 PCE. Similarly, for Wistar rats, the vehicle control range (360 animals) was 1.3-5.3 with a mean of 2.6 MNPCE per 1000 PCE and the positive control range (240 animals) was 10.4-33.8 MNPCE per 1000 PCE. Vehicle control ranges reported here are comparable to the literature database and the positive control response was > or = 4-fold over vehicle control, in all studies. These data demonstrate the reproducibility of positive control response in MN assay in our laboratory and support the MN Assay Expert Panel's view that the use of positive control may not be necessary in every study.