- Open Access
Weight of evidence approach using a TK gene mutation assay with human TK6 cells for follow-up of positive results in Ames tests: a collaborative study by MMS/JEMS
Genes and Environment volume 43, Article number: 7 (2021)
Conflicting results between bacterial mutagenicity tests (the Ames test) and mammalian carcinogenicity tests might be due to species differences in metabolism, genome structure, and DNA repair systems. Mutagenicity assays using human cells are thought to be an advantage as follow-up studies for positive results in Ames tests. In this collaborative study, a thymidine kinase gene mutation study (TK6 assay) using human lymphoblastoid TK6 cells, established in OECD TG490, was used to examine 10 chemicals that have conflicting results in mutagenicity studies (a positive Ames test and a negative result in rodent carcinogenicity studies).
Two of 10 test substances were negative in the overall judgment (20% effective as a follow-up test). Three of these eight positive substances were negative after the short-term treatment and positive after the 24 h treatment, despite identical treatment conditions without S9. A toxicoproteomic analysis of TK6 cells treated with 4-nitroanthranilic acid was thus used to aid the interpretation of the test results. This analysis using differentially expressed proteins after the 24 h treatment indicated that in vitro specific oxidative stress is involved in false positive response in the TK6 assay.
The usefulness of the TK6 assay, by current methods that have not been combined with new technologies such as proteomics, was found to be limited as a follow-up test, although it still may help to reduce some false positive results (20%) in Ames tests. Thus, the combination analysis with toxicoproteomics may be useful for interpreting false positive results raised by 24 h specific reactions in the assay, resulting in the more reduction (> 20%) of false positives in Ames test.
DNA reactive substances may directly damage DNA even when present at low levels leading to mutations and, therefore, potentially initiating cancer. A positive result in the Ames test has a significant influence on the development of new drugs and chemical substances, and in many cases, mutagenic chemicals are dropped from drug/chemical development at an early stage. In the guidelines from the ICH M7 (ICH: International Council for Harmonization of Technical Requirements for Pharmaceuticals for Human Use), Ames-positive substances are considered as DNA reactive substances  and require a great deal of labor for subsequent development and manufacturing. Continued development of components with Ames-positive results requires in vivo tests, such as transgenic rodent mutation assays as follow-up testings. However, problems such as costs and labor burden of in vivo testing can be prohibitive.
Furthermore, such testing does not fit the 3R principles for animal welfare [2, 3]. Positive results in the Ames test correlate well with carcinogenicity in rodents, with the concordance (Ames-negative and carcinogenic-negative) of approximately 80% [4, 5]. This indicates that there are approximately 20% of chemical substances that are positive in Ames tests and negative for carcinogenicity test. In fact, it is considered that some of these substances are unrelated to human carcinogenesis. For example, fexinidazole, a drug for sleeping sickness, is positive for the Ames test but negative for all in vitro (micronucleus test in human lymphocytes) and in vivo genotoxicity tests (ex vivo unscheduled DNA synthesis in rats; bone marrow micronucleus test in mice). Tweats et al.  demonstrated that fexinidazole is metabolically activated by a bacterial-specific nitroreductase reaction and is thus positive in the Ames test alone. Thus, if it is possible to prove that the Ames test positive is a bacteria-specific reaction and has a low risk of carcinogenicity in humans, useless a follow-up testing in whole animals can be avoided.
The consideration of the mode of action is the critical establishment of non-animal testing for the safety evaluation of chemicals. The Organization for Economic Co-operation and Development (OECD) have been vigorously developing “adverse outcome pathway (AOP)” and “integrated approaches to testing and assessment (IATA)” combining in silico and in vitro information based on AOP . The AOP and IATA would contribute to a precise toxicological evaluation based on the weight of evidence (WoE) including genotoxicity [7, 8] and the derivation of regulatory conclusions. Since the in vivo testing is prohibited for the safety evaluation of cosmetic ingredients, the use of a WoE approach based on in vitro testing is relatively advanced in the cosmetic industry . An example is the toxicity evaluation of Basic Brown 17, used as a hair coloring agent for cosmetics. Basic Brown 17 is positive in Ames test, but negative in mutation assays using mouse lymphoma (two loci of TK and HPRT) and in in vitro micronucleus test using mammalian cells. Furthermore, Basic Brown 17 is negative in the comet assay with 3-D reconstituted human skin cells [10, 11]. Thus, based on WoE , Basic Brown 17 can be considered to show negligible potential for in vivo genotoxicity, and no additional testing is reported . Furthermore, omics technologies, such as transcriptome and proteome, play a crucial role in implementing a WoE approach. Ates et al.  used an in vitro transcriptomics approach to evaluate the genotoxicity predictions for cosmetic ingredients that were negative by in silico analysis but positive by the Ames test. Transcriptomics data were reported to provide needed mechanistic information for toxicity assessments at the gene expression and metabolic pathway levels. In addition, Kirkland et al. [14, 15] also reported that follow-up in vitro tests, such as for gene expression profiling, may aid significantly in interpretation of the relevance for humans of the in vitro genotoxicity results vis-à-vis in vivo genotoxicity or carcinogenicity. Thus, the WoE approaches are also useful for a follow-up strategy for positive Ames tests, shifting steadily toward new evaluation strategies for human risk while reducing dependence on animal experiments.
Based on the current literature, we focused on the utility of the thymidine kinase gene (TK) mutation assay using human TK6 cells for the follow-up of positive Ames tests. Conflicting results between Ames tests and mammalian carcinogenicity tests might be caused by species differences in metabolism, genome structure, and DNA repair systems. TK6 cells express human metabolic enzymes and have a chromatin structure. Furthermore, the cells are positive for p53 protein-related functions (competent DNA repair systems) . TK6 cells might be useful for in a WoE approach to scrutinize positive results in the Ames test. However, little is known about the utility of the assay for this purpose. Thus, the present study used the TK6 assay for the follow-up of Ames test results with 10 non-carcinogenic chemicals that were Ames-positive in a collaborative study with 10 laboratories in Japan. Assays were conducted for a short-term (4 h in the presence and absence of rat liver S9) and long-term (24 h in the absence of S9) treatments. Furthermore, we explored the integration of toxicoproteomics analysis with TK6 assays to help interpret test results and increase the utility of WoE.
Materials and methods
Participants and test substances
Ten laboratories in Japan, including pharmaceutical and chemical companies, contract laboratories, and public institutes, conducted TK6 assays for 10 test substances in the collaborative study (Table 1). We purchased the test substances from FUJIFILM Wako Pure Chemical Corp. (Osaka, Japan), Tokyo Chemical Industry Co., LTD. (Tokyo, Japan), or Sigma-Aldrich (St. Louis, USA). Most of the test substances were dissolved in dimethyl sulfoxide (DMSO) as a solvent, except that ethanol was used for 1-nitronaphthalene (Table 1).
The human lymphoblast cell line TK6 was purchased from the Japanese Collection of Research Bioresources cell bank and the American Type Culture Collection. Cells were cultured at 37 °C with 5% CO2 in an RPMI medium containing 10% horse serum (JRH Bioscience), 200 μg/mL sodium pyruvate (Wako Pure Chemical Industries, Ltd. and Thermo Fisher (Gibco)), 100 U/mL penicillin, and 100 μg/mL streptomycin (Nacalai Tesque Inc. and Thermo Fisher (Gibco)). Reagent manufacturers are not necessarily unified, and some research institutions have procured equivalent reagents from other manufacturers.
The treatment of test substances and dose finding
In principle, this collaborative study conducted the treatment of test substances, the dose finding, and the main test of TK6 assays, according to the OECD Guideline TG490 . We optimized the protocol for TK6 assay prior to this collaborative study. The laboratory of the National Institute of Health Sciences carefully supported the experimental procedures such as chemical treatments and colony counting for TK6 assay performed by each laboratory. Briefly, test substances were dissolved in an appropriate solvent (DMSO or ethanol) and then serially diluted to prepare final concentrations. Details of 10 test substances are shown in Table 2. In the absence (150 mM KCl) or presence of rat liver S9 mix (final concentration 4.5%), cells (2 × 107 cells) were exposed to a test substance and cultured for 4 h. Rat liver S9 was purchased from Oriental Yeast Co., Ltd. and IEDA TRADING Corp. In cases where the short-term treatment (4 h) showed negative results, the 24 h treatment without S9 mix was conducted. Cyclophosphamide (CP, Fujifilm Wako Pure Chemical Industries, Ltd.) was used as a positive control (2.5–3 μg/mL) for metabolic activation, and methyl methanesulfonate (MMS, Tokyo Chemical Industry Co., Ltd., and Sigma Aldrich) was used as a positive control (3–5 μg/mL for 4 h and 1–2.5 μg/mL for 24 h treatments) for non-metabolic activation. The 4 h and 24 h treatments were conducted by shaking and static exposures, respectively. After treatment with test substances, cells were centrifuged (1000 rpm, 5 min), supernatants were removed, and cells were washed with a serum-free medium or Hank’s Balanced Salt Solution. After centrifugation (1000 rpm, 5 min) to remove the supernatant, cells (2 × 107 cells) were dispersed in an RPMI medium containing 10% serum, and cell concentration was measured. Treated cells were cultured at 37 °C with 5% CO2 condition and used for TK6 assays. At that time, to calculate the cloning efficiency (CE) (Eq. 1), the cells were cultured in a 96-well microplate at a concentration of about 1.6 cells/well for 2 weeks.
CE, the cell colony formation rate, was calculated using Eq. 1 according to the Poisson distribution equation. EW was the number of wells without colonies, and TW was the total number of wells. N was the average number of cells per well (N = 1.6).
Additionally, CE was adjusted by the following calculation (Eq. 2) due to cell loss when exposure caused cytotoxicity. “The number of cells at the end of the treatment” was the number of cells obtained after centrifugation at the end of the treatment in the above-described exposure treatment of the test substance. The “cell number at the start of treatment” was 2 × 107.
CE0 seeding, in which cells were seeded immediately after treatment, and CE3 seeding, in which cells were seeded 3 days after treatment, to examine cell viability in the TK6 assay described below. The relative cell viability RS0 (%) just after the treatment with the test substance was calculated from CE0, following Eq. 3. The survival rate of the negative control group was defined as 100%.
The main test of the TK6 assay
The TK6 assay is conducted between 20 and 10% RS0 as a maximum dose if cytotoxicity is observed. If no precipitate or limiting cytotoxicity was observed, the maximum dose was the lowest concentration among 10 mM, 2 mg/mL, or 2 μL/mL of the test substance. When the main TK6 assay was difficult to conduct under the condition of RS0 values from 20 to 10%, the test was conducted between 20 and 10% relative total growth (RTG), described later, as a maximum dose.
Besides RS0, relative suspension growth (RSG) and RTG as other cytotoxicity indices (Eq. 4) were used. When cells were cultured for 3 days after treatment with the test substance, suspension cell growth ratio 1 (SG1) was the growth ratio from day 0 to day 1 (cell concentration on day 1/cell concentration on day 0). The suspension cell growth ratio 2 (SG2) was the growth ratio from day 1 to day 2 (cell concentration on day 2/cell concentration on day 1). The RSG value was calculated by dividing the total SG value of the untreated group for 3 days by the total SG value of the treated group for the same 3 days (Eq. 4).
On the third day of culture, CE3 plates for determining plating efficiency and mutant frequency (MF) plates for mutation detection were prepared. CE3 plates were seeded in 96-well microplates at a concentration of about 1.6 cells/well. The MF plates were seeded in 96-well microplates at 40,000 cells/well in the presence of 3 μg/mL of trifluorothymidine (Sigma Aldrich).
Mutant colonies in MF plates were calculated using Eq. 7 using the Poisson distribution. EW was the number of wells without colonies, and TW was the total number of wells. N was the average number of cells per well (N = 40,000 in this study). The data were statistically analyzed by Omori’s method, a modified Dunnett’s procedure for identifying clear negatives, a Simpson–Margolin procedure for detecting downturn data, and a trend test to evaluate the dose dependency [42, 43]. The acceptability criterion for the test was that the MF value of the positive control group of each laboratory was increased with statistical significance compared with that of the concurrent negative control group.
Sample preparation for proteomic analysis
We conducted a proteomic analysis using TK6 cells treated by 4-nitroanthranilic acid that was negative in the short-term treatment and positive in the 24 h treatment. HC Blue No. 2 gave also same test results. 4-Nitroanthranilic acid, but not HC Blue No. 2, increased clearly the mutant efficiency in dose-dependent manner even at the low-dose range in the 24 h treatment (Additional file 1 (Table S1 (No.8)), we selected the chemical for the proteomics. Cell treatment with 4-nitroanthranilic acid was the same as the short-time treatment (4 h, −S9mix) and long-term treatment (24 h, −S9mix) in the TK6 assay. Treatment concentrations were 0, 400, and 800 μg/mL, relatively low toxicity doses, to clearly measure biological responses to the chemical. After treatment, the supernatant was removed by centrifugation (1000 rpm, 5 min). The cell pellet (6 × 106 cells) was washed with a 10 mL of ice-cold phosphate-buffered saline (PBS) and centrifuged (repeated twice on ice throughout). The cells were resuspended in 3 mL of PBS containing the protease inhibitor cocktail (Roche), complete EDTA-free (Roche), and 2 × 106 cells each were dispensed into three pre-cooled 2.0 mL tubes. After centrifugation, supernatants were removed again, and tubes containing cell pellets were immersed in liquid nitrogen to freeze the cell pellet and stored at − 80 °C.
Each frozen cell pellet was mixed with PTS (phase transfer surfactant) buffer and boiled at 95 °C for 5 min . Lysates were further sonicated three times (15 min per cycle) with a Bioruptor sonicator (Cosmo Bio, Tokyo, Japan). Samples were then reduced with 10 mM TCEP (tris(2-carboxyethyl) phosphine), alkylated with 20 mM iodoacetamide, and sequentially, digested with trypsin (protein weight: 1/50) and Lys-C (protein weight: 1/50) for 16 h at 37 °C. Peptides were elucidated by centrifugation for 10 min at 20,000 g and desalted on a C18-SCX StageTips .
LC-MS/MS was conducted by coupling an UltiMate 3000 Nano LC system (Thermo Scientific, Bremen, Germany) and an HTC-PAL autosampler (CTC Analytics, Zwingen, Switzerland) to a Q Exactive hybrid quadrupole-Orbitrap mass spectrometer (Thermo Scientific). Peptides were loaded on an analytical column (75 μm × 30 cm, packed in-house with ReproSil-Pur C18-AQ, 1.9 μm resin, Dr. Maisch, Ammerbuch, Germany) and separated at a flow rate of 280 nL/min using a 45 min gradient from 5 to 35% of solvent B (solvent A, 0.1% FA and 2% acetonitrile; solvent B, 0.1% FA and 90% acetonitrile). The Q Exactive instrument was operated in the data-dependent mode. Survey full-scan MS spectra (m/z 350–1800) were acquired in the Orbitrap with 70,000 resolution after the accumulation of ions to a 3 × 106 target value. Dynamic exclusion was set to 10 s. The 12 most intense multiplied charged ions (z ≥ 2) were sequentially accumulated to a 1 × 105 target value and fragmented in the collision cell by higher-energy collisional dissociation (HCD) with a maximum injection time of 120 ms and 35,000 resolution. Typical mass spectrometric conditions were: spray voltage, 2 kV; heated capillary temperature, 250 °C; and normalized HCD collision energy, 25%. The MS/MS ion selection threshold was set to 2.5 × 104 counts. A 2.0 Da isolation width was chosen.
Raw MS data were processed by MaxQuant (version 18.104.22.168) supported by the Andromeda search engine for peak detection and quantification. MS/MS spectra were searched against the UniProt human database with the following search parameters: full tryptic specificity, up to two missed cleavage sites, carbamidomethylation of cysteine residues set as a fixed modification, and N-terminal protein acetylation and methionine oxidation as variable modifications. Search results were filtered to a maximum FDR (false discovery rate) of 0.01 at protein, and PSM levels.
Extraction of differentially expressed proteins
All NaN values of log2 LFQ intensities were converted to − 7, which is approximately equivalent to the minimum value of the LFQ intensities observed in the present study. Then, fold change values were calculated by dividing LFQ intensities (antilogarithms of the log2 LFQ intensities) into sample treated groups by intensities in non-treated (control) groups, respectively (4 and 24 h treated groups). Then, log2 fold changes were calculated. Finally, differentially expressed proteins (DEPs) that were notably expressed in each sample treated group were identified based on log2 fold change of ≥2. Next, to analyze specific effects in repeated treatments at the highest dose, DEPs in 24 h 800 μg/mL group was subjected to further screening based on log2 fold change between 24-h 800 μg/mL group and other groups ≥1. Thereby, the specific DEPs in 24 h 800 μg/mL group were thus identified. Log2 fold changes in LFQ intensities were used to generate a heatmap using the R package (heatmap3).
Enrichment analysis was used to interpret the biological processes and molecular functions of DEPs. DAVID bioinformatics v6.8 (https://david.ncifcrf.gov/) was used for enrichment analyses to annotate DEPs to their correlated GO (Gene Ontology) terms and pathways. Specifically, Gene Ontology (GOTERM_BP_DIRECT, GOTERM_CC_DIRECT, GOTERM_MF_DIRECT), and Protein_Domains (INTERPRO, PIR_SUPERFMILY, SMART) were analyzed. Moreover, since DNA damage and oxidative stress were considered as important mechanisms to interpret genotoxicity test results, enrichment scores (the Expression Analysis Systematic Explorer (EASE) p-value) of some related terms, such as GO:0006974 (cellular response to DNA damage stimulus) and GO:0006979 (response to oxidative stress) were respectively evaluated even if they were not statistically significant. Statistical significance was expressed as −log (the EASE p-value) and was illustrated as bar charts for comparison.
Negative and positive control data
The TK6 assay conducted by each laboratory were presented in Fig. 1 (No.1–10) and Table 3. The mutant frequencies of negative control were 2.5 to 16.9 × 10− 6 (mean 6.36 × 10− 6) for the short treatment without S9 mix, 1.7 to 16.8 × 10− 6 (mean 6.78 × 10− 6) for the short treatment with S9 mix, and 2.3 to 15.5 × 10− 6 (mean 9.09 × 10− 6) for the 24 h treatment without S9 mix. According to the laboratory of the National Institute of Health Sciences, the historical spontaneous mutant frequency is 4 to 10 × 10− 6, these mean values almost accepted the criteria by OECD TG490 . In addition, the mutant frequencies of the concurrent positive controls were 9.4 to 71.7 × 10− 6 (mean 30.1 × 10− 6) for MMS during the short treatment without S9 mix, 10.4 to 57.8 × 10− 6 (mean 28.6 × 10− 6) for CP during the short treatment with S9 mix, and 18.2 to 120 × 10− 6 (mean 61.6 × 10− 6) for MMS during the 24 h treatment without S9 mix. Thus, the concurrent positive controls produced a statistically significant increase compared with the concurrent negative control. For more details, see Additional file 1 (Table S1 (No.1–10)) attached in this paper. The summary of statistics analysis by Omori’s method in each test substance was provided in Additional file 2 (Table S2).
4-(Chloroacetyl) acetanilide was prepared by dissolving in DMSO. Dose levels of 4-(chloroacetyl) acetanilide were set at 0.5, 1, 1.5 and 2 μg/mL in the short-term treatment (−S9), 5, 10, 14, 16, and 17 μg/mL with the short-term treatment (+S9), and 0.5, 1.0, 1.2, and 1.4 μg/mL with the long-term treatment based on the results of the dose range-finding tests. A positive response was statistically significant in the Dunnett type test in the short-term (−S9) and long-term treatments. However, no statistically significant linear trends were observed (Additional file 2 (Table S2)). Both MF values clearly exceeded those of the solvent control group. Therefore, based on expert judgment, 4-(chloroacetyl) acetanilide was judged to be positive.
2-(Chloromethyl) pyridine HCl
About 80% cytotoxicity (relative colony efficiency) was observed at 60 μg/mL in the absence or at 100 μg/mL in the presence of S9. Regardless of metabolic activation, a clear concentration-dependent increase of MF was observed under the condition used in this study. Long-term treatment was not applied. The test chemical, 2-(chloromethyl) pyridine HCl, was judged as positive.
More than 80% cytotoxicity was observed at 2000 μg/mL in the long-term treatment. Statistically significant increases of MF were observed after the short-term treatment with metabolic activation and after the long-term treatment. Thus, 2,6-diaminotoluene was judged to be positive.
About 80% of cytotoxicity was observed at 20.0 μg/mL in the absence or at 200 μg/mL in the presence of S9 for the short-term treatment and 10.0 μg/mL for the long-term treatment. Statistically significant increases of mutation frequency were not observed after the short-term treatment in the absence or presence of metabolic activation and after the long-term treatment. Thus, 2,5-diaminotoluene was judged to be negative.
HC Blue No. 2
Based on the results of the dose finding study, concentrations were determined. In the 4 h treatment with and without S9, 62.5–2000 μg/mL were selected. In the 24 h treatment without S9, 60–240 μg/mL were selected. No precipitation was observed at any concentration in all groups. About 80% cytotoxicity was observed at 220 μg/mL in the 24 h treatment without S9 and about 40% cytotoxicity was observed at 2000 μg/mL in 4 h treatment with S9. No cytotoxicity was observed in 4 h treatment without S9. There was no significant difference in MF after 4 h treatment with and without S9. After the 24 h treatment without S9, MF increased in a concentration-dependent manner as compared with that of the solvent control group and showed a statistically significant difference. HC Blue No. 2 was judged to have gene mutagenicity.
After the short-term treatment (+S9), 8-Hydroxyquinoline showed cytotoxicity at the dose level of 10.0 μg/mL, at which the RS was 19.3%. Mutagenic responses to 8-hydroxyquinoline were significant in both Dunnett type test and linear trend tests with S9. After the short-term treatment (−S9), 8-hydroxyquinoline showed cytotoxicity at a dose level of 100 μg/mL, at which the RS was 18.9%. Mutagenic responses to 8-hydroxyquinoline were not significant in the Dunnett type test in without S9. Therefore, based on the above results, it was concluded that 8-hydroxyquinoline had mutagenic potential in TK6 cells under the conditions of this study. Long-term treatment was not applied.
After the 24 h treatment without S9, Iodoform showed cytotoxicity at the dose level of 40.0 μg/mL, at which the RS was 10.9%. Conversely, after 4 h treatment with/without S9, because of the strong toxicity during the expression period, mutation frequency at the concentration with RS of 10–20% could not be evaluated. Therefore, an additional study was conducted with treatment concentration at which RTG was 10–20% set as maximum concentration. Iodoform showed cytotoxicity at a dose level of 100 (−S9) and 200 (+S9) μg/mL, at which the RTGs were 19.1 and 20.1%, respectively. A dose-dependent increase in MF was observed after the long-term treatment at the high-dose levels (50–70 μg/mL), which showed RS < 10%. These dose levels were excluded from the statistical judgment because of high cytotoxicity. MF values from 10 to 40 μg/mL were significant in the Dunnett modified test (Additional file 2 (Table S2)) but were not increased dose-dependently within the dose range. Hence, we suggest that the significant result obtained from the Dunnett modified test was not biologically relevant in this study. Therefore, the mutagenic potential of iodoform was negative as expert judgment.
More than 50% cytotoxicity was observed at 1138 and 1821 μg/mL without metabolic activation, and approximately 100% cytotoxicity was observed at 1821 μg/mL with metabolic activation for the short-term treatment. More than 80% cytotoxicity was observed at 1821 μg/mL for the long-term treatment. There was no increase in mutation frequency after the short-term treatment with or without metabolic activation. A statistically significant increase in mutation frequency after the long-term treatment was observed.
Doses showing 80% cytotoxicity in the absence of S9 were between 120 and 125 μg/mL. Doses showing 80% cytotoxicity with S9 were between 17.5 and 20 μg/mL. Regardless of metabolic activation and exposure time, a significant increase in MF was observed under the condition used in this study. Long-term treatment was not applied. Hence, this test chemical, 1-nitronaphthalene, was judged as positive.
More than 80% cytotoxicity was observed at 144 μg/mL and above for the long-term treatment. MF was increased significantly in the statistical analysis compared with the negative control value for short-term (−S9) and long-term assays. The judgment for this chemical was positive. Cells exposed to 770 and 1540 μg/mL (+/−S9) were abandoned because of high cytotoxicity.
A total of 1078 DEPs (4 h 400 μg/mL group: 359 proteins, 4 h 800 μg/mL group: 506 proteins, 24 h 400 μg/mL group: 358 proteins, and 24 h 800 μg/mL group: 420 proteins) were identified based on log2 fold change of ≥2 (Fig. 2a). Furthermore, 168 specific DEPs in the 24 h 800 μg/mL group were extracted from 420 DEPs in 24 h 800 μg/mL group based on a log2 fold change between 24 h 800 μg/mL group and other groups ≥1 (Fig. 2b).
In the analysis of specific DEPs in 24 h 800 μg/mL group, the enrichment analysis of upregulated proteins revealed that the expression level of proteins that associated with transcription, cellular stress and cell division was remarkably upregulated (e.g., GO:0034244 ~ negative regulation of transcription elongation from RNA polymerase II promoter, GO:0008631 ~ intrinsic apoptotic signaling pathway in response to oxidative stress, GO:0051301 ~ cell division, and GO:2000145 ~ regulation of cell motility) (Table 4A). Furthermore, the analysis of downregulated proteins revealed proteins related to rRNA processing and cell division (e.g., GO:0000462 ~ maturation of SSU-rRNA from tricistronic rRNA transcript (SSU-rRNA, 5.8S rRNA, and LSU-rRNA) and GO:0051301 ~ cell division).
In the target analysis of specific GO term among groups (Fig. 3), P-values of GO:0006974 (cellular response to DNA damage stimulus), GO:0006979 (response to oxidative stress), and GO:0006281 (DNA repair) were less than 0.05, indicating no significant difference. Conversely, enrichment scores (−log (p-value)) of GO:0008631 ~ intrinsic apoptotic signaling pathway in response to oxidative stress were drastically increased in DEPs in 24 h 800 μg/mL group and specific DEPs in 24 h 800 μg/mL group, indicating involvement with oxidative stress.
In DEPs in 24 h 800 μg/mL group, three proteins ([Pyruvate dehydrogenase (acetyl-transferring)] kinase isozyme 1, mitochondrial phosphoinositide-dependent kinase 1: PDK1, superoxide dismutase 2, mitochondrial: SOD2, direct IAP-binding protein with low PI: DIABLO) that belong to GO:0008631 ~ intrinsic apoptotic signaling pathway in response to oxidative stress were included (Figs. 3 and 4). The expression level of these proteins is elevated only after the 24 h treatment. Furthermore, the expression level of SOD2 was elevated dose-dependently, and PDK1 and DIABLO were only elevated at 800 μg/mL.
Usefulness of the TK6 assay as a follow-up for testing Ames-positive compounds
The 10 test substances used in this collaborative study were mostly polycyclic aromatic compounds containing amino groups or nitro groups. Characteristics of chemical structures were examined for any regularity or correlation with the results of TK6 assays. The assay can certainly be positive for 8-hydroxyquinoline, 1-nitronaphthalene, and 4-nitro-o-phenylenediamine with specific activity values exceeding 10,000 revertants per milligram obtained from Ames test, even three compounds (4-(chloroacetyl)-acetanilide, 2-(chloromethyl) pyridine HCl, and HC Blue No. 2) with small specific activity values of several hundred revertants per milligram were detected as positive (Tables 2 and 3). Thus, the results of TK6 assays are not related to the strength of the specific activity value of the Ames test. Based on our results obtained in the present study, no regularity could be found between mutagenicity results and chemical structures of test substances.
The TK6 assay showed that two substances, 2,5-diaminotoluene and iodoform, were negative (20% effective as a follow-up test), and the remaining eight substances were positive (Table 3). Seven of these substances were reported by MLA using L5178Y cells (Table 3), and all seven showed positive results, including an “inconclusive” judgment for 2,5-diaminotoluene that was reported negative and positive by the Scientific Committee on Consumer Safety (SCCS)  and the National Toxicology Program (NTP) database , respectively. Compared to p53-proficient TK6 cells, p53-deficient L5178Y cells are more sensitive to spindle inhibitors due to disruption of spindle checkpoints and apoptosis . Moreover, Whitwell et al. reported that the use of human TK6 cells may be preferable over rodent L5178Y cells to help reduce false positive results in in vitro assay . We believe that the TK6 assay, due to negative here for 2,5-diaminotoluene in this study, may indicate a possible improvement over MLA to help follow-up on false-positive results from Ames testing.
In general, it is difficult to follow up the positive results of the Ames test, which detects DNA-reactive substances, with in vitro mammalian cell gene mutation test, as it may lead to similar results to the Ames test in terms of the principle of detecting mutations [4, 57]. In this present study, 20% effective (2 negative results of 10 substances in the TK6 assay) suggests that the TK6 assay may be able to follow up on the positive results of the Ames test. Six substances (4-(chloroacetyl)-acetanilide, 2-(chloromethyl) pyridine HCl, 2,6-diaminotoluene, 8-hydroxyquinoline, 1-nitronaphthalene, and 4-nitro-o-phenylenediamine) that are negative in in vivo genotoxicity tests (chromosomal Aberration test, micronucleus test, and transgenic rodent gene mutation assay) are not considered to be carcinogenic [48, 58,59,60,61,62,63], because their genotoxicity is eliminated by biological reactions (e.g., ADME; absorption, distribution, metabolism and excretion) even if they show DNA reactivity in vitro such as TK6 assay. Thus, Ames-positive substances, such as 2,5-diaminotoluene, that are negative in the TK6 assay, not via ADME, and negative in the in vivo test may have bacterial-specific DNA reactivity. Kirkland et al. [14, 15] reported that, in the case of an Ames-positive chemical, negative results in two in vitro mammalian cell tests covering both mutation and clastogenicity/aneugenicity endpoints should be considered as indicative of absence of in vivo genotoxic or carcinogenic potential. As the mutation endpoint, the human TK6 assay, even by current methods that have not been combined with new technologies such as proteomics, may have sufficient potential to reduce some false positive results (20%) as a follow-up test.
Possible mechanisms of chemical mutagenesis involving reactive oxygen species (ROS) production in the long-term treatment with TK6 cells
Interestingly, three (2,6-diaminotoluene, HC Blue No. 2, and 4-nitroanthranilic acid) of the eight positive substances in the present study were negative after the short-term treatment and positive after the 24 h treatment (Table 3), despite identical treatment conditions in the absence of S9. Thus, an extended treatment time of only 20 h long, the judgment of these substances was changed from negative to positive (Table 3). Based on the chemical structures of these three compounds, we discussed the mechanisms of positives results after the long-term treatment in the absence of S9.
Metabolism of hepatic microsomes in general (such as +S9) oxidize arylamines to hydroxylamines by P450 and O-acetylized to acetoxyarylamines with acetyltransferase. These further break down spontaneously into arylnitrenium ions that form adducts with bases in nucleic acids [64, 65]. Additionally, arylamines are metabolized to hydroxylamines with amine N-oxidase and flavin-containing monooxygenase (FMO) . Mutagenicity was confirmed in human TK6 cells after 96 h of exposure to benzidine, one of the arylamines, in the absence of S9 . Additionally, the mutagenicity of 2,6-diaminotoluene occurs in MLA using L5178Y mouse lymphoma cells (4 h, −S9) . A slight expression of P450 has been reported in TK6 cells . In the present study, the enzyme expression for P450s, amine N-oxidase, and FMO was not detected; however, the formation of proteins related to O-acetylization (N-acetyltransferase) was measured in the proteomic analysis of vehicle TK6 control (Additional file 3 (Table S3)). These findings suggest the possibility of DNA adduct formation by arylnitrenium ions via oxidation of arylamine with some enzymes in TK6 cells (24 h, −S9) . Furthermore, the induction of DNA damage mediated by ROS is highly plausible (Fig. 5) [65, 70,71,72].
2,4-Diaminotoluene and 2,6-diaminotoluene are considered carcinogenic and non-carcinogenic, respectively, based on a report of the National Toxicology Program (NTP), USA , although one study indicated that it is not at all clear that 2,6-diaminotoluene is non-carcinogenic . Significant DNA adducts were detected in the livers of rats exposed to 2,4-diaminotoluene; otherwise, a marginal amount of adducts was detected in the case 2,6-diaminotoluene . In studies on the comparison with 2,4-diaminotoluene, negative results for 2,6-diaminotoluene with respect to hepatocellular proliferation in rats , an unscheduled DNA synthesis (UDS) assay in rats , a mutagenic assay with LacI transgenic mice , and a guanine-hypoxanthine phosphoribosyltransferase (gpt) mutagenic assay with transgenic rats [23, 78] have been reported, although positive results were also confirmed at a high dose of 2,6-diaminotoluene via a UDS assay . Furthermore, the negative result was confirmed by a comet assay of various mouse organs . 2,6-Diaminotoluene might be more efficiently detoxified than 2,4-diaminotoluene in vivo because its para site at position 4 can be oxidized and subsequently conjugated by phase II enzymes . The formation of the DNA adduct via the nitrenium ion is important for the Ames assay, and it has been reported that the mutagenicity (+S9) of 2,5-diaminotoluene is lower than that of 2,6-diaminotoluene . In the present study, the results showed that 2,5-diaminotoluene was not mutagenic and 2,6-diaminotoluene was mutagenic according to the TK assay in the 4-h treatment with S9 and in the 24-h treatment without S9, which was assumed to affect the formation of the DNA adducts via the arylnitrenium ion.
ROS are formed and mediated by oxidation to quinonediimines in keratinocytes, in which the expression of oxidative enzymes is minimal for arylamines . Additionally, ROS are formed via hydroxylamines and aminophenols, which are P450-mediated oxidative metabolites for arylamines [65, 71, 72]. For arylamines such as 2,6-dimethylaniline and 3,5-dimethylaniline, the principal mechanism of mutagenic action is likely associated with aminophenol/quinone imine redox cycling to generate ROS rather than the formation of covalent DNA adducts via the arylnitrenium ion [71, 72]. 2,6-diaminoltoluene may be metabolized in the same pathway as arylamines. Metabolites of 2,6-diaminoltoluene that contribute to ROS formation and accumulation are unknown.
HC Blue No. 2
Nitro compounds are reduced to hydroxylamines with nitroreductases, such as NAD(P) H quinone oxidoreductase (NQO1) and P450 oxidoreductase (POR), and are O-acetylized to acetoxyarylamines with acetyltransferase. These metabolites break down spontaneously into arylnitrenium ions that form adducts with bases in nucleic acids [64, 65]. NQO1 is expressed in TK6 in the normal state , and the proteomics analysis in the present study, measured nitroreductases, such as NQO1 and POR, and proteins related to O-acetylization (N-acetyltransferase) in the vehicle TK6 control (Additional file 3 (Table S3)). DNA adduct formation via the arylnitrenium ion and ROS formation via hydroxylamine likely contribute to DNA damage caused by HC Blue No. 2 in TK6 cells after 24 h of exposure in the absence of S9 (Fig. 5) .
HC blue No. 1 has been determined to be a carcinogen, whereas HC blue No.2 has been classified as a non-carcinogenic. However, reinterpretation of bioassays showed that the results did not justify HC blue No. 2 being classified as a carcinogenic or as definitely a non-carcinogenic . The MLA (4 h) for HC blue No. 2 was positive in the presence of S9 [27, 29, 68], and the mutagenicity in MLA (+S9) for carcinogenic HC blue No. 1 has also been reported to be positive with weak response . Meanwhile, a negative result has been reported for carcinogenic HC blue No. 1 with high purity in MLA (−S9) and other mutagenic assays [27, 85]. For HC blue No. 2 with high purity, a positive result was obtained in MLA (−S9) though the evaluation of carcinogenicity has not been reported . The differences between in vitro and in vivo metabolites may be responsible for the different results of HC blue No. 1 related to mutagenicity (negative in the absence of S9) and carcinogenicity (carcinogenic) . The in vivo metabolite of HC blue No. 2 was single, and the non-carcinogenic result of HC blue No. 2 in animal experiments may be due to the conjugation and the excretion of the metabolite .
This chemical possesses three functional groups: nitro, amino, and carboxyl. However, it is presumably metabolized and excreted relatively quickly because of the carboxyl group. Therefore, a negative result of chemical in TK6 assays (4 h, +S9) in the present study was likely due to the facilitation of its oxidative metabolism. Furthermore, 4-nitroanthranilic acid showed a negative result upon exposure for 4 h and a positive result after 24 h in the absence of S9. DNA damage was likely induced by the arylnitrenium ion and ROS via the reductive metabolism of the nitro group that required longer time frames 24 h (Fig. 5). Major oxidative enzymes that act on amino groups, such as P450s, amine N-oxidase, and FMO, were not confirmed, and nitro reductive enzymes, such as NQO1 and POR, were detected via proteomic analysis of TK6 control cells (Additional file 3 (Table S3)) [65, 69, 83].
Oxidative stress revealed by proteomics analysis
The enrichment analysis of DEPs in the 24 h treatment condition suggested that oxidative stress plays a key role in the enhancement of cytotoxicity and genotoxicity in continuous exposure conditions (Fig. 4). PDK1, SOD2, and DIABLO were expressed prominently after a 24 h exposure of cells to 800 μg/mL 4-nitroanthranilic acid in the present study. These proteins are all involved in response to oxidative stress. A dose- and time-dependent increasing tendency of major antioxidant enzymes (catalase and GSR) was confirmed in the proteomics analysis (Additional file 3 (Table S3) and 4 (Figure S1)). After the short-term treatment, antioxidant defenses such as glutathione and related enzymes in TK6 cells were able to suppress ROS damage, but with the long-term treatment, the increase of oxidative stress caused by depletion of antioxidant enzymes is expected. The prominent decrease of several antioxidant enzymes (GPX4, MGST3, TXN2, TXNRD1, etc.) at 24 h in contrast to 4 h in the proteomics analysis may reflect such depletion (Additional file 3 (Table S3) and 4 (Figure S1)). Additionally, oxidative stress has been considered as a common underlying mechanism for in vitro specific genotoxicity. Generally, mammalian tissues in vivo are likely to display greater antioxidant defenses than cells in culture. Chemicals that induce genotoxicity via the production of ROS, in such case, would damage DNA directly but would be expected to have a threshold . Further analyses are needed to confirm whether these same features of DEPs are observed in other chemicals with conflicting test results. If the same features are observed, quantification of these proteins (or genes) or ROS would provide a promising solution to discriminating in vitro specific positive results from a mode of action point of view.
The usefulness of the TK6 assay, by current methods that have not been combined with new technologies such as proteomics, was found to be limited as a follow-up test, although it still may help to reduce some false positive results (20%) in Ames tests. Furthermore, in vitro specific genotoxicity was prominently exhibited in the long-term treatment, but proteomics analysis of a false-positive compound revealed a mode of action, in which genotoxicity caused was induced by oxidative stress during the long-term treatment. Therefore, an integration of omics technology and TK6 assays may contribute to interpreting irregular results in 24 h specific positives and might lead to further reduction (> 20%) of false positives in follow-up to Ames tests.
Availability of data and materials
- TK6 assay:
Thymidine kinase gene mutation assay
Weight of evidence
The Japanese Environmental Mutagen Society
The Mammalian Mutagenicity Study group
The Organization for Economic Co-operation and Development
Adverse outcome pathway
Integrated approaches to testing and assessment
Scientific Committee on Consumer Safety
Mouse lymphoma assay
Relative suspension growth
Relative total growth
Not a number
Differentially expressed protein
ICH. ICH guideline M7 - genotoxic impurities - assessment and control of DNA reactive (mutagenic) impurities to limit potential carcinogenic risk. Guideline. 2014;44:30 Available from: http://www.ich.org/products/guidelines/multidisciplinary/article/multidisciplinary-guidelines.html.
Hardy A, Benford D, Halldorsson T, Jeger M, Knutsen HK, More S, et al. Clarification of some aspects related to genotoxicity assessment. EFSA J. 2017;15(12):e05113.
ECHA. Chapter R.7a: endpoint specific guidance. In: Guid. Inf. Requir. Chem. Saf. Assessment. Version 6.0; 2017.
Kirkland D, Aardema M, Henderson L, Müller L. Evaluation of the ability of a battery of three in vitro genotoxicity tests to discriminate rodent carcinogens and non-carcinogens: I. sensitivity, specificity and relative predictivity. Mutat Res Genet Toxicol Environ Mutagen. 2005;584:1–256.
Matthews EJ, Kruhlak NL, Cimino MC, Benz RD, Contrera JF. An analysis of genetic toxicity, reproductive and developmental toxicity, and carcinogenicity data: I. identification of carcinogens using surrogate endpoints. Regul Toxicol Pharmacol. 2006;44:83–96.
Tweats D, Bourdin Trunz B, Torreele E. Genotoxicity profile of fexinidazole--a drug candidate in clinical development for human African trypanomiasis (sleeping sickness). Mutagenesis. 2012;27(5):523–32. https://doi.org/10.1093/mutage/ges015 Epub 2012 Apr 26. PMID: 22539226.
OECD. Guidance document for the use of adverse outcome pathways in developing Integrated Approaches to Testing and Assessment (IATA). OECD Guidel Test Chem. 2016
Fujita Y, Honda H, Yamane M, Morita T, Matsuda T, Morita O. A decision tree-based integrated testing strategy for tailor-made carcinogenicity evaluation of test substances using genotoxicity test results and chemical spaces. Mutagenesis. 2019;34(1):101–9.
European Commission. SCCS (Scientific Committee on Consumer Safety), SCCS notes of guidance for the testing of cosmetic ingredients and their safety evaluation 10th revision: SCCS; 2018. Available from: http://ec.europa.eu/health//sites/health/files/scientific_committees/consumer_safety/docs/sccs_o_190.pdf
Committee S, Sccs CS. Scientific Committee on Consumer Safety Basic Brown 17 COLIPA n ° B007. 2014.
Pfuhler S, Pirow R, Downs TR, Haase A, Hewitt N, Luch A, et al. Validation of the 3D reconstructed human skin comet assay, an animal-free alternative for following-up positive results from standard in vitro genotoxicity assays. Mutagenesis. 2020;geaa009. https://doi.org/10.1093/mutage/geaa009.
SCCP. ADDENDUM to the SCCS’s notes of guidance for the testing of cosmetic ingredients and their safety evaluation, 8th revision. The SCCS adopted this Addendum on 9 April 2014 by written procedure, SCCS/1501/12 (2014).
Ates G, Raitano G, Heymans A, Van Bossuyt M, Vanparys P, Mertens B, et al. In silico tools and transcriptomics analyses in the mutagenicity assessment of cosmetic ingredients: a proof-of-principle on how to add weight to the evidence. Mutagenesis. 2016;31:453–61 Oxford Academic.
Kirkland, et al. Can in vitro mammalian cell genotoxicity test results be used to complement positive results in the Ames test and help predict carcinogenic or in vivo genotoxic activity? I. Reports of individual databases presented at an EURL ECVAM Workshop. Mutat Res Genet Toxicol Environ Mutagen. 2014;775–776:55–68.
Kirkland, et al. Can in vitro mammalian cell genotoxicity test results be used to complement positive results in the Ames test and help predict carcinogenic or in vivo genotoxic activity? II. Construction and analysis of a consolidated database. Mutat Res Genet Toxicol Environ Mutagen. 2014;775–776:69–80.
Negrini S, Gorgoulis VG, Halazonetis TD. Genomic instability an evolving hallmark of cancer. Nat Rev Mol Cell Biol. 2010;11:220–8.
OECD. Test No. 490: In vitro mammalian cell gene mutation tests using the thymidine kinase gene: OECD Guidelines for the Testing of Chemicals, OECD, Paris (2016). https://doi.org/10.1787/9789264264908-en.
Zeiger E, Anderson B, Haworth S, Lawlor T, Mortelmans K. Salmonella mutagenicity tests: V. Results from the testing of 311 chemicals. Environ Mol Mutagen. 1992;19:2–141 John Wiley & Sons, Ltd.
Dunkel VC, Zeiger E, Brusick D, McCoy E, McGregor D, Mortelmans K, et al. Reproducibility of microbial mutagenicity assays: II. Testing of carcinogens and noncarcinogens in Salmonella typhimurium and Escherichia coli. Environ Mutagen. 1985;7:1–19.
National Cancer Institute CARCINOGENESIS Technical Report Series 33 No. 177. 1979. Available from: https://ntp.niehs.nih.gov/ntp/htdocs/lt_rpts/tr177.pdf. Accessed 16 Sept 2020.
Claxton LD, Dearfield KL, Spanggord RJ, Riccio ES, Mortelmans K. Comparative mutagenicity of halogenated pyridines in the Salmonella typhimurium/mammalian microsome test. Mutat Res. 1987;176:185–98.
National Cancer Institute CARCINOGENESIS Technical Report Series No. 178. 1979. Available from: https://ntp.niehs.nih.gov/ntp/htdocs/lt_rpts/tr178.pdf. Accessed 16 Sept 2020.
Toyoda-Hokaiwado N, Inoue T, Masumura K, Hayashi H, Kawamura Y, Kurata Y, et al. Integration of in vivo genotoxicity and short-term carcinogenicity assays using F344 gpt delta transgenic rats: In vivo mutagenicity of 2,4-diaminotoluene and 2,6-diaminotoluene structural isomers. Toxicol Sci. 2010;114:71–8.
National Cancer Institute CARCINOGENESIS Technical Report Series No. 200 NTP No. 80–20. 1980. Available from: https://ntp.niehs.nih.gov/ntp/htdocs/lt_rpts/tr200.pdf. Accessed 16 Sept 2020.
Chung KT, Murdock CA, Stevens SE, Li YS, Wei CI, Huang TS, et al. Mutagenicity and toxicity studies of p-phenylenediamine and its derivatives. Toxicol Lett. 1995;81:23–32.
National Cancer Institute CARCINOGENESIS Technical Report Series No. 126. 1978. Available from: https://ntp.niehs.nih.gov/ntp/htdocs/lt_rpts/tr126.pdf. Accessed 16 Sept 2020.
Oberly TJ, Kokkino AJ, Bewsey BJ, Richardson KK. Mutagenicity evaluation of HC Blue No. 1 and HC Blue No. 2. III. Effects in the Salmonella typhimurium/Escherichia coli reversion assay and the mouse lymphoma L5178Y TK+/− forward mutation assay. Mutat Res. 1990;241:151–9.
Zeiger E, Anderson B, Haworth S, Lawlor T, Mortelmans K. Salmonella mutagenicity tests: IV. Results from the testing of 300 chemicals. Environ Mol Mutagen. 1988;11:1–157.
National Toxicology Program Technical Report Series No. 293. 1985. Available from: https://ntp.niehs.nih.gov/ntp/htdocs/lt_rpts/tr293.pdf. Accessed 16 Sept 2020.
National Toxicology Program Technical Report Series No. 276. 1985. Available from: https://ntp.niehs.nih.gov/ntp/htdocs/lt_rpts/tr276.pdf. Accessed 16 Sept 2020.
Rumsby P, Lawton L, Ewence A, Rockett L, Hall T, James H, Jackson P. Review of the current toxicological and occurrence information available on iodinated disinfection by-products. DEFRA 7883.03 (DWI 70/2/233). Swindon; 2009. www.wrcplc.co.uk.
National Cancer Institute CARCINOGENESIS Technical Report Series No. 110. 1978. Available from: https://ntp.niehs.nih.gov/ntp/htdocs/lt_rpts/tr110.pdf. Accessed 16 Sept 2020.
National Toxicology Program. 4-Nitroanthranilic acid (619–17-0) | Chemical Effects in Biological Systems. Available from: https://manticore.niehs.nih.gov/cebssearch/test_article/619-17-0. Cited 2020 Sep 14.
National Cancer Institute CARCINOGENESIS Technical Report Series No. 109. 1978. Available from: https://ntp.niehs.nih.gov/ntp/htdocs/lt_rpts/tr109.pdf. Accessed 16 Sept 2020.
Mortelmans K, Haworth S, Lawlor T, Speck W, Tainer B, Zeiger E. Salmonella mutagenicity tests: II. Results from the testing of 270 chemicals. Environ Mutagen. 1986;8:1–119.
National Cancer Institute CARCINOGENESIS Technical Report Series No. 64. 1978. Available from: https://ntp.niehs.nih.gov/ntp/htdocs/lt_rpts/tr064.pdf. Accessed 16 Sept 2020.
Shahin MM, Von Borstel RC. Comparisons of mutation induction in reversion systems of Saccharomyces cerevisiae and Salmonella typhimurium. Mutat Res. 1978;53:1–10.
IARC Monographs on the evaluation of the carcinogenic risk of chemicals to man; Some aromatic amines and related nitro compounds –hair dyes, colouring agents and miscellaneous industrial chemicals. Lyon; 1978;16. ISBN; 92 832 1216 9.
Searle CE, Harnden DG, Venitt S, Gyde OHB. Carcinogenicity and mutagenicity tests of some hair colourants and constituents. Nature. 1975;255:506–7.
U.S. Department of HealthEducation, and Welfare, National Toxicology Program. Bioassay of 4-Nitro-o-phenylenediamine for possible carcinogenicity. National Cancer Institute Carcinogenesis technical report series no. 180. 1979.
National Cancer Institute CARCINOGENESIS Technical Report Series No. 180. 1979. Available from: https://ntp.niehs.nih.gov/ntp/htdocs/lt_rpts/tr180.pdf. Accessed 16 Sept 2020.
Omori T, Honma M, Hayashi M, Honda Y, Yoshimura I. A new statistical method for evaluation of L5178Ytk+/− mammalian cell mutation data using microwell method. Mutat Res Genet Toxicol Environ Mutagen. 2002;517:199–208.
Simpson DG, Margolin BH. Recursive nonparametric testing for dose-response relationship subject to downturns at high dose. Biometrika. 1986;73:589–96.
Masuda T, Tomita M, Ishihama Y. Phase transfer surfactant-aided trypsin digestion for membrane proteome analysis. J Proteome Res. 2008;7:731–40 American Chemical Society.
Adachi J, Hashiguchi K, Nagano M, Sato M, Sato A, Fukamizu K, et al. Improved proteome and phosphoproteome analysis on a cation exchanger by a combined acid and salt gradient. Anal Chem. 2016;88:7899–903.
Scientific Committee on Consumer Safety, Opinion on toluene-2,5-diamine and its sulfate. Colipa n° A5. Brussels: European Commission; 2012.
Mitchell AD, Auletta AE, Clive D, Kirby PE, Moore MM, Myhr BC. The L5178Y/tk(+/−) mouse lymphoma specific gene and chromosomal mutations assay: a phase III report of the U.S. environmental protection agency Gene-Tox program. Mutat Res. 1997;394:177–303.
Environemental Protection Agency. Provisional peer reviewed toxicity values for 2,6-Toluenediamine. 2005; Available from: https://cfpub.epa.gov/ncea/pprtv/documents/Toluenediamine26.pdf.
EURL ECVAM genotoxicity and carcinogenicity consolidated database. Available from: https://ec.europa.eu/jrc/en/scientific-tools/Ames. Accessed 22 Dec 2020.
McGregor DB, Brown A, Cattanach P, Edwards I, McBride D, Caspary WJ. Responses of the L5178Y tk+/tk− mouse lymphoma cell forward mutation assay ii: 18 coded chemicals. Environ Mol Mutagen. 1988;11:91–118.
National Toxicology Program. Mouse lymphoma assay; Aliquot No. 721150. Available from: https://manticore.niehs.nih.gov/cebssearch/test_article/619-17-0. Accessed 22 Dec 2020.
Mitchell AD, Rudd CJ, Caspary WJ. Evaluation of the L5178Y mouse lymphoma cell mutagenesis assay: Intralaboratory results for sixty-three coded chemicals tested at sri international. Environ Mutagen Environ Mol Mutagen. 1988;12:37–101.
Myhr BC, Caspary WJ. Evaluation of the L5178Y mouse lymphoma cell mutagenesis assay: Intralaboratory results for sixty-three coded chemicals tested at litton bionetics, inc. Environ Mol Mutagen. 1988;12:103–94.
NTP database; 2,5-Toluenediamine sulfate 10736-T. Available from: https://ntp.niehs.nih.gov/whatwestudy/testpgm/status/ts-10736-t.html?utm_source=direct&utm_medium=prod&utm_campaign=ntpgolinks&utm_term=ts-10736-t. Cited 2020 Aug 14.
Honma M, Hayashi M. Comparison of in vitro micronucleus and gene mutation assay results for p53-competent versus p53-deficient human lymphoblastoid cells. Environ Mol Mutagen. 2011;52(5):373–84. https://doi.org/10.1002/em.20634 Epub 2010 Oct 20. PMID: 20963812.
Whitwell J, Smith R, Jenner K, Lyon H, Wood D, Clements J, Aschcroft-Hawley K, Gollapudi B, Kirkland D, Lorge E, Pfuhler S, Tanir JY, Thybaud V. Relationships between p53 status, apoptosis and induction of micronuclei in different human and mouse cell lines in vitro: Implications for improving existing assays. Mutat Res Genet Toxicol Environ Mutagen. 2015;789–790:7–27. https://doi.org/10.1016/j.mrgentox.2015.05.011 Epub 2015 May 30. PMID: 26232254.
Zeiger E. Identification of rodent carcinogens and noncarcinogens using genetic toxicity tests: premises, promises, and performance. Regul Toxicol Pharmacol. 1998;28(2):85–95. https://doi.org/10.1006/rtph.1998.1234 PMID: 9927558.
Morita T, Hamada S, Masumura K, Wakata A, Maniwa J, Takasawa H, Yasunaga K, Hashizume T, Honma M. Evaluation of the sensitivity and specificity of in vivo erythrocyte micronucleus and transgenic rodent gene mutation tests to detect rodent carcinogens. Mutat Res Genet Toxicol Environ Mutagen. 2016;802:1–29. https://doi.org/10.1016/j.mrgentox.2016.03.008 Epub 2016 Mar 21. PMID: 27169373.
National Toxicology Program. Testing Status of 4-(Chloroacetyl) acetanilide 10592-P, No.223043. Available from: https://ntp.niehs.nih.gov/whatwestudy/testpgm/status/ts-10592-p.html?utm_source=direct&utm_medium=prod&utm_campaign=ntpgolinks&utm_term=ts-10592-p. Accessed 22 Dec 2020.
National Toxicology Program. Testing status of 2-Chloromethylpyridine hydrochloride 10738-C, No.145263, 448657, A65556, 861777, and 145263, NTP. Available from: https://ntp.niehs.nih.gov/whatwestudy/testpgm/status/ts-10738-c.html?utm_source=direct&utm_medium=prod&utm_campaign=ntpgolinks&utm_term=ts-10738-c. Accessed 22 Dec 2020.
National Toxicology Program. Testing status of 8-Hydroxyquinoline 10598-N, No. 470651, 592699, 616445, 375977, 980523, and 470651. Available from: https://ntp.niehs.nih.gov/whatwestudy/testpgm/status/ts-10598-n.html?utm_source=direct&utm_medium=prod&utm_campaign=ntpgolinks&utm_term=ts-10598-n. Accessed 22 Dec 2020.
National Toxicology Program. Testing status of 1-Nitronaphthalene 10425-R, No. 153730. Available from: https://ntp.niehs.nih.gov/whatwestudy/testpgm/status/ts-10425-r.html?utm_source=direct&utm_medium=prod&utm_campaign=ntpgolinks&utm_term=ts-10425-r. Accessed 22 Dec 2020.
National Toxicology Program. Testing status of 4-Nitro-o-phenylenediamine 10476-J, No. 704945, 549927, A82327, and 704945, NTP. Available from: https://ntp.niehs.nih.gov/whatwestudy/testpgm/status/ts-10476-j.html?utm_source=direct&utm_medium=prod&utm_campaign=ntpgolinks&utm_term=ts-10476-j. Accessed 22 Dec 2020.
Martin CN, Beland FA, Roth RW, Kadlubar FF. Covalent binding of benzidine and n-acetylbenzidine to dna at the c-8 atom of deoxyguanosine in vivo and in vitro. Cancer Res. 1982;42:2678–86.
Murata M, Kawanishi S. Mechanisms of oxidative DNA damage induced by carcinogenic arylamines. Front Biosci. 2011;16:1132–43.
Ziegler DM, Ansher SS, Nagata T, Kadlubar FF, Jakoby WB. N-methylation: potential mechanism for metabolic activation of carcinogenic primary arylamines. Proc Natl Acad Sci U S A. 1988;85:2514–7.
O’Brien KAF, Gatehouse DG, Tiley M. Induction of mutations in TK6 human lymphoblastoid cells by ethyl methanesulphonate, benzo[ a ]pyrene and benzidine. Mutagenesis. 1990;5:55–60 Oxford Academic.
Myhr BC, Caspary WJ, Holden HE. Chemical mutagenesis at the thymidine kinase locus in L5178Y mouse lymphoma cells: results for 31 coded compounds in the national toxicology program. Environ Mol Mutagen. 1991;18:51–83.
McGregor DB, Edwards I, Roland Wolf C, Forrester LM, Caspary WJ. Endogenous xenobiotic enzyme levels in mammalian cells. Mutat Res. 1991;261:29–39.
Zanoni TB, Hudari F, Munnia A, Peluso M, Godschalk RW, Zanoni MVB, et al. The oxidation of p-phenylenediamine, an ingredient used for permanent hair dyeing purposes, leads to the formation of hydroxyl radicals: oxidative stress and DNA damage in human immortalized keratinocytes. Toxicol Lett. 2015;239:194–204. https://doi.org/10.1016/j.toxlet.2015.09.026 Elsevier Ireland Ltd.
Chao MW, Kim MY, Ye W, Ge J, Trudel LJ, Belanger CL, et al. Genotoxicity of 2,6- and 3,5-dimethylaniline in cultured mammalian cells: the role of reactive oxygen species. Toxicol Sci. 2012;130:48–59.
Chao MW, Erkekoglu P, Tseng CY, Ye W, Trudel LJ, Skipper PL, et al. Intracellular generation of ROS by 3,5-dimethylaminophenol: persistence, cellular response, and impact of molecular toxicity. Toxicol Sci. 2014;141:300–13.
Gold LS, de Veciana M, Backman GM, Magaw R, Lopipero P, Smith M, et al. Chronological supplement to the carcinogenic potency database: standardized results of animal bioassays published through December 1982. Environ Health Perspect. 1986;67:161–200.
Taningher M, Peluso M, Parodi S, Ledda-Columbano GM, Columbano A. Genotoxic and non-genotoxi activities of 2,4- and 2,6-diaminotoluene, as evaluated in Fischer-344 rat liver. Toxicology. 1995;99:1–10.
Cunningham ML, Foley J, Maronpot RR, Matthews HB. Correlation of hepatocellular proliferation with hepatocarcinogenicity induced by the mutagenic noncarcinogen: carcinogen pair—2,6- and 2,4-diaminotoluene. Toxicol Appl Pharmacol. 1991;107:562–7.
George E, Westmoreland C. Evaluation of the in vivo genotoxicity of the structural analogues 2,6-diaminotoluene and 2,4-diaminotoluene using the rat micronucleus test and rat liver UDS assay. Carcinogenesis. 1991;12:2233–7.
Cunningham ML, Hayward JJ, Shane BS, Tindall KR. Distinction of mutagenic carcinogens from a mutagenic noncarcinogen in the Big Blue transgenic mouse. Environ Health Perspect. 1996;104(Suppl):683–6.
Sui H, Ohta R, Shiragiku T, Akahori A, Suzuki K, Nakajima M, et al. Evaluation of in vivo mutagenicity by 2,4-diaminotoluene and 2,6-diaminotoluene in liver of F344 gpt delta transgenic rat dosed for 28 days: a collaborative study of the gpt delta transgenic rat mutation assay. Genes Environ. 2012;34:25–33.
Allavena A, Martelli A, Robbiano L, Brambilla G. Evaluation in a battery of in vivo assays of four in vitro genotoxins proved to be noncarcinogens in rodents. Teratog Carcinog Mutagen. 1992;12:31–41.
Sasaki YF, Fujikawa K, Ishida K, Kawamura N, Nishikawa Y, Ohta S, et al. The alkaline single cell gel electrophoresis assay with mouse multiple organs: results with 30 aromatic amines evaluated by the IARC and U.S. NTP. Mutat Res. 1999;440:1–18.
Cunningham ML, Burka LT, Matthews HB. Metabolism, disposition, and mutagenicity of 2,6-diaminotoluene, a mutagenic noncarcinogen. Drug Metab Dispos. 1989;17:612–7.
Hatch FT, Knize MG, Colvin ME. Extended quantitative structure-activity relationships for 80 aromatic and heterocyclic amines: structural, electronic, and hydropathic factors affecting mutagenic potency. Environ Mol Mutagen. 2001;38:268–91.
Revollo J, Petibone DM, McKinzie P, Knox B, Morris SM, Ning B, Dobrovolsky VN. Whole genome and normalized mRNA sequencing reveal genetic status of TK6, WTK1, and NH32 human B-lymphoblastoid cell lines. Mutat Res Genet Toxicol Environ Mutagen. 2016;795:60–9.
Prival MJ, Dunkel VC. Reevaluation of the mutagenicity and carcinogenicity of chemicals previously identified as “false positives” in the Salmonella typhimurium mutagenicity assay. Environ Mol Mutagen. 1989;13:1–24.
Burnett CM, Corbett JF. Failure of short-term in vitro mutagenicity tests to predict the animal carcinogenicity of hair dyes. Food Chem Toxicol. 1987;25:703–7.
Kirkland DJ, Aardema M, Banduhn N, Carmichael P, Fautz R, Meunier JR, et al. In vitro approaches to develop weight of evidence (WoE) and mode of action (MoA) discussions with positive in vitro genotoxicity results. Mutagenesis. 2007;22:161–75.
The authors are grateful to Dr. Mariko Murata (Department of Environmental and Molecular Medicine, Mie University) for helpful comments and supporting the figure preparation. The authors would like to thank Enago (www.enago.com) for the English language review.
This study was supported by a Grant-in-Aid for Scientific Research from the Ministry of Health, Labor and Welfare (H30-Food-General-003).
Ethics approval and consent to participate
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Additional file 1.
Additional file 2.
Additional file 3.
Additional file 4.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
About this article
Cite this article
Yasui, M., Fukuda, T., Ukai, A. et al. Weight of evidence approach using a TK gene mutation assay with human TK6 cells for follow-up of positive results in Ames tests: a collaborative study by MMS/JEMS. Genes and Environ 43, 7 (2021). https://doi.org/10.1186/s41021-021-00179-1
- TK6 assay
- Human lymphoblastoid TK6 cells
- Ames test
- Weight of evidence approach