Cytokinesis-block micronucleus assay evolves into a "cytome" assay of chromosomal instability, mitotic dysfunction and cell death

The cytokinesis-block micronucleus (CBMN) assay was originally developed as an ideal system for measuring micronuclei (MNi) however it can also be used to measure nucleoplasmic bridges (NPBs), nuclear buds (NBUDs), cell death (necrosis or apoptosis) and nuclear division rate. Current evidence suggests that (a) NPBs originate from dicentric chromosomes in which the centromeres have been pulled to the opposite poles of the cell at anaphase and are therefore indicative of DNA mis-repair, chromosome rearrangement or telomere end-fusions, (b) NPBs may break to form MNi, (c) the nuclear budding process is the mechanism by which cells remove amplified and/or excess DNA and is therefore a marker of gene amplification and/or altered gene dosage, (d) cell cycle checkpoint defects result in micronucleus formation and (e) hypomethylation of DNA, induced nutritionally or by inhibition of DNA methyl transferase can lead to micronucleus formation either via chromosome loss or chromosome breakage. The strong correlation between micronucleus formation, nuclear budding and NPBs (r = 0.75–0.77, P < 0.001) induced by either folic acid deficiency or exposure to ionising radiation is supportive of the hypothesis that folic acid deficiency and/or ionising radiation cause genomic instability and gene amplification by the initiation of breakage–fusion–bridge cycles. In its comprehensive mode, the CBMN assay measures all cells including necrotic and apoptotic cells as well as number of nuclei per cell to provide a measure of cytotoxicity and mitotic activity. The CBMN assay has in fact evolved into a “cytome” method for measuring comprehensively chromosomal instability phenotype and altered cellular viability caused by genetic defects and/or nutrional deficiencies and/or exogenous genotoxins thus opening up an exciting future for the use of this methodology in the emerging fields of nutrigenomics and toxicogenomics and their combinations.     

Cytokinesis-block micronucleus cytome assay

The cytokinesis-block micronucleus cytome assay is a comprehensive system for measuring DNA damage, cytostasis and cytotoxicity. DNA damage events are scored specifically in once-divided binucleated (BN) cells and include (a) micronuclei (MNi), a biomarker of chromosome breakage and/or whole chromosome loss, (b) nucleoplasmic bridges (NPBs), a biomarker of DNA misrepair and/or telomere end-fusions, and (c) nuclear buds (NBUDs), a biomarker of elimination of amplified DNA and/or DNA repair complexes. Cytostatic effects are measured via the proportion of mono-, bi- and multinucleated cells and cytotoxicity via necrotic and/or apoptotic cell ratios. Further information regarding mechanisms leading to MNi, NPBs and NBUDs formation is obtained using centromere and/or telomere probes. The assay is being applied successfully for biomonitoring of in vivo genotoxin exposure, in vitro genotoxicity testing and in diverse research fields such as nutrigenomics and pharmacogenomics as well as a predictor of normal tissue and tumor radiation sensitivity and cancer risk. The procedure can take up to 5 days to complete.     

Mussel micronucleus cytome assay

The micronucleus (MN) assay is one of the most widely used genotoxicity biomarkers in aquatic organisms, providing an efficient measure of chromosomal DNA damage occurring as a result of either chromosome breakage or chromosome mis-segregation during mitosis. The MN assay is today applied in laboratory and field studies using hemocytes and gill cells from bivalves, mainly from the genera Mytilus. These represent 'sentinel' organisms because of their ability to survive under polluted conditions and to accumulate both organic and inorganic pollutants. Because the mussel MN assay also includes scoring of different cell types, including necrotic and apoptotic cells and other nuclear anomalies, it is in effect an MN cytome assay. The mussel MN cytome (MUMNcyt) assay protocol we describe here reports the recommended experimental design, sample size, cell preparation, cell fixation and staining methods. The protocol also includes criteria and photomicrographs for identifying different cell types and scoring criteria for micronuclei (MNi) and nuclear buds. The complete procedure requires approximately 10 h for each experimental point/sampling station (ten animals).     

The use of the in vitro micronucleus assay to detect and assess the aneugenic activity of chemicals

The successful validation of the in vitro micronucleus assay by the SFTG now provides the opportunity for this highly cost effective assay to be used to screen chemicals for their ability to induce both structural (clastogenic) and numerical (aneugenic) chromosome changes using interphase cells. The use of interphase cells and a relatively simple experimental protocol provides the opportunity to greatly increase the statistical power of cytogenetic studies on chemical interactions. The application of molecular probes capable of detecting kinetochores and centromeres provides the opportunity to classify mechanisms of micronucleus induction into those which are primarily due to chromosome loss or breakage. When a predominant mechanism of micronucleus induction has been shown to be based upon chromosome loss then further investigation can involve the determination of the role of non-disjunction in the induction of aneuploidy. The binucleate cell modification of the in vitro micronucleus assay can be combined with the use of chromosome specific centromere probes to determine the segregation of individual chromosomes into daughter nuclei. The combination of these methods provides us with powerful tools for the investigation of mechanisms of genotoxicity particularly in the low dose regions.     

A lifetime passion for micronucleus cytome assays—Reflections from Down Under

A brief account of an improbable career in the field of genetic toxicology is given, extending from my early years in Malta through a life-changing decision to study in Australia (Down Under). I describe the circumstances that led to the discovery of the cytokinesis-block micronucleus (CBMN) assay and its evolution into a cytome assay of chromosome breakage and loss (micronuclei), asymmetrical chromosome rearrangements or telomere end fusions (nucleoplasmic bridges), gene amplification (nuclear buds), cell death (necrosis, apoptosis) and cytostasis (nuclear division index). This paper also describes the role of my laboratory in the beginning of the HUMN project, its achievements, and the applications of CBMN cytome assays in the fields of radiation biology, genetic toxicology, epidemiology, biodosimetry and genome health nutrigenomics, leading to the Genome Health Clinic concept. Along the way I mention my encounters with some of the influential people in the field of mutagenesis who provided me with the motivation and guidance needed to realise these achievements. I hope this account provides some inspiration to the next generation of scientists who may be fortunate to see the realisation of the application of the principles of mutagenesis in health optimisation or disease prevention and eventually in mainstream medicine.     

The nitroxide antioxidant Tempol affects metal-induced cyto- and genotoxicity in human lymphocytes in vitro

Stable, membrane-permeating nitroxide radicals have been reported to possess antioxidant activity in various experimental systems while, in parallel, they have been considered to be evidently harmful oxidants. The aim of this study was to evaluate the role of the piperidine nitroxide 4-hydroxy-2,2,6,6-tetramethylpiperidine-1-oxyl (Tempol) in the modulation of cyto- and genotoxicity in human lymphocytes in vitro by cadmium and chromium, which depend, at least in part, on formation of reactive oxygen species. The cytokinesis-block micronucleus (CBMN) assay to measure micronucleus (MN) formation, the nuclear division index (NDI) and the percentage of apoptotic and necrotic cells were assessed after exposure human lymphocytes to Cd(II), Cr(III) and Cr(VI) or co-incubation with these metals and Tempol. We found a significant ability of 5-50 microM Tempol to diminish toxic effects of the agents tested. In every system studied, Tempol decreased the micronucleus frequency and the percentage of apoptotic and necrotic cells, while it increased the nuclear division index (p<0.05). We observed adverse effects when 0.1-1 mM Tempol alone was used: inhibition of cell growth, induction of apoptotic and necrotic cell death and chromosomal damage (p<0.05). Collectively, we demonstrated that Tempol can be considered as a potent anti-apoptotic and antigenotoxic agent, but also as a cytotoxic and clastogenic chemical, depending on the concentration applied.     

Assessment of the genetic risks of a metallic alloy used in medical implants

The use of artificial implants provides a palliative or permanent solution for individuals who have lost some bodily function through disease, an accident or natural wear. This functional loss can be compensated for by the use of medical devices produced from special biomaterials. Titanium alloy (Ti-6Al-4V) is a well-established primary metallic biomaterial for orthopedic implants, but the toxicity of the chemical components of this alloy has become an issue of concern. In this work, we used the MTT assay and micronucleus assay to examine the cytotoxicity and genotoxicity, respectively, of an extract obtained from this alloy. The MTT assay indicated that the mitochondrial activity and cell viability of CHO-K1 cells were unaffected by exposure to the extract. However, the micronucleus assay revealed DNA damage and an increase in micronucleus frequency at all of the concentrations tested. These results show that ions released from Ti-6Al-4V alloy can cause DNA and nuclear damage and reinforce the importance of assessing the safety of metallic medical devices constructed from biomaterials.     

Studying the genotoxic effects induced by two kinds of bentonite particles on human B lymphoblast cells in vitro

The aim of the present study was to evaluate the genotoxic effects induced by native and active bentonite particles (BPs) on human B lymphoblast cells using comet assay and cytokinesis-block micronucleus (CBMN) assay in vitro. The cells were exposed to BPs at the concentrations of 30, 60, 120 and 240μg/ml for 24, 48 and 72h, respectively. The quartz contents of native and active BPs were 6.80±0.20 and 6.50±0.10%, respectively. Gypsum and DQ-12 quartz served as negative and positive controls. The results of comet assay showed that DNA damage induced by native and active BPs was significantly higher than that induced by gypsum control (P<0.05 or <0.01), and increased with exposure concentration and duration. When the cells were exposed to BPs at the doses of 120 and 240μg/ml for 72h, DNA damage induced by active BPs and native BPs was significantly higher than that induced by DQ-12 quartz (P<0.01), and DNA damage induced by active BPs enhanced significantly, as compared with native BPs (P<0.01). The results of CBMN assay demonstrated that both native BPs and active BPs could induce significant micronuclei, as compared with gypsum control (P<0.05 or <0.01). However, there was no significant difference of micronucleus frequency (MNF) among native BPs, active BPs and DQ-12 quartz. The water-soluble fractions from two kinds of BPs did not induce significant DNA damage and micronuclei. These findings indicated that the genotoxicity induced by active BPs and native BPs could be detected in comet assay and CBMN assay in vitro, the insoluble particle fractions from BPs may play a main role in the genotoxic effects induced by BPs.     

Cyto- and genotoxicity of ultrafine TiO2 particles in cultured human lymphoblastoid cells

Titanium dioxide is frequently used in the production of paints, paper, plastics, welding rod-coating material, and cosmetics, because of its low toxicity. However, recent studies have shown that nano-sized or ultrafine TiO(2) (UF-TiO(2)) (<100 nm in diameter) can generate pulmonary fibrosis and lung tumor in rats. Cytotoxicity induced by UF-TiO(2) in rat lung alveolar macrophages was also observed. This generates great concern about the possible adverse effects of UF-TiO(2) for humans. The cytotoxicity and genotoxicity of UF-TiO(2) were investigated using the methyl tetrazolium cytotoxicity (MTT) assay, the population growth assay, the apoptosis assay by flow cytometry, the cytokinesis block micronucleus (CBMN) assay, the comet assay, and the hypoxanthine-guanine phosphoribosyltransferase (HPRT) gene mutation assay. WIL2-NS cells were incubated for 6, 24 and 48 h with 0, 26, 65 and 130 microg/ml UF-TiO(2). Significant decreases in viability were seen in the MTT assay at higher doses; for example, 61, 7 and 2% relative viability at 130 microg/ml for 6, 24 and 48-h exposure (P<0.01). A dose-dependent relationship was observed, while a time-dependent relationship was seen only at the highest dose (130 microg/ml) after exposure for 24 and 48 h. Treatment with 130 microg/ml UF-TiO(2) induced approximately 2.5-fold increases in the frequency of micronucleated binucleated cells (P<0.01). In addition, a significant reduction in the cytokinesis block proliferation index was observed by the CBMN assay (P<0.05). In the comet assay, treatment with 65 microg/ml UF-TiO(2) induced approximately 5-fold increases in olive tail moment (P<0.05). In the HPRT mutation assay, treatment with 130 microg/ml UF-TiO(2) induced approximately 2.5-fold increases in the mutation frequency (P<0.05). The results of this study indicate that UF-TiO(2) can cause genotoxicity and cytotoxicity in cultured human cells.     

The detection and assessment of the aneugenic potential of selected oestrogens, progestins and androgens using the in vitro cytokinesis blocked micronucleus assay

The use of 17-beta-oestradiol, testosterone, progesterone, zearanol, trenbolone acetate and melengesterol acetate in animal feed as growth promoters has been banned in the European Union since 1989. However, the data available on their genotoxicity is limited. To bridge this gap the present study was carried out with the aim of evaluating these hormones for their ability to induce aneuploidy. Aneuploidy has been recently considered sufficiently important to be included in the routine testing of chemicals and radiation. These types of numerical chromosomal aberrations may arise by at least two mechanisms, chromosome loss and non-disjunction. Over the past few years, the cytokinesis blocked micronucleus (CBMN) technique has evolved into a robust assay for the detection of aneuploidy induction. At the present time, it is the only assay which can reliably detect both chromosome loss and non-disjunction when the basic methodology is coupled with appropriate molecular probing techniques such as immunoflourescent labelling of kinetochores and Fluorescence in situ Hybridisation. In this present study, aneuploidy induction by three groups of hormones was studied using CBMN assay coupled with Fluorescence in situ Hybridisation. The results from the present study demonstrate that 17-beta-oestradiol, diethylstilboestrol, progesterone and testosterone are genotoxic and induce aneuploidy by non-disjunctional mechanism, whereas trenbolone is also genotoxic by a clastogenic mechanism. However, melengesterol acetate and zearanol proved to be non-genotoxic in vitro.     

Chromosomal changes: induction, detection methods and applicability in human biomonitoring

The objective of this state of the art paper is to review the mechanisms of induction, the fate, the methodology, the sensitivity/specificity and predictivity of two major cytogenetic endpoints applied for genotoxicity studies and biomonitoring purposes: chromosome aberrations and micronuclei. Chromosomal aberrations (CAs) are changes in normal chromosome structure or number that can occur spontaneously or as a result of chemical/radiation treatment. Structural CAs in peripheral blood lymphocytes (PBLs), as assessed by the chromosome aberration (CA) assay, have been used for over 30 years in occupational and environmental settings as a biomarker of early effects of genotoxic carcinogens. A high frequency of structural CAs in lymphocytes (reporter tissue) is predictive of increased cancer risk, irrespective of the cause of the initial CA increase. Micronuclei (MN) are small, extranuclear bodies that arise in dividing cells from acentric chromosome/chromatid fragments or whole chromosomes/chromatids that lag behind in anaphase and are not included in the daughter nuclei in telophase. The cytokinesis-block micronucleus (CBMN) assay is the most extensively used method for measuring MN in human lymphocytes, and can be considered as a "cytome" assay covering cell proliferation, cell death and chromosomal changes. The key advantages of the CBMN assay lie in its ability to detect both clastogenic and aneugenic events and to identify cells which divided once in culture. Evaluation of the mechanistic origin of individual MN by centromere and kinetochore identification contributes to the high sensitivity of the method. A number of findings support the hypothesis of a predictive association between the frequency of MN in cytokinesis-blocked lymphocytes and cancer development. Recent advances in fluorescence in situ hybridization (FISH) and microarray technologies are modifying the nature of cytogenetics, allowing chromosome and gene identification on metaphase as well as in interphase. Automated scoring by flow cytometry and/or image analysis will enhance their applicability.     

In vitro short-term test evaluation of catecholestrogens genotoxicity

Estrogens are indicated as being the most important etiological factors for the development and progression of breast cancer. The implication of estrogen in breast cancer has been associated mostly with the estrogen receptors that mediate cell proliferation. Evidence also exists to support the hypothesis of a direct role of estrogens as tumor initiators. However, the role of estrogen genotoxicity in breast cancer is still questionable.

In this study the genotoxic activity of catecholestrogens and 16-hydroxy estrone has been investigated by performing Salmonella strain TA98 and TA100 Ames tests, sister chromatide exchange assays (SCE) and micronucleus assays on human peripheral lymphocytes (CBMN and ARA/CBMN). We found a lack of positive results with micronucleus assays, except for 2-hydroxy estradiol (2-OHE₂), which shows a peculiar “bell shaped” trend of micronucleus number versus concentrations. SCE assay suggests weak genotoxic activity of all tested catechol metabolites, except 4-hydroxy estrone (4-OHE₁), which also showed negative results by ARA/CBMN.

In this open debate, our results support the hypothesis of a weak genotoxicity, not correlated with the carcinogenetic potential of estrogens.     

Juglone, a naphthoquinone from walnut, exerts cytotoxic and genotoxic effects against cultured melanoma tumor cells

This study demonstrates cytotoxic and genotoxic potential of juglone, a chief constituent of walnut, and its underlying mechanisms against melanoma cells. MTT assay and clonogenic assay were used to study cytotoxicity, micronucleus assay to assess genotoxicity, glutathione (GSH) assay and 2′,7′-dicholorofluorescein diacetate (DCFH-DA) assay to evaluate the oxidative stress induction. Apoptosis/necrosis induction was analysed by flow cytometry. We observed a concentration-dependent decrease in cell survival with a corresponding increase in the lactate dehydrogenase levels. A dose-dependent increase in the frequency of micronucleated binucleate cells indicated the potential of juglone to induce cytogenetic damage in melanoma tumor cells. Moreover, results of the micronuclei study indicated division delay in the proliferating cell population by showing decrease in the cytokinesis blocked proliferation index. Further, juglone-induced apoptosis and necrosis could be demonstrated by oligonucleosomal ladder formation, microscopic analysis, increase in the hypodiploid fraction (sub Go peak in DNA histogram), as well as an increased percentage of AnnexinV(+)/PI(+) cells detected by flow cytometry. A significant concentration-dependent decrease in the glutathione levels and increase in dichlorofluorescein (DCF) fluorescence after juglone treatment confirmed the ability of juglone to generate intracellular reactive oxygen species. The cytotoxic effect of juglone can be attributed to mechanisms including the induction of oxidative stress, cell membrane damage, and a clastogenic action leading to cell death by both apoptosis and necrosis.     

Genotoxicity testing of PLGA-PEO nanoparticles in TK6 cells by the comet assay and the cytokinesis-block micronucleus assay

The in vitro genotoxicity of PLGA-PEO (poly-lactic-co-glycolic acid-polyethylene oxide copolymer) nanoparticles was assessed in TK6 cells using the comet assay as well as cytokinesis-block micronucleus (CBMN) assay. The cells were exposed to 0.12-75μg/cm2 of PLGA-PEO nanoparticles during 2 and 24h for analysis in the comet assay, and to 3-75μg/cm2 of these nanoparticles during 4, 24, 48 and 72h, respectively, for analysis in the CBMN assay. Two different protocols for treatment with cytochalasin B were used. We found that PLGA-PEO was neither cytotoxic (measured by relative cell growth activity and cytokinesis-block proliferation index (CBPI)), nor did it induce DNA strand-breaks (detected by the comet assay) or oxidative DNA lesions (measured by the comet assay modified with lesion-specific enzyme formamidopyrimidine-DNA-glycosylase). There were no statistically significant differences in the frequencies of micronucleated binucleated cells (MNBNCs) between untreated and treated cells in either of the conditions used. This suggests that PLGA-PEO did not have potential genotoxicity. However, using two experimental protocols of the micronucleus assay, PLGA-PEO nanoparticles showed a weak but significant increase in the level of MN in mononucleated cells, in cells treated for 48h with PLGA-PEO nanoparticles when cytochalasin B was added for the last 24h (1st protocol), and in cells treated for 24h with PLGA-PEO nanoparticles followed by washing of NPs and addition of cytochalasin B for another 24h (2nd protocol). It remains unclear whether the increase of MNMNC after treatment with PLGA-PEO nanoparticles is the effect of a possible, weak aneugenic potential or early effect of these particles, or due to another reason. These results suggest that aneugenicity in addition to clastogenicity may be considered as an important biomarker when assessing the genotoxic potential of polymeric nanoparticles.     

Radiation-induced genomic instability in Caenorhabditis elegans

Currently, the cosmetics industry relies on the results of in vitro genotoxicity tests to assess the safety of chemicals. Although the cytokinesis-block micronucleus (CBMN) test for the detection of cells that have divided once is routinely used and currently accepted by regulatory agencies, it has some limitations. Reconstituted human epidermis (RHE) is widely used in safety assessments because its physiological properties resemble those of the skin, and because it allows testing of substances such as hydrophobic compounds. Thus, the micronucleus test is being adapted for application in RHE-reconstructed tissues. Here we investigated whether two different reconstructed epidermis models (EPI/001 from Straticell, and RHE/S/17 from Skinethic) are suitable for application of the micronucleus test. We found that acetone does not modify micronucleus frequency, cell viability, and model structure, compared with non-treated RHE. Treatment of the EPI/001 model with mitomycin C and vinblastine resulted in a dose-dependent increase of micronucleus frequency as well as a decrease of tissue viability and of binucleated cell rate, while no changes of the epidermal structure were observed. The number of binucleated cells obtained with the RHE/S/17 model was too small to permit micronucleus testing. These results indicate that the proliferative rate of the tissue used is a critical parameter in performing the micronucleus test on a 3D model.     

Induction of Excision Repairable DNA Lesions in Lymphocytes Exposed to Lead and ALA In Vitro

Numbers of studies have been carried out on the potential of lead genotoxicity. The mechanisms of lead genotoxicity are not fully known but partly attributed to the formation of highly reactive oxygen metabolites (ROM). However, lead ions have no ability to generate ROM. In this study, we have investigated the ability of lead and ALA to induce excision repairable DNA lesions by using cytosine arabinoside or cytokinesis block micronucleus (ARA-C/CBMN) assay. N-methyl-N-nitrosourea was used as a positive control which is a mutagen and known to induce excision repair. The results of the ARA-C/CBMN assay show that ALA exposures have significantly (p < 0.01) increased the ratio of excision repairable DNA lesions in peripheral blood lymphocytes; however, lead have not. Accordingly, accumulation of ALA should be considered as an effective partner of lead induced DNA damage in lead exposure.     

A delay in the Saccharomyces cerevisiae cell cycle that is induced by a dicentric chromosome and dependent upon mitotic checkpoints.

Dicentric chromosomes are genetically unstable and depress the rate of cell division in Saccharomyces cerevisiae. We have characterized the effects of a conditionally dicentric chromosome on the cell division cycle by using microscopy, flow cytometry, and an assay for histone H1 kinase activity. Activating the dicentric chromosome induced a delay in the cell cycle after DNA replication and before anaphase. The delay occurred in the absence of RAD9, a gene required to arrest cell division in response to DNA damage. The rate of dicentric chromosome loss, however, was elevated in the rad9 mutant. A mutation in BUB2, a gene required for arrest of cell division in response to loss of microtubule function, diminished the delay. Both RAD9 and BUB2 appear to be involved in the cellular response to a dicentric chromosome, since the conditionally dicentric rad9 bub2 double mutant was highly inviable. We conclude that a dicentric chromosome results in chromosome breakage and spindle aberrations prior to nuclear division that normally activate mitotic checkpoints, thereby delaying the onset of anaphase.     

Micronuclei and other nuclear anomalies in buccal smears: methods development

Laboratory work aimed at improving the epidemiologic utility of an innovative genotoxicity assay is described. The exfoliated cell micronucleus assay involves microscopic analysis of epithelial smears to determine the prevalence of micronucleation, an indicator of structural or numerical chromosome aberrations. While the assay holds promise for the study of epithelial carcinogens, it is hampered by the fact that exfoliated cells are moribund and undergo degenerative phenomena that can produce extranuclear objects difficult to distinguish from classical micronuclei. Modifications in the protocol were assessed in sample buccal smears from several study populations: radiotherapy patients, nonusers of tobacco, and snuff users. Refinements in micronucleus scoring criteria and the inclusion of other nuclear anomalies in the scoring system are proposed. We demonstrate that our criteria are successful in detecting excess micronucleation in positive controls. We also provide evidence that other nuclear anomalies are at least as common as micronucleation and that therefore there is the potential for extensive misclassification. Reliability was assessed in duplicate readings.     

Micronuclei and other nuclear anomalies in buccal smears: methods development.

Laboratory work aimed at improving the epidemiologic utility of an innovative genotoxicity assay is described. The exfoliated cell micronucleus assay involves microscopic analysis of epithelial smears to determine the prevalence of micronucleation, an indicator of structural or numerical chromosome aberrations. While the assay holds promise for the study of epithelial carcinogens, it is hampered by the fact that exfoliated cells are moribund and undergo degenerative phenomena that can produce extranuclear objects difficult to distinguish from classical micronuclei. Modifications in the protocol were assessed in sample buccal smears from several study populations: radiotherapy patients, nonusers of tobacco, and snuff users. Refinements in micronucleus scoring criteria and the inclusion of other nuclear anomalies in the scoring system are proposed. We demonstrate that our criteria are successful in detecting excess micronucleation in positive controls. We also provide evidence that other nuclear anomalies are at least as common as micronucleation and that therefore there is the potential for extensive misclassification. Reliability was assessed in duplicate readings.