- Email: [email protected]
Chromosome aberrations in Chinese hamster and human cells: A comparison using compounds with various genotoxicity profiles
Mutation Research 616 (2007) 103–118
Chromosome aberrations in Chinese hamster and human cells: A comparison using compounds with various genotoxicity profiles Cathy Hilliard ∗ , Rosina Hill, Michael Armstrong, Clint Fleckenstein, Jessica Crowley, Elizabeth Freeland, Danielle Duffy, Sheila M. Galloway Merck Research Laboratories, Department of Genetic and Cellular Toxicology, West Point, PA 19486, United States Available online 14 December 2006
Abstract Chromosome aberrations (Cabs) can be induced in vitro by non-DNA damaging compounds, often associated with cytotoxicity and DNA synthesis inhibition, and under conditions that would not be relevant in vivo. Such misleading positive results are reported both in Chinese hamster cell lines and in human peripheral blood lymphocytes (HL). We assessed the response of HL to compounds with varied genetic toxicity profiles, all of which induced Cabs in CHO cells Seven of 10 compounds were negative or equivocal in HL. Results in purified lymphocytes for four verified that the difference was not due to the presence of blood in cultures. Two compounds that were weakly positive in the Ames test and one that induced DNA adducts were negative or equivocal in the HL assay; their overall mutagenic potential in vivo is not clear. Of four Ames-negative compounds, three of which inhibited DNA synthesis in CHO cells, three were negative and one was equivocal in the HL assay. A potent Cab inducer, which also induced micronuclei in vivo (but was negative in the Ames test) was clearly positive in the HL assay. Two compounds were clearly positive in HL only when the mitotic indices (MI) were below 50% of control. These are genotoxic in other assays but our evidence suggests that Cab induction is related more to toxicity than to primary DNA damage. For this limited set of 10 compounds, HL were more likely than CHO cells to give negative or equivocal results. It is likely that more stringent checkpoint controls in human cells prevent damaged cells reaching mitosis, and may also influence the reported greater sensitivity to induction of aneuploidy and polyploidy of normal rodent compared with human cells. In the studies reported here, two strong inducers of polyploidy in CHO cells gave weaker increases in HL. Human lymphocytes have disadvantages as a routine screening assay (finding donors, known individual variability, increased time required and the inadequacy of the MI as a toxicity measure), but may be useful in follow-up testing to assess weight of evidence about genotoxic risk to humans, for compounds that are positive in the Chinese hamster cell Cabs assays. © 2006 Elsevier B.V. All rights reserved. Keywords: Chromosome aberrations; Genotoxicity testing; Human lymphocytes; Hamster cells; Indirect mechanisms
1. Introduction It is now widely recognized that in vitro assays for chromosome aberrations may give misleading positive results, especially when testing extends into the toxic ∗ Corresponding author. Tel.: +1 215 652 5666; fax: +1 215 652 4944. E-mail address: catherine [email protected] (C. Hilliard).
0027-5107/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.mrfmmm.2006.11.013
dose range. In many cases the evidence indicates these positive results occur without direct attack on DNA, and under conditions that would not be relevant in vivo. Such positive results are reported both in permanent hamster cell lines such as CHO and CHL, and also in cultured human peripheral blood lymphocytes (HL) [1]. We have found that chromosome aberrations can be induced by non-DNA damaging compounds, often associated with cytotoxicity and inhibition of DNA synthesis
104
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
[2,3]. Since the polymerase inhibitor aphidicolin induces chromosome aberrations in normal human cells [4,5], we were interested in assessing the response of lymphocytes to drug candidates that induced aberrations in CHO cells associated with inhibition of DNA synthesis. We extended the comparison to compounds with a variety of genotoxicity profiles encountered in drug development, including those with weak positive Ames test results, compounds that induced polyploidy in CHO cells, compounds that induced aberrations at toxic doses in CHO cells but did not inhibit DNA synthesis, and compounds thought to have more than one mechanism of genotoxicity. Since some drug candidates are known to be highly protein bound, and we had demonstrated marked shifts in the dose response to higher concentrations when we increased the serum concentrations in the treatment medium from 10% to, e.g., 50% serum, some experiments were done to compare results at different serum concentrations. Similarly, in certain cases experiments were done both with whole blood cultures and with purified lymphocytes, to affirm that a difference in results between CHO cells and lymphocytes was due to cell type difference, and not interference from blood components. 2. Materials and methods 2.1. Test chemicals and solvents All compounds were from Merck & Co., Inc. except for aphidicolin, which was from Sigma Chemical Company (St. Louis, MO, USA). All stocks were prepared fresh for each experiment. All compounds tested, including aphidicolin, were prepared as 100× concentrations to give a final concentration of 1% DMSO. DMSO was from Sigma Chemical Company. Results are presented with compounds grouped by therapeutic class because typically the molecules in each group share structural features, and knowledge of genotoxicity profiles of additional related compounds can in some cases inform the discussion of the overall genotoxicity profile. 2.2. Cell culture and treatment 2.2.1. CHO cells Chinese hamster ovary (CHO) cells, clone WBL, were cultured in a humidified atmosphere of 5% CO2 in air at 37 ◦ C in McCoy’s 5A medium (Invitrogen, Grand Island, New York, USA), supplemented with 10% fetal bovine serum (Hyclone, Logan, Utah, USA), 2 mM l-glutamine, and 100 U/mL penicillin, 100 U/mL streptomycin (all Invitrogen). Compounds G, H, I, and N were also tested using 50% pooled human serum (type AB, from SeraCare Inc., Oceanside, CA, USA). Cells were seeded the day before treatment at 1.2 × 106 /10 mL/75-
cm2 flask (Corning, Corning, NY, USA). Cultures treated with the metabolic activation system were re-fed with serum-free S-9 mix just before treatment. Test compounds were dissolved in DMSO or dH2 O, added to cultures, and the cultures were incubated at 37 ◦ C. Treatments were for either 3 or 20 h, and aberrations were scored in cells harvested 20 h from the beginning of treatment. At the termination of the 3-h treatments, the cultures were washed twice with Hanks Balanced Salt Solution (HBSS) (Invitrogen) and re-fed with fresh complete medium. Cultures were incubated for a further 17 h. Colcemid (0.1 g/mL; Invitrogen) was added 1–2 h before monolayers were harvested by trypsinization. A cell sample was counted by Coulter counter (Beckman Coulter, Inc., Miami, FL, USA) to determine cell number as an indicator of cytotoxicity. (We verified that these counts represented viable cells by checking samples for trypan blue dye exclusion.) The cells were treated with hypotonic KCl (75 mM; Sigma Chemical Company) for 1–3 min at room temperature, washed twice with fixative (methanol:glacial acetic acid, 3:1, v/v), dropped onto slides, air-dried, and stained with Gurr Giemsa stain (BDH Chemicals LTD, Poole, England, UK). 2.2.2. Lymphocytes Human blood obtained from healthy adult donors (aged 21–45, nonsmokers without a history of radiotherapy or chemotherapy and lacking current viral infections) was cultured in a humidified atmosphere of 5% CO2 in air at 37 ◦ C in RPMI 1640 medium (Invitrogen), supplemented with 15% fetal bovine serum, 2 mM l-glutamine, 100 U/mL penicillin, 100 U/mL streptomycin, and 2% phytohemaglutanin M (all Invitrogen). Whole blood cultures were set up approximately 48 h before treatment, with 1 mL blood per 9 mL medium in standing 25-cm2 flasks (Corning). Purified lymphocyte cultures were obtained either by Ficoll gradient (Compound K only) or by buffy coat separation. For Ficoll gradient preparations, whole blood was diluted four-fold with phosphate-buffered saline (PBS; Invitrogen, Carlsbad, CA, USA) and layered over Ficoll–Hypaque (Mediatech, Inc., Herndon, VA, USA). After centrifugation at 400 × g for 40 min, the lymphocyte layer was removed and washed twice with PBS containing 3–5% fetal bovine serum. For preparation by buffy coat separation, whole blood was allowed to settle in sodium heparin tubes for approximately 3 h at 37 ◦ C before removing the white cell layer, which was spun down and resuspended in fresh complete medium. For both methods, cell suspensions were seeded at approximately 2.0 × 106 /10 mL/25-cm2 flasks (total white blood cells) and incubated for 48 h before dosing. Cultures treated with the metabolic activation system were spun and resuspended in serum-free medium containing S-9 mix before treatment (cultures used for Compound K were resuspended in medium containing 3–5% serum and S-9 mix). Test compounds were dissolved in DMSO or dH2 O, added to cultures, and the cultures were incubated at 37 ◦ C for either 3 or 24 h.
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
Treatments were for either 3 or 24 h, and aberrations were scored in cells harvested 24 h from the beginning of treatment. (The 24-h treatment length, compared with 20 h for CHO cells, was chosen to allow for the slightly longer average cell cycle length in cultured lymphocytes.) At the termination of the 3-h treatments, the cultures were centrifuged at about 225 × g for 5 min, washed once with HBSS, and re-fed with fresh complete medium. Cultures were incubated for a further 21 h. Colcemid (0.1 g/mL) was added 2–3 h before cultures were harvested. The cells were treated with hypotonic KCl (75 mM) for 10–15 min at 37 ◦ C, washed twice with fixative (methanol:glacial acetic acid, 3:1 v/v), dropped onto slides, air-dried, and stained with Gurr Giemsa stain. 2.3. Metabolic activation system Compounds which gave positive results in CHO cells only with S-9 metabolic activation, or gave more substantial increases with S-9, were tested in HL with S-9 activation. Otherwise, HL were tested only without S-9. S-9 was produced using liver homogenate (S-9 fraction) from phenobarbital/βnapthoflavone-treated male Sprague–Dawley rats (Charles River Laboratories, Raleigh, NC, USA). The S-9 used for these experiments was purchased from MolTox (Boone, NC, USA). S-9 was stored at −70 to −80 ◦ C and thawed immediately before use. S-9 was mixed with sodium NADP (Boehringer Mannheim, Indianapolis, IN, USA), and trisodium isocitrate (Sigma Chemical Company) in serum-free culture medium (culture medium with 3% serum used for Compound K) to give final concentrations of: S-9, 15 L/mL; NADP, 1.05 mM; and trisodium isocitrate, 5.8 mM. 2.4. Dose selection The maximum dose tested was 10 mM. Cytotoxicity was used to select doses, scoring doses that gave a cell growth reduction not greatly exceeding 50% compared with concurrent controls. For CHO cells toxicity was assessed as reductions in population doubling [6] except for Compound P for which reductions in cell counts were used. For HL mitotic index was used. In cases where there was little or no cytotoxicity, doses scored were those that did not produce excessive precipitate in culture medium.
105
of mitotic index and 500 metaphases for the frequency of polyploidy. 2.6. Criteria for a positive result A positive point is one where the percentage of cells with aberrations is significantly greater than the concurrent controls (P ≤ 0.05), and outside the historical control range. The analysis is a pairwise comparison of the percentage of cells with aberrations at each dose level with the controls by Fisher’s exact test, with the P values adjusted for multiple comparisons against a common control by the method of Dunnett [7]. We generally consider a test positive if there are two positive points, or if a single positive point is reproduced in an independent assay. Evidence for a dose relation and increased frequencies of aberrations per cell (presence of cells with more than one aberration) are also taken into account. Although our top dose is usually selected such that growth reduction does not greatly exceed 50%, for the purposes of the current experiments dose levels with mitotic index of less than 50% of controls were included to see if aberrations were induced only with marked growth suppression. 2.7. Flow cytometry for detecting DNA synthesis inhibition Chinese hamster ovary (CHO) cells were treated in parallel cultures to those used in the aberration assay with 0.1 M 5bromodeoxyuridine (BrdUrd; Sigma Chemical Company) for the last 30 min of incubation with test compound. After removal of test compound, cultures were immediately harvested with trypsin. Cells were processed as described in [2], and stained for BrdUrd detection (FITC labeled secondary antibody to mouse anti-BrdUrd) and total cell DNA (propidium iodide, PI) using the iU-4 anti-BrdUrd monoclonal antibody. Analysis of these dual stained CHO cells was performed using an EPICS Elite ESP flow cytometer (Coulter Electronics, Hialeah, FL), equipped with an Innova 90-6 argon-ion laser (Coherent, Inc., Palo Alto, CA) tuned to 488 nm to excite the two fluorochromes (FITC and PI). Two parameter BrdUrd-FITC/DNA-PI histograms were constructed using a linear scale on the x-axis to display the DNA content (PI, red integrated fluorescence [IRFL] and a logarithmic scale on the y-axis to display BrdUrd incorporation (FITC, green integrated fluorescence [LIGFL]. Information from 104 cells was collected for each sample.
2.5. Chromosome aberration scoring Slides for each compound were scored under code by one observer. For each treatment, 200 metaphase cells containing 19–23 (CHO) or 44–48 (HL) chromosomes were scored, unless aberration frequencies were high or too few cells were available for analysis due to mitotic suppression. Gaps (achromatic region less than or equal to the width of a chromatid), polyploidy, and endoreduplicated cells, and pulverized chromosomes were noted but not included in structural aberration totals. Five hundred cells were scored for assessment
2.7.1. Relative BrdUrd incorporation Inhibition was measured as suppression of BrdUrd uptake, i.e. decreased green fluorescence intensity. The relative quantity of BrdUrd incorporated per S phase cell was calculated from the difference in the mean fluorescence intensity (channel number) between the BrdUrd-positive cell population (S phase cells) and the BrdUrd-negative population (G1 , G2 , and M cells). This number calculated for each treated culture was divided by the corresponding number for the controls to give fluorescent intensity as a percentage of control.
106
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
3. Results 3.1. Compounds that induced aberrations and inhibited DNA synthesis in CHO cells 3.1.1. Therapeutic class I 3.1.1.1. Ames test negative compounds. This set of compounds is from a therapeutic class in which the first candidate (data not shown) was positive in the CHO cell chromosome aberration assay but negative in vitro in Ames, hepatocyte DNA stand break (alkaline elution) and TK6 mammalian cell tk mutation assays, and in vivo in the mouse bone marrow micronucleus assay. This compound was also known to be highly proteinbound, and the dose response for aberration induction was markedly shifted to higher concentrations (up to 20fold) when the serum content of the treatment medium was increased from 0 to 50% human serum. A series of related compounds was subsequently tested for aberrations in CHO cells, and three that induced aberrations and DNA synthesis inhibition but were negative in the Ames test are described here. Results reported without serum (FBS) are from incubations with S-9 mix in serum-free medium unless noted otherwise. Compound G induced statistically significant increases in aberrations in CHO cells after 3-h treatments (both with and without S-9, Fig. 1) and 20-h treatment (data not shown) at dose levels between 100 and 1000 M. Aberration induction was not dependent on S-9 activation. Adding 50% human serum instead of the routine 10% FBS shifted the active concentration range to slightly higher doses. The high frequencies of aberrations (numbers in the columns) indicate the presence of some cells with multiple aberrations, a common finding with compounds that inhibit DNA synthesis. Compound G inhibited DNA synthesis dramatically. For example, BrdUrd uptake was suppressed to about 7% of controls at the end of a 3-h treatment with 100 M Compound G without serum, or 500 M with 50% serum. In whole blood, in a first experiment mitotic suppression of 50% or greater was achieved, and there was no increase in aberrations up to 800 M without serum (without S-9, data not shown). With 15% serum (standard treatment conditions for lymphocytes) the value of 2.5% cells with aberrations at 2000 M (Fig. 1e) was statistically significant compared with the concurrent control of 0.25%, but is not considered meaningful compared with historical control data (historical control mean: 0.7; range: 0.0–2.0), and precipitate of drug was visible during treatment. Also, the mitotic index was only 37% of control. In a repeat experiment without FBS (Fig. 1d) a weak but
statistically significant increase in aberrations was seen at 2000 M (6.0% cells with aberrations with a mitotic index at 52% of controls), and again this exceeded the solubility limit for the drug. An experiment repeated with purified lymphocytes was negative at doses up to 2000 M (with mitotic index at 38% of controls). The overall conclusion for lymphocytes for Compound G is equivocal, and most likely negative under standard testing conditions. The cells with multiple aberrations which were seen in CHO cultures were not seen in lymphocytes. Compound H induced weak but statistically significant increases in aberrations in CHO cells after 3-h treatments both with and without S-9 (Fig. 2). Some cells had multiple aberrations. Twenty-hour treatments induced significant increases in aberrations with 10% fetal bovine serum (7.0% and 6.0% cells with aberrations at 15 and 20 M, respectively). Increasing the concentration of serum in the culture medium to 50% human serum suppressed aberration induction, with no significant increases at up to 800 M (Fig. 2c). DNA synthesis was suppressed, the greatest suppression being to about 54% of controls. Levels of DNA synthesis about 60% of controls were seen at 60–100 M without serum, and at 400–600 M with 50% serum. Compound H was tested for micronuclei in vivo and was negative. In HL there were no increases in aberrations either in whole blood cultures with or without serum or in purified lymphocytes (Fig. 2). The top doses scored were limited by the appearance of precipitate in the culture medium in addition to mitotic index suppression. Compound I induced statistically significant increases in aberrations in CHO cells after 3-h treatments both with and without S-9 (Fig. 3). There was no increase in the percentage of cells with aberrations up to 20 M after a 20-h treatment without S-9 (data not shown) but there was one cell with multiple aberrations. Increased serum (50% human serum) shifted the dose response but results were still clearly positive (Fig. 3c). DNA synthesis was markedly suppressed, to less than 5% of controls at the higher doses, about 120 M without serum and about 500 M with 50% serum. In whole blood cultures, 2% cells with aberrations were seen at 200 and 250 M, compared with 0% in controls and there was no mitotic suppression after a 3-h treatment in normal serum (15% FBS) up to 250 M, a dose with visible precipitate (Fig. 3). A repeat experiment confirmed that this result was negative (data not shown) and the assay in purified lymphocytes (buffy coat) was also negative since the value of 4.0% aberrations was not significantly higher than the control value of 1.5% (Fig. 3f).
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
107
Fig. 1. The percentage of cells with chromosomal aberrations in cultures treated with Compound G (concurrent controls for CHO: 0.5–2.5; concurrent controls for lymphocytes: 0.0–1.0). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. No precipitate observed in CHO culture medium at doses scored; precipitate observed at 2000 M and above in lymphocyte culture medium. For lymphocytes without FBS, 800 M was from a separate experiment. ** Statistically significant increase over concurrent control (P ≤ 0.01). * Statistically significant increase over concurrent control (P ≤ 0.05). CHO positive not markedly affected by serum (even 50% human serum). Very weak aberration induction in lymphocytes when compared with CHO even at doses with marked mitotic suppression.
3.1.1.2. Ames test positive compound. The fourth compound in this class, Compound J, differed from the first three in that it was weakly positive in the Ames test, giving a reproducible increase of two- to three-fold in revertants in Salmonella TA1535 at 3000–6000 g/plate. (Our background revertant counts were about 21–27 per plate). No DNA adducts could be detected by 32 P post-labeling in the Salmonella, and the compound was
negative in the in vitro DNA strand break assay in rat hepatocytes (alkaline elution assay.) Thus it is not clear whether DNA damage is a factor in the induction of aberrations seen reproducibly in CHO cells after a 20-h treatment (Fig. 4). Aberrations were not seen after the 3-h treatments, but concentrations with about 50% toxicity were not obtained in that protocol (data not shown). Compound J also inhibited DNA synthesis after 20-h
108
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
Fig. 2. The percentage of cells with chromosomal aberrations in cultures treated with Compound H (concurrent controls for CHO: 0.5–3.0; concurrent controls for lymphocytes: 0.0–2.0). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. Precipitate observed at 80 M and above without FBS and at 600 M with 10% FBS in CHO culture medium. Precipitate observed at 100 M and above without FBS and at 400 M and above with FBS in lymphocyte culture medium. ** Statistically significant increase over concurrent control (P ≤ 0.01). CHO positive suppressed by higher serum concentrations. Purified lymphocytes and whole blood negative.
treatment, with suppression to about 11% of controls at 1000 M. In lymphocytes, after a 3-h treatment there was moderate mitotic suppression and no aberration induction (data not shown), and after a 24-h treatment there were weak but significant increases in aberrations at 750 M (3.0%) and 1000 M (5.0%) with mitotic indices suppressed to 44 and 52% of controls. To see
how this related to toxicity, aberrations were scored at a dose, 1500 M, which suppressed the mitotic index to 16% of control (about 1% compared with controls of about 6–9%). In two experiments, substantial increases in aberrations were seen, to 8.5% (Fig. 4), and to 21% (data not shown). There were also increased frequencies of aberrations per cell. Overall, this compound was pos-
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
109
Fig. 3. The percentage of cells with chromosomal aberrations in cultures treated with Compound I (concurrent controls for CHO: 1.0–3.0; concurrent controls for lymphocytes: 0.0–2.0). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. No precipitate observed in CHO culture medium at doses scored. Precipitate observed in whole blood lymphocyte culture medium with FBS at 250 M. ** Statistically significant increase over concurrent control (P ≤ 0.01). CHO positive found at higher doses with higher concentrations of serum. Purified lymphocytes and whole blood negative.
itive in the lymphocyte assay but the result was weaker than that seen in CHO cells when toxicity in HL was limited to about 50% mitotic suppression. 3.1.2. Therapeutic class II The first compound in this therapeutic class, an anticancer agent, Compound K, gave borderline positive results in the Ames test (reproducible two-fold increases that shifted from TA100 to TA97a in repeat assays with
a pre-incubation protocol, and were not S-9 dependent). The structure suggests it may have some intrinsic genotoxicity. The result of the in vivo mouse bone marrow micronucleus assay was equivocal, since the levels of micronucleated PCE (2.6–2.7 MN-PCE/1000 PCE) after single oral administrations of 1000 and 2000 mg/kg were significant compared with the concurrent control mean of about 1.0 MN-PCE/1000, but within our historical control range (mean about 1.5 MN-PCE/1000 PCE, typi-
110
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
Fig. 4. The percentage of cells with chromosomal aberrations in cultures treated with Compound J (concurrent controls for CHO: 1.0–2.0; concurrent controls for lymphocytes: 0.0–0.5). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. ** Statistically significant increase over concurrent control (P ≤ 0.01). * Statistically significant increase over concurrent control (P ≤ 0.05). CHO positive with 20-h treatment. Lymphocytes positive with 24-h treatment at slightly higher doses.
cal range (99th percentiles) of 0.3–2.8). Strong induction of chromosome aberrations was seen in CHO cells after 3-h but not 20-h treatment (Fig. 5), but it was completely negative in HL, both in whole blood (data not shown) and Ficoll–Hypaque purified lymphocytes (Fig. 5). The frequencies of aberrations per cell were also high in CHO cells, and Compound K markedly inhibited DNA synthesis in CHO cells, e.g., without S-9 at 400–600 M BrdUrd uptake per S phase cells was 17 to 5% of controls. This drug class affects cell cycle checkpoints and apoptosis pathways, so the different results may reflect differences in these pathways between rodent and human cells, and/or between tumorigenic (CHO) and normal cells [8,9]. Two additional compounds with the same therapeutic target but different structures, Compound L and Compound M, gave weak positive and positive results, respectively, in the CHO aberration assay (Figs. 6 and 7) and both had dramatic increases in polyploidy (Figs. 6b and c and 7a and b). The weak positive result with Compound L was reproduced in a second experiment (data not shown). Neither of these compounds induced aberrations in lymphocytes, but mitotic indices at some dose levels were increased, and increases in polyploidy were significant (range of percent polyploidy cells from 0.4 to 7.0 compared to a concurrent control range of 0.0–0.4) but much lower than in CHO cells. 3.2. Compounds which induced aberrations but did not inhibit DNA synthesis in CHO cells Compound N (therapeutic class III) induced aberrations in CHO cells but was negative in vitro in the Ames
test and DNA strand break assay in rat hepatocytes (alkaline elution assay), and in vivo in the mouse bone marrow micronucleus assay and an alkaline elution assay in rat liver. A further assay for DNA damage in CHO cells was also negative, i.e., DNA adducts assessed by the 32 P post-labeling method. Moderate increases in aberrations were seen in CHO cells (Fig. 8) that were not S-9 dependent, appeared associated with growth suppression, and were suppressed in 50% human serum. The increase seen with a 3-h treatment was reproduced in two further experiments and the results after the 20h treatment were equivocal. Surprisingly, little or no inhibition of DNA synthesis was seen by flow cytometry, e.g. BrdUrd uptake per S phase cell was 87% of control at 50 M, a concentration that reduced cell growth to 46% of control and had 7.0% cells with aberrations. In lymphocytes, after the 3-h treatment there was little or no mitotic suppression so only the top dose, 150 M, was scored and was negative (1.5% cells with aberrations). After the 24-h treatment no significant increase in aberrations was seen at any doses up to the solubility limit even when doses were scored with mitotic indices suppressed to less than 50% of controls (Fig. 8). Compound O, in a different therapeutic and structural class (IV) was a potent inducer of aberrations in CHO cells especially after the 20-h treatment (Fig. 9). There was little cytotoxicity, and little or no suppression of DNA synthesis after either 3- or 20-h treatments up to 15 M (data not shown). This candidate was negative in the Ames test and the in vitro DNA strand break (alkaline elution) assay in rat hepatocytes. However its chromosome damaging ability
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
111
Fig. 5. The percentage of cells with chromosomal aberrations in cultures treated with Compound K (concurrent controls for CHO: 1.0–2.0; concurrent controls for lymphocytes: 0.0–0.5). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. ** Statistically significant increase over concurrent control (P ≤ 0.01). Lymphocyte assays conducted at Covance Laboratories. CHO positive not markedly affected by serum. Purified lymphocytes negative even with 22-h treatment. Whole blood was also negative (data not shown).
was confirmed in vivo in a bone marrow micronucleus assay in rats. Micronuclei were increased to 5.6 and 7.6 MN-PCE/1000 at 375 and 1000 mg/kg/day after a month of daily administration (controls were 2.5/1000). Compound O was clearly positive in the human lymphocyte aberration assay with 24-h treatment but did not induce significant increases of aberrations after the 3-h treatment.
3.3. Compound with multiple mechanisms of genotoxicity Compound P (therapeutic class V) was tested because there was thought to be more than one mechanism of genotoxicity. This compound induced aberrations in CHO cells at toxic doses associated with inhibition of DNA synthesis. It was negative in the Ames test, but
112
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
Fig. 6. The percentage of cells with chromosomal aberrations in cultures treated with Compound L (concurrent controls for CHO: 0.5–2.0; concurrent controls for lymphocytes: 1.0–1.5). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. ** Statistically significant increase over concurrent control (P ≤ 0.01). PP = percent polyploid cells. CHO weak positive with 20-h treatment; polyploidy seen. Lymphocytes negative in both 3- and 20-h treatments; weaker increase in polyploidy.
induced DNA strand breaks in the in vitro alkaline elution assay in rat hepatocytes, and also induced DNA adducts detectable by 32 P post-labeling in calf thymus DNA with S-9, in rat hepatocytes, and in vivo in rodent liver. In contrast, the in vivo alkaline elution assay in liver and the in vivo bone marrow micronucleus assays were negative. Adducts were also not detectable in bone marrow. It seemed likely that the aberration induction was associated with toxicity and DNA synthesis inhibition
rather than with DNA adducts, since no DNA adducts could be detected in CHO cells, and adduct induction was dependent upon S-9 activation, while aberrations were induced with and without S-9. This compound contained an indole in the structure; genotoxicity of naturally occurring indoles has been discussed in [10]. In parallel with the observations on Compound P, melatonin was shown to induce chromosome aberrations associated with toxicity and DNA synthesis inhibition, and aber-
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
113
Fig. 7. The percentage of cells with chromosomal aberrations in cultures treated with Compound M (concurrent controls for CHO: 1.0–4.5; concurrent controls for lymphocytes: 0.0–0.5). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. ** Statistically significant increase over concurrent control (P ≤ 0.01). PP = percent polyploid cells. CHO positive with 20-h treatment. Lymphocytes negative both 3- and 24-h treatments. Greatly increased polyploidy in CHO; smaller increase in polyploidy in lymphocytes. Increase in mitotic index as doses increase in lymphocytes.
rations were not S-9 dependent, whereas DNA adduct induction required metabolic activation [10]. In support of the role of cytotoxicity in aberration induction in CHO cells, we note that in our initial studies in CHO cells, we used cell counts expressed as % control to assess cytotoxicity. Subsequently we began using population doublings as a more appropriate measure of toxicity [6]. With Compound P, increases in aberrations (>25% cells with aberrations) were generally seen with cell counts between 49% and 59% of controls (Fig. 10a). In a later a study with growth suppression assessed based on population doublings, the only significant increase (30%) was seen at 250 M with no growth in the cultures (Fig. 10b). In lymphocytes, no increases in aberrations were seen at concentrations up to 350 M after 3-h treatments with or without S-9, but there was no consistent mitotic suppression (data not shown). The assay was repeated up to concentrations with visible precipitate, and weak increases in aberrations were seen. Fig. 10c and d shows the data for lymphocytes with-
out S-9. With S-9 the maximum levels of aberrations were 3.5 and 4.0%, which were not statistically significant compared with the concurrent control of 1.25% but higher than typical controls (data not shown). In a separate experiment, after a 24-h treatment without S-9 there was a small increase at 50 M (5.0% cells with aberrations) and a more convincing increase at 200 M at a more toxic concentration (mitotic index 34% of controls). Overall, the lymphocyte assay is weakly positive and confirms that substantial increases in aberrations are seen only with marked growth suppression. 3.4. DNA synthesis inhibitor, aphidicolin Aberration data for both CHO [2] and HL [4,5] treated with aphidicolin have been published, but the published lymphocyte studies used different protocols from the one we use now. Here we confirmed that aphidicolin induces aberrations under the standard test protocol conditions (3-h treatment of whole blood cultures without S-9), with
114
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
Fig. 8. The percentage of cells with chromosomal aberrations in cultures treated with Compound N (concurrent controls for CHO: 0.0–2.5; concurrent controls for lymphocytes: 1.0–1.5). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. ** Statistically significant increase over concurrent control (P ≤ 0.01). CHO positive with 3- and 20-h treatments, but not with human serum. Lymphocytes negative with 3- and 24-h treatments.
20.0–38.0% cells with aberrations in the range from 5 to 10 M, with mitotic indices of 49% down to 39% of controls respectively (data not shown). 4. Discussion Our results (summarized in Table 1) show seven cases where aberrations were induced in CHO cells, but negative or equivocal results were obtained in lymphocytes.
It is known that the presence of red blood cells or other blood components can affect the results of cytogenetics assays [11,12]. We examined whether purified lymphocytes might give different results from whole blood cultures for several of our test compounds, and did not find a difference in these cases. There appears to be an intrinsic difference between the cell types. With two compounds that induced aberrations in CHO cells and were weakly positive in the Ames test,
Table 1 Summary of genotoxicity profiles on the test compounds CHO Abs
HL Abs
DNA synthesis inhibition
Ames test
AE rat hepatocytes vitro/vivo
DNA adducts vitro/vivo
In vivo rodent micronucleus (bone marrow)
Compound G (Class I) Compound H (Class I) Compound I (Class I) Compound J (Class I) Compound K (Class II) Compound L (Class II) Compound M (Class II) Compound N (Class III)
Pos
Equivocal
Pos
Neg
Neg
Not done
Not done
Pos
Neg
Pos
Neg
Neg (neg TK6 assay)
Not done
Neg
Pos
Neg
Pos
Neg
Neg
Not done
Not done
Pos (20 h treatment–S9)
Pos
Pos
Weak Pos
Neg vitro
Negative in Salmonella
Not done
Pos
Neg
Pos
Weak Pos
Not done
Not done
Equivocal
Weak Pos (↑↑PP)
Neg (↑PP)
Not done
Not done
Not done
Not done
Not done
Pos (↑↑PP)
Neg (↑PP)
Not done
Not done
Not done
Not done
Not done
Pos
Neg
Neg
Neg
Neg in CHO cells
Neg
Compound O (Class IV)
Pos
Pos
Neg
Neg
Not done
Pos
Compound P (Class V)
Pos+− S9 Also pos human S9
Pos
Little or no DNA synthesis inhibition Little or no DNA synthesis inhibition Pos
Neg
Pos vitro, Neg vivo
Neg-
Aphidicolin
Pos
Pos
Pos
Not done
Not done
Pos, calf thymus DNA +S9, rat hepatocytes in vitro; in vivo liver pos but bone marrow neg Negative in CHO cells Not done
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
Compound (therapeutic class)
Not done
AE = alkaline elution assay for DNA strand breaks; PP = polyploidy.
115
116
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
Fig. 9. The percentage of cells with chromosomal aberrations in cultures treated with Compound O (concurrent controls for CHO: 0.5–1.5; concurrent controls for lymphocytes: 0.0–1.0). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. Note different scale of graph due to substantially higher increase in aberrations. ** Statistically significant increase over concurrent control (P ≤ 0.01). * Statistically significant increase over concurrent control (P ≤ 0.05). CHO positive with 3- and 20-h treatments. Lymphocytes positive with 24-h treatment.
one was negative in the lymphocyte aberration assay (K) and the other was weakly positive (J) when the mitotic index was ≥50% of controls (see below). Compound K, is known to affect tumor cells differently from normal cells, and the differential aberration induction in CHO cells and HL may be related to effects on cell cycle checkpoints and apoptosis.
Several of our candidates inhibited DNA synthesis strongly in CHO cells. Early studies that demonstrated the induction of aberrations by inhibitors of DNA synthesis used plants, but this was then demonstrated in normal human cells including lymphocytes [13]. Bender and Moore [5] pointed out that HL were relatively resistant to inhibition of DNA synthesis by aphidicolin compared
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
117
Fig. 10. The percentage of cells with chromosomal aberrations in cultures treated with Compound P (concurrent controls for CHO: 1.0–2.5; concurrent controls for lymphocytes: 1.0). Numbers on tops of columns are cell counts, taken at harvest, as percentages of controls (CHO cells, graph a) population doublings taken at harvest, as percentages of controls (CHO cells, graph b) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. No precipitate observed in CHO culture medium at doses scored. Precipitate observed in whole blood culture medium at 600 M for 3-h treatment. ** Statistically significant increase over concurrent control (P ≤ 0.01). CHO positive at similar doses with cell counts or population doubling (PD) as cytotoxicity assessment but toxicity seen to be much greater by PD method. Lymphocytes positive with 3- and 24-h treatments.
with transformed cells such as mouse L1210 and HeLa, requiring a 100 times higher aphidicolin concentration to achieve the same degree of inhibition of 3 HTdR uptake. However, we found here that aberrations were induced in lymphocytes in the same dose range of aphidicolin that induces aberrations in CHO cells. It is likely that other factors play into the differential aberration induction with some compounds in lymphocytes compared with CHO cells, including more stringent checkpoint controls in human cells that may be more likely to prevent damaged cells reaching mitosis compared with rodent cells [8,9]. Differences in checkpoint control may also be related to the greater sensitivity to induction of aneuploidy and polyploidy of rodent compared with human cells. Tsutsui et al. [14] showed that normal early passage Syrian hamster fibroblasts displayed more aneuploidy and polyploidy than normal human fibroblasts when treated with classical spindle poisons or with the estrogenic compounds 17- estradiol and diethylstilbestrol. We have
also found previously that two compounds that induced high levels of polyploidy (but not structural aberrations) in CHO cells were completely negative in early passage normal human fibroblasts (unpublished data). In the current experiments, two candidates that strongly induced polyploidy in CHO cells, L and M, gave weaker (but still significant) increases in polyploidy in the human lymphocytes. Lymphocytes are sensitive to aberration induction by classical DNA damaging agents ranging from ionizing radiation to alkylating agents, and to clastogens such as topisomerase inhibitors. We found here that a potent aberration inducer (Compound O), which also induced micronuclei in vivo (although it was negative in the Ames test) was also clearly positive in the lymphocyte aberration assay. Two compounds were found to reach convincing aberration levels in the lymphocyte assay only when the mitotic indices were suppressed below 50% of control, Compound J and Compound P. One of these, Compound
118
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
J, is weakly positive in the Ames test and the second, Compound P, although Ames test negative, is capable of inducing DNA strand breaks and DNA adducts, but it appears likely that aberration induction is related more to toxicity and inhibition of DNA synthesis than it is to primary DNA damage. The studies here indicate that for this limited set of compounds, lymphocytes are more likely to give negative or equivocal results for compounds that are positive in the CHO assay associated with toxicity. We note that two compounds that are weakly positive in the Ames test (Compound K and Compound J) and one that induced DNA adducts were negative or equivocal in the lymphocyte aberration assay; it is not clear what the overall mutagenic potential in vivo is for these, and no carcinogenicity data are available. When “pushed” to doses that decrease the mitotic index by more than half compared with controls, additional positive results may also occur in lymphocytes. These studies again pinpoint the unsatisfactory nature of mitotic index as a measure of cyotoxicity/cell growth. The present experiments also demonstrate once again that mitotic index can actually increase at a certain times after treatment. The need to find suitable donors, variability among individuals, and increased time required for the assay along with the inadequacy of the MI as a toxicity measure make the lymphocyte assay less attractive for routine screening. However, studies in lymphocytes or other normal human cells may play a useful part in the follow-up strategy to assess weight of evidence about genotoxic risk for humans for compounds that are positive in the Chinese hamster cell aberration assays. Acknowledgement The blood lymphocyte assays on Compound K were conducted at Covance Laboratories.
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
References [14] [1] S. Galloway, Cytotoxicity and chromosome aberrations in vitro: experience in industry and the case for an upper limit on toxicity
in the aberration assay, Environ. Mol. Mutag. 35 (2000) 191– 201. S. Galloway, J. Miller, M. Armstrong, C. Bean, T. Skopek, W. Nichols, DNA synthesis inhibition as an indirect mechanism of chromosome aberrations: comparison of DNA-reactive and non-DNA-reactive clastogens, Mutat. Res. 400 (1998) 169– 186. C. Hilliard, M. Armstrong, C. Bradt, R. Hill, S. Greenwood, S. Galloway, Chromosome aberrations in vitro related to cytotoxicity of nonmutagenic chemicals and metabolic poisons, Environ. Mol. Mutag. 31 (1998) 316–326. M. Bender, J. Preston, Role of base damage in aberration formation interactions of aphidicolin and X-rays, Progr. Mutat. Res. 4 (1982) 37–46. M. Bender, R. Moore, Dose relationships for different effects of aphidicolin in human peripheral blood leukocytes, Mutat. Res. 198 (1988) 227–231. S. Greenwood, R. Hill, J. Sun, M. Armstrong, T. Johnson, J. Gara, S. Galloway, Population doubling: a simple and more accurate estimation of cell growth suppression in the in vitro assay for chromosomal aberrations that reduces irrelevant positive results, Environ. Mol. Mutag. 43 (1) (2004) 36–44. C. Dunnett, A multiple comparison procedure for comparing several treatments with a control, J. Am. Stat. Assoc. 50 (1955) 1096–1121. A. Kung, S. Sherwood, R. Schimke, Cell line-specific differences in the control of cell cycle progression in the absence of mitosis, Proc. Natl. Acad. Sci. 87 (1990) 9553–9557. R. Schimke, A. Kung, D. Rush, S. Sherwood, Differences in Mitotic Control Among Mammalian Cells, in: Cold Spring Harbor Symposia on Quantitative Biology Vol LVI, 1991, p. 417. V. Reddy, R. Storer, G. Laws, M. Armstrong, J. Barnum, J. Gara, C. McKnight, T. Skopek, J. Sina, J. DeLuca, S. Galloway, Genotoxicity of naturally occurring indole compounds: correlation between covalent DNA binding and other genotoxicity tests, Environ. Mol. Mutag. 40 (1) (2002) 1–17. J. Wilmer, G. Erexson, A. Kligerman, Implication of an elevated sister-chromatid exchange frequency in rat lymphocytes cultured in the absences of erythrocytes, Mutat. Res. 109 (1983) 231–248. K. Mehnert, R. During, W. Vogel, G. Speit, Differences in the induction of SCEs between human whole blood cultures and purified lymphocyte cultures and the effect of an S-9 mix, Mutat. Res. 130 (1984) 403–410. B. Kihlman, W. Nichols, A. Levan, The effect of deoxyadenosine and cytosine arabinoside on the chromosomes of human leukocytes in vitro, Hereditas 50 (1963) 139–143. T. Tsutsui, N. Suzuki, H. Maizumi, C. Carrett, Aneuploidy induction in human fibroblasts: comparison with results in Syrian hamster fibroblasts, Mutat. Res. 240 (1990) 241–249.
Chromosome aberrations in Chinese hamster and human cells: A comparison using compounds with various genotoxicity profiles Cathy Hilliard ∗ , Rosina Hill, Michael Armstrong, Clint Fleckenstein, Jessica Crowley, Elizabeth Freeland, Danielle Duffy, Sheila M. Galloway Merck Research Laboratories, Department of Genetic and Cellular Toxicology, West Point, PA 19486, United States Available online 14 December 2006
Abstract Chromosome aberrations (Cabs) can be induced in vitro by non-DNA damaging compounds, often associated with cytotoxicity and DNA synthesis inhibition, and under conditions that would not be relevant in vivo. Such misleading positive results are reported both in Chinese hamster cell lines and in human peripheral blood lymphocytes (HL). We assessed the response of HL to compounds with varied genetic toxicity profiles, all of which induced Cabs in CHO cells Seven of 10 compounds were negative or equivocal in HL. Results in purified lymphocytes for four verified that the difference was not due to the presence of blood in cultures. Two compounds that were weakly positive in the Ames test and one that induced DNA adducts were negative or equivocal in the HL assay; their overall mutagenic potential in vivo is not clear. Of four Ames-negative compounds, three of which inhibited DNA synthesis in CHO cells, three were negative and one was equivocal in the HL assay. A potent Cab inducer, which also induced micronuclei in vivo (but was negative in the Ames test) was clearly positive in the HL assay. Two compounds were clearly positive in HL only when the mitotic indices (MI) were below 50% of control. These are genotoxic in other assays but our evidence suggests that Cab induction is related more to toxicity than to primary DNA damage. For this limited set of 10 compounds, HL were more likely than CHO cells to give negative or equivocal results. It is likely that more stringent checkpoint controls in human cells prevent damaged cells reaching mitosis, and may also influence the reported greater sensitivity to induction of aneuploidy and polyploidy of normal rodent compared with human cells. In the studies reported here, two strong inducers of polyploidy in CHO cells gave weaker increases in HL. Human lymphocytes have disadvantages as a routine screening assay (finding donors, known individual variability, increased time required and the inadequacy of the MI as a toxicity measure), but may be useful in follow-up testing to assess weight of evidence about genotoxic risk to humans, for compounds that are positive in the Chinese hamster cell Cabs assays. © 2006 Elsevier B.V. All rights reserved. Keywords: Chromosome aberrations; Genotoxicity testing; Human lymphocytes; Hamster cells; Indirect mechanisms
1. Introduction It is now widely recognized that in vitro assays for chromosome aberrations may give misleading positive results, especially when testing extends into the toxic ∗ Corresponding author. Tel.: +1 215 652 5666; fax: +1 215 652 4944. E-mail address: catherine [email protected] (C. Hilliard).
0027-5107/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.mrfmmm.2006.11.013
dose range. In many cases the evidence indicates these positive results occur without direct attack on DNA, and under conditions that would not be relevant in vivo. Such positive results are reported both in permanent hamster cell lines such as CHO and CHL, and also in cultured human peripheral blood lymphocytes (HL) [1]. We have found that chromosome aberrations can be induced by non-DNA damaging compounds, often associated with cytotoxicity and inhibition of DNA synthesis
104
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
[2,3]. Since the polymerase inhibitor aphidicolin induces chromosome aberrations in normal human cells [4,5], we were interested in assessing the response of lymphocytes to drug candidates that induced aberrations in CHO cells associated with inhibition of DNA synthesis. We extended the comparison to compounds with a variety of genotoxicity profiles encountered in drug development, including those with weak positive Ames test results, compounds that induced polyploidy in CHO cells, compounds that induced aberrations at toxic doses in CHO cells but did not inhibit DNA synthesis, and compounds thought to have more than one mechanism of genotoxicity. Since some drug candidates are known to be highly protein bound, and we had demonstrated marked shifts in the dose response to higher concentrations when we increased the serum concentrations in the treatment medium from 10% to, e.g., 50% serum, some experiments were done to compare results at different serum concentrations. Similarly, in certain cases experiments were done both with whole blood cultures and with purified lymphocytes, to affirm that a difference in results between CHO cells and lymphocytes was due to cell type difference, and not interference from blood components. 2. Materials and methods 2.1. Test chemicals and solvents All compounds were from Merck & Co., Inc. except for aphidicolin, which was from Sigma Chemical Company (St. Louis, MO, USA). All stocks were prepared fresh for each experiment. All compounds tested, including aphidicolin, were prepared as 100× concentrations to give a final concentration of 1% DMSO. DMSO was from Sigma Chemical Company. Results are presented with compounds grouped by therapeutic class because typically the molecules in each group share structural features, and knowledge of genotoxicity profiles of additional related compounds can in some cases inform the discussion of the overall genotoxicity profile. 2.2. Cell culture and treatment 2.2.1. CHO cells Chinese hamster ovary (CHO) cells, clone WBL, were cultured in a humidified atmosphere of 5% CO2 in air at 37 ◦ C in McCoy’s 5A medium (Invitrogen, Grand Island, New York, USA), supplemented with 10% fetal bovine serum (Hyclone, Logan, Utah, USA), 2 mM l-glutamine, and 100 U/mL penicillin, 100 U/mL streptomycin (all Invitrogen). Compounds G, H, I, and N were also tested using 50% pooled human serum (type AB, from SeraCare Inc., Oceanside, CA, USA). Cells were seeded the day before treatment at 1.2 × 106 /10 mL/75-
cm2 flask (Corning, Corning, NY, USA). Cultures treated with the metabolic activation system were re-fed with serum-free S-9 mix just before treatment. Test compounds were dissolved in DMSO or dH2 O, added to cultures, and the cultures were incubated at 37 ◦ C. Treatments were for either 3 or 20 h, and aberrations were scored in cells harvested 20 h from the beginning of treatment. At the termination of the 3-h treatments, the cultures were washed twice with Hanks Balanced Salt Solution (HBSS) (Invitrogen) and re-fed with fresh complete medium. Cultures were incubated for a further 17 h. Colcemid (0.1 g/mL; Invitrogen) was added 1–2 h before monolayers were harvested by trypsinization. A cell sample was counted by Coulter counter (Beckman Coulter, Inc., Miami, FL, USA) to determine cell number as an indicator of cytotoxicity. (We verified that these counts represented viable cells by checking samples for trypan blue dye exclusion.) The cells were treated with hypotonic KCl (75 mM; Sigma Chemical Company) for 1–3 min at room temperature, washed twice with fixative (methanol:glacial acetic acid, 3:1, v/v), dropped onto slides, air-dried, and stained with Gurr Giemsa stain (BDH Chemicals LTD, Poole, England, UK). 2.2.2. Lymphocytes Human blood obtained from healthy adult donors (aged 21–45, nonsmokers without a history of radiotherapy or chemotherapy and lacking current viral infections) was cultured in a humidified atmosphere of 5% CO2 in air at 37 ◦ C in RPMI 1640 medium (Invitrogen), supplemented with 15% fetal bovine serum, 2 mM l-glutamine, 100 U/mL penicillin, 100 U/mL streptomycin, and 2% phytohemaglutanin M (all Invitrogen). Whole blood cultures were set up approximately 48 h before treatment, with 1 mL blood per 9 mL medium in standing 25-cm2 flasks (Corning). Purified lymphocyte cultures were obtained either by Ficoll gradient (Compound K only) or by buffy coat separation. For Ficoll gradient preparations, whole blood was diluted four-fold with phosphate-buffered saline (PBS; Invitrogen, Carlsbad, CA, USA) and layered over Ficoll–Hypaque (Mediatech, Inc., Herndon, VA, USA). After centrifugation at 400 × g for 40 min, the lymphocyte layer was removed and washed twice with PBS containing 3–5% fetal bovine serum. For preparation by buffy coat separation, whole blood was allowed to settle in sodium heparin tubes for approximately 3 h at 37 ◦ C before removing the white cell layer, which was spun down and resuspended in fresh complete medium. For both methods, cell suspensions were seeded at approximately 2.0 × 106 /10 mL/25-cm2 flasks (total white blood cells) and incubated for 48 h before dosing. Cultures treated with the metabolic activation system were spun and resuspended in serum-free medium containing S-9 mix before treatment (cultures used for Compound K were resuspended in medium containing 3–5% serum and S-9 mix). Test compounds were dissolved in DMSO or dH2 O, added to cultures, and the cultures were incubated at 37 ◦ C for either 3 or 24 h.
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
Treatments were for either 3 or 24 h, and aberrations were scored in cells harvested 24 h from the beginning of treatment. (The 24-h treatment length, compared with 20 h for CHO cells, was chosen to allow for the slightly longer average cell cycle length in cultured lymphocytes.) At the termination of the 3-h treatments, the cultures were centrifuged at about 225 × g for 5 min, washed once with HBSS, and re-fed with fresh complete medium. Cultures were incubated for a further 21 h. Colcemid (0.1 g/mL) was added 2–3 h before cultures were harvested. The cells were treated with hypotonic KCl (75 mM) for 10–15 min at 37 ◦ C, washed twice with fixative (methanol:glacial acetic acid, 3:1 v/v), dropped onto slides, air-dried, and stained with Gurr Giemsa stain. 2.3. Metabolic activation system Compounds which gave positive results in CHO cells only with S-9 metabolic activation, or gave more substantial increases with S-9, were tested in HL with S-9 activation. Otherwise, HL were tested only without S-9. S-9 was produced using liver homogenate (S-9 fraction) from phenobarbital/βnapthoflavone-treated male Sprague–Dawley rats (Charles River Laboratories, Raleigh, NC, USA). The S-9 used for these experiments was purchased from MolTox (Boone, NC, USA). S-9 was stored at −70 to −80 ◦ C and thawed immediately before use. S-9 was mixed with sodium NADP (Boehringer Mannheim, Indianapolis, IN, USA), and trisodium isocitrate (Sigma Chemical Company) in serum-free culture medium (culture medium with 3% serum used for Compound K) to give final concentrations of: S-9, 15 L/mL; NADP, 1.05 mM; and trisodium isocitrate, 5.8 mM. 2.4. Dose selection The maximum dose tested was 10 mM. Cytotoxicity was used to select doses, scoring doses that gave a cell growth reduction not greatly exceeding 50% compared with concurrent controls. For CHO cells toxicity was assessed as reductions in population doubling [6] except for Compound P for which reductions in cell counts were used. For HL mitotic index was used. In cases where there was little or no cytotoxicity, doses scored were those that did not produce excessive precipitate in culture medium.
105
of mitotic index and 500 metaphases for the frequency of polyploidy. 2.6. Criteria for a positive result A positive point is one where the percentage of cells with aberrations is significantly greater than the concurrent controls (P ≤ 0.05), and outside the historical control range. The analysis is a pairwise comparison of the percentage of cells with aberrations at each dose level with the controls by Fisher’s exact test, with the P values adjusted for multiple comparisons against a common control by the method of Dunnett [7]. We generally consider a test positive if there are two positive points, or if a single positive point is reproduced in an independent assay. Evidence for a dose relation and increased frequencies of aberrations per cell (presence of cells with more than one aberration) are also taken into account. Although our top dose is usually selected such that growth reduction does not greatly exceed 50%, for the purposes of the current experiments dose levels with mitotic index of less than 50% of controls were included to see if aberrations were induced only with marked growth suppression. 2.7. Flow cytometry for detecting DNA synthesis inhibition Chinese hamster ovary (CHO) cells were treated in parallel cultures to those used in the aberration assay with 0.1 M 5bromodeoxyuridine (BrdUrd; Sigma Chemical Company) for the last 30 min of incubation with test compound. After removal of test compound, cultures were immediately harvested with trypsin. Cells were processed as described in [2], and stained for BrdUrd detection (FITC labeled secondary antibody to mouse anti-BrdUrd) and total cell DNA (propidium iodide, PI) using the iU-4 anti-BrdUrd monoclonal antibody. Analysis of these dual stained CHO cells was performed using an EPICS Elite ESP flow cytometer (Coulter Electronics, Hialeah, FL), equipped with an Innova 90-6 argon-ion laser (Coherent, Inc., Palo Alto, CA) tuned to 488 nm to excite the two fluorochromes (FITC and PI). Two parameter BrdUrd-FITC/DNA-PI histograms were constructed using a linear scale on the x-axis to display the DNA content (PI, red integrated fluorescence [IRFL] and a logarithmic scale on the y-axis to display BrdUrd incorporation (FITC, green integrated fluorescence [LIGFL]. Information from 104 cells was collected for each sample.
2.5. Chromosome aberration scoring Slides for each compound were scored under code by one observer. For each treatment, 200 metaphase cells containing 19–23 (CHO) or 44–48 (HL) chromosomes were scored, unless aberration frequencies were high or too few cells were available for analysis due to mitotic suppression. Gaps (achromatic region less than or equal to the width of a chromatid), polyploidy, and endoreduplicated cells, and pulverized chromosomes were noted but not included in structural aberration totals. Five hundred cells were scored for assessment
2.7.1. Relative BrdUrd incorporation Inhibition was measured as suppression of BrdUrd uptake, i.e. decreased green fluorescence intensity. The relative quantity of BrdUrd incorporated per S phase cell was calculated from the difference in the mean fluorescence intensity (channel number) between the BrdUrd-positive cell population (S phase cells) and the BrdUrd-negative population (G1 , G2 , and M cells). This number calculated for each treated culture was divided by the corresponding number for the controls to give fluorescent intensity as a percentage of control.
106
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
3. Results 3.1. Compounds that induced aberrations and inhibited DNA synthesis in CHO cells 3.1.1. Therapeutic class I 3.1.1.1. Ames test negative compounds. This set of compounds is from a therapeutic class in which the first candidate (data not shown) was positive in the CHO cell chromosome aberration assay but negative in vitro in Ames, hepatocyte DNA stand break (alkaline elution) and TK6 mammalian cell tk mutation assays, and in vivo in the mouse bone marrow micronucleus assay. This compound was also known to be highly proteinbound, and the dose response for aberration induction was markedly shifted to higher concentrations (up to 20fold) when the serum content of the treatment medium was increased from 0 to 50% human serum. A series of related compounds was subsequently tested for aberrations in CHO cells, and three that induced aberrations and DNA synthesis inhibition but were negative in the Ames test are described here. Results reported without serum (FBS) are from incubations with S-9 mix in serum-free medium unless noted otherwise. Compound G induced statistically significant increases in aberrations in CHO cells after 3-h treatments (both with and without S-9, Fig. 1) and 20-h treatment (data not shown) at dose levels between 100 and 1000 M. Aberration induction was not dependent on S-9 activation. Adding 50% human serum instead of the routine 10% FBS shifted the active concentration range to slightly higher doses. The high frequencies of aberrations (numbers in the columns) indicate the presence of some cells with multiple aberrations, a common finding with compounds that inhibit DNA synthesis. Compound G inhibited DNA synthesis dramatically. For example, BrdUrd uptake was suppressed to about 7% of controls at the end of a 3-h treatment with 100 M Compound G without serum, or 500 M with 50% serum. In whole blood, in a first experiment mitotic suppression of 50% or greater was achieved, and there was no increase in aberrations up to 800 M without serum (without S-9, data not shown). With 15% serum (standard treatment conditions for lymphocytes) the value of 2.5% cells with aberrations at 2000 M (Fig. 1e) was statistically significant compared with the concurrent control of 0.25%, but is not considered meaningful compared with historical control data (historical control mean: 0.7; range: 0.0–2.0), and precipitate of drug was visible during treatment. Also, the mitotic index was only 37% of control. In a repeat experiment without FBS (Fig. 1d) a weak but
statistically significant increase in aberrations was seen at 2000 M (6.0% cells with aberrations with a mitotic index at 52% of controls), and again this exceeded the solubility limit for the drug. An experiment repeated with purified lymphocytes was negative at doses up to 2000 M (with mitotic index at 38% of controls). The overall conclusion for lymphocytes for Compound G is equivocal, and most likely negative under standard testing conditions. The cells with multiple aberrations which were seen in CHO cultures were not seen in lymphocytes. Compound H induced weak but statistically significant increases in aberrations in CHO cells after 3-h treatments both with and without S-9 (Fig. 2). Some cells had multiple aberrations. Twenty-hour treatments induced significant increases in aberrations with 10% fetal bovine serum (7.0% and 6.0% cells with aberrations at 15 and 20 M, respectively). Increasing the concentration of serum in the culture medium to 50% human serum suppressed aberration induction, with no significant increases at up to 800 M (Fig. 2c). DNA synthesis was suppressed, the greatest suppression being to about 54% of controls. Levels of DNA synthesis about 60% of controls were seen at 60–100 M without serum, and at 400–600 M with 50% serum. Compound H was tested for micronuclei in vivo and was negative. In HL there were no increases in aberrations either in whole blood cultures with or without serum or in purified lymphocytes (Fig. 2). The top doses scored were limited by the appearance of precipitate in the culture medium in addition to mitotic index suppression. Compound I induced statistically significant increases in aberrations in CHO cells after 3-h treatments both with and without S-9 (Fig. 3). There was no increase in the percentage of cells with aberrations up to 20 M after a 20-h treatment without S-9 (data not shown) but there was one cell with multiple aberrations. Increased serum (50% human serum) shifted the dose response but results were still clearly positive (Fig. 3c). DNA synthesis was markedly suppressed, to less than 5% of controls at the higher doses, about 120 M without serum and about 500 M with 50% serum. In whole blood cultures, 2% cells with aberrations were seen at 200 and 250 M, compared with 0% in controls and there was no mitotic suppression after a 3-h treatment in normal serum (15% FBS) up to 250 M, a dose with visible precipitate (Fig. 3). A repeat experiment confirmed that this result was negative (data not shown) and the assay in purified lymphocytes (buffy coat) was also negative since the value of 4.0% aberrations was not significantly higher than the control value of 1.5% (Fig. 3f).
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
107
Fig. 1. The percentage of cells with chromosomal aberrations in cultures treated with Compound G (concurrent controls for CHO: 0.5–2.5; concurrent controls for lymphocytes: 0.0–1.0). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. No precipitate observed in CHO culture medium at doses scored; precipitate observed at 2000 M and above in lymphocyte culture medium. For lymphocytes without FBS, 800 M was from a separate experiment. ** Statistically significant increase over concurrent control (P ≤ 0.01). * Statistically significant increase over concurrent control (P ≤ 0.05). CHO positive not markedly affected by serum (even 50% human serum). Very weak aberration induction in lymphocytes when compared with CHO even at doses with marked mitotic suppression.
3.1.1.2. Ames test positive compound. The fourth compound in this class, Compound J, differed from the first three in that it was weakly positive in the Ames test, giving a reproducible increase of two- to three-fold in revertants in Salmonella TA1535 at 3000–6000 g/plate. (Our background revertant counts were about 21–27 per plate). No DNA adducts could be detected by 32 P post-labeling in the Salmonella, and the compound was
negative in the in vitro DNA strand break assay in rat hepatocytes (alkaline elution assay.) Thus it is not clear whether DNA damage is a factor in the induction of aberrations seen reproducibly in CHO cells after a 20-h treatment (Fig. 4). Aberrations were not seen after the 3-h treatments, but concentrations with about 50% toxicity were not obtained in that protocol (data not shown). Compound J also inhibited DNA synthesis after 20-h
108
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
Fig. 2. The percentage of cells with chromosomal aberrations in cultures treated with Compound H (concurrent controls for CHO: 0.5–3.0; concurrent controls for lymphocytes: 0.0–2.0). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. Precipitate observed at 80 M and above without FBS and at 600 M with 10% FBS in CHO culture medium. Precipitate observed at 100 M and above without FBS and at 400 M and above with FBS in lymphocyte culture medium. ** Statistically significant increase over concurrent control (P ≤ 0.01). CHO positive suppressed by higher serum concentrations. Purified lymphocytes and whole blood negative.
treatment, with suppression to about 11% of controls at 1000 M. In lymphocytes, after a 3-h treatment there was moderate mitotic suppression and no aberration induction (data not shown), and after a 24-h treatment there were weak but significant increases in aberrations at 750 M (3.0%) and 1000 M (5.0%) with mitotic indices suppressed to 44 and 52% of controls. To see
how this related to toxicity, aberrations were scored at a dose, 1500 M, which suppressed the mitotic index to 16% of control (about 1% compared with controls of about 6–9%). In two experiments, substantial increases in aberrations were seen, to 8.5% (Fig. 4), and to 21% (data not shown). There were also increased frequencies of aberrations per cell. Overall, this compound was pos-
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
109
Fig. 3. The percentage of cells with chromosomal aberrations in cultures treated with Compound I (concurrent controls for CHO: 1.0–3.0; concurrent controls for lymphocytes: 0.0–2.0). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. No precipitate observed in CHO culture medium at doses scored. Precipitate observed in whole blood lymphocyte culture medium with FBS at 250 M. ** Statistically significant increase over concurrent control (P ≤ 0.01). CHO positive found at higher doses with higher concentrations of serum. Purified lymphocytes and whole blood negative.
itive in the lymphocyte assay but the result was weaker than that seen in CHO cells when toxicity in HL was limited to about 50% mitotic suppression. 3.1.2. Therapeutic class II The first compound in this therapeutic class, an anticancer agent, Compound K, gave borderline positive results in the Ames test (reproducible two-fold increases that shifted from TA100 to TA97a in repeat assays with
a pre-incubation protocol, and were not S-9 dependent). The structure suggests it may have some intrinsic genotoxicity. The result of the in vivo mouse bone marrow micronucleus assay was equivocal, since the levels of micronucleated PCE (2.6–2.7 MN-PCE/1000 PCE) after single oral administrations of 1000 and 2000 mg/kg were significant compared with the concurrent control mean of about 1.0 MN-PCE/1000, but within our historical control range (mean about 1.5 MN-PCE/1000 PCE, typi-
110
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
Fig. 4. The percentage of cells with chromosomal aberrations in cultures treated with Compound J (concurrent controls for CHO: 1.0–2.0; concurrent controls for lymphocytes: 0.0–0.5). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. ** Statistically significant increase over concurrent control (P ≤ 0.01). * Statistically significant increase over concurrent control (P ≤ 0.05). CHO positive with 20-h treatment. Lymphocytes positive with 24-h treatment at slightly higher doses.
cal range (99th percentiles) of 0.3–2.8). Strong induction of chromosome aberrations was seen in CHO cells after 3-h but not 20-h treatment (Fig. 5), but it was completely negative in HL, both in whole blood (data not shown) and Ficoll–Hypaque purified lymphocytes (Fig. 5). The frequencies of aberrations per cell were also high in CHO cells, and Compound K markedly inhibited DNA synthesis in CHO cells, e.g., without S-9 at 400–600 M BrdUrd uptake per S phase cells was 17 to 5% of controls. This drug class affects cell cycle checkpoints and apoptosis pathways, so the different results may reflect differences in these pathways between rodent and human cells, and/or between tumorigenic (CHO) and normal cells [8,9]. Two additional compounds with the same therapeutic target but different structures, Compound L and Compound M, gave weak positive and positive results, respectively, in the CHO aberration assay (Figs. 6 and 7) and both had dramatic increases in polyploidy (Figs. 6b and c and 7a and b). The weak positive result with Compound L was reproduced in a second experiment (data not shown). Neither of these compounds induced aberrations in lymphocytes, but mitotic indices at some dose levels were increased, and increases in polyploidy were significant (range of percent polyploidy cells from 0.4 to 7.0 compared to a concurrent control range of 0.0–0.4) but much lower than in CHO cells. 3.2. Compounds which induced aberrations but did not inhibit DNA synthesis in CHO cells Compound N (therapeutic class III) induced aberrations in CHO cells but was negative in vitro in the Ames
test and DNA strand break assay in rat hepatocytes (alkaline elution assay), and in vivo in the mouse bone marrow micronucleus assay and an alkaline elution assay in rat liver. A further assay for DNA damage in CHO cells was also negative, i.e., DNA adducts assessed by the 32 P post-labeling method. Moderate increases in aberrations were seen in CHO cells (Fig. 8) that were not S-9 dependent, appeared associated with growth suppression, and were suppressed in 50% human serum. The increase seen with a 3-h treatment was reproduced in two further experiments and the results after the 20h treatment were equivocal. Surprisingly, little or no inhibition of DNA synthesis was seen by flow cytometry, e.g. BrdUrd uptake per S phase cell was 87% of control at 50 M, a concentration that reduced cell growth to 46% of control and had 7.0% cells with aberrations. In lymphocytes, after the 3-h treatment there was little or no mitotic suppression so only the top dose, 150 M, was scored and was negative (1.5% cells with aberrations). After the 24-h treatment no significant increase in aberrations was seen at any doses up to the solubility limit even when doses were scored with mitotic indices suppressed to less than 50% of controls (Fig. 8). Compound O, in a different therapeutic and structural class (IV) was a potent inducer of aberrations in CHO cells especially after the 20-h treatment (Fig. 9). There was little cytotoxicity, and little or no suppression of DNA synthesis after either 3- or 20-h treatments up to 15 M (data not shown). This candidate was negative in the Ames test and the in vitro DNA strand break (alkaline elution) assay in rat hepatocytes. However its chromosome damaging ability
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
111
Fig. 5. The percentage of cells with chromosomal aberrations in cultures treated with Compound K (concurrent controls for CHO: 1.0–2.0; concurrent controls for lymphocytes: 0.0–0.5). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. ** Statistically significant increase over concurrent control (P ≤ 0.01). Lymphocyte assays conducted at Covance Laboratories. CHO positive not markedly affected by serum. Purified lymphocytes negative even with 22-h treatment. Whole blood was also negative (data not shown).
was confirmed in vivo in a bone marrow micronucleus assay in rats. Micronuclei were increased to 5.6 and 7.6 MN-PCE/1000 at 375 and 1000 mg/kg/day after a month of daily administration (controls were 2.5/1000). Compound O was clearly positive in the human lymphocyte aberration assay with 24-h treatment but did not induce significant increases of aberrations after the 3-h treatment.
3.3. Compound with multiple mechanisms of genotoxicity Compound P (therapeutic class V) was tested because there was thought to be more than one mechanism of genotoxicity. This compound induced aberrations in CHO cells at toxic doses associated with inhibition of DNA synthesis. It was negative in the Ames test, but
112
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
Fig. 6. The percentage of cells with chromosomal aberrations in cultures treated with Compound L (concurrent controls for CHO: 0.5–2.0; concurrent controls for lymphocytes: 1.0–1.5). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. ** Statistically significant increase over concurrent control (P ≤ 0.01). PP = percent polyploid cells. CHO weak positive with 20-h treatment; polyploidy seen. Lymphocytes negative in both 3- and 20-h treatments; weaker increase in polyploidy.
induced DNA strand breaks in the in vitro alkaline elution assay in rat hepatocytes, and also induced DNA adducts detectable by 32 P post-labeling in calf thymus DNA with S-9, in rat hepatocytes, and in vivo in rodent liver. In contrast, the in vivo alkaline elution assay in liver and the in vivo bone marrow micronucleus assays were negative. Adducts were also not detectable in bone marrow. It seemed likely that the aberration induction was associated with toxicity and DNA synthesis inhibition
rather than with DNA adducts, since no DNA adducts could be detected in CHO cells, and adduct induction was dependent upon S-9 activation, while aberrations were induced with and without S-9. This compound contained an indole in the structure; genotoxicity of naturally occurring indoles has been discussed in [10]. In parallel with the observations on Compound P, melatonin was shown to induce chromosome aberrations associated with toxicity and DNA synthesis inhibition, and aber-
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
113
Fig. 7. The percentage of cells with chromosomal aberrations in cultures treated with Compound M (concurrent controls for CHO: 1.0–4.5; concurrent controls for lymphocytes: 0.0–0.5). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. ** Statistically significant increase over concurrent control (P ≤ 0.01). PP = percent polyploid cells. CHO positive with 20-h treatment. Lymphocytes negative both 3- and 24-h treatments. Greatly increased polyploidy in CHO; smaller increase in polyploidy in lymphocytes. Increase in mitotic index as doses increase in lymphocytes.
rations were not S-9 dependent, whereas DNA adduct induction required metabolic activation [10]. In support of the role of cytotoxicity in aberration induction in CHO cells, we note that in our initial studies in CHO cells, we used cell counts expressed as % control to assess cytotoxicity. Subsequently we began using population doublings as a more appropriate measure of toxicity [6]. With Compound P, increases in aberrations (>25% cells with aberrations) were generally seen with cell counts between 49% and 59% of controls (Fig. 10a). In a later a study with growth suppression assessed based on population doublings, the only significant increase (30%) was seen at 250 M with no growth in the cultures (Fig. 10b). In lymphocytes, no increases in aberrations were seen at concentrations up to 350 M after 3-h treatments with or without S-9, but there was no consistent mitotic suppression (data not shown). The assay was repeated up to concentrations with visible precipitate, and weak increases in aberrations were seen. Fig. 10c and d shows the data for lymphocytes with-
out S-9. With S-9 the maximum levels of aberrations were 3.5 and 4.0%, which were not statistically significant compared with the concurrent control of 1.25% but higher than typical controls (data not shown). In a separate experiment, after a 24-h treatment without S-9 there was a small increase at 50 M (5.0% cells with aberrations) and a more convincing increase at 200 M at a more toxic concentration (mitotic index 34% of controls). Overall, the lymphocyte assay is weakly positive and confirms that substantial increases in aberrations are seen only with marked growth suppression. 3.4. DNA synthesis inhibitor, aphidicolin Aberration data for both CHO [2] and HL [4,5] treated with aphidicolin have been published, but the published lymphocyte studies used different protocols from the one we use now. Here we confirmed that aphidicolin induces aberrations under the standard test protocol conditions (3-h treatment of whole blood cultures without S-9), with
114
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
Fig. 8. The percentage of cells with chromosomal aberrations in cultures treated with Compound N (concurrent controls for CHO: 0.0–2.5; concurrent controls for lymphocytes: 1.0–1.5). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. ** Statistically significant increase over concurrent control (P ≤ 0.01). CHO positive with 3- and 20-h treatments, but not with human serum. Lymphocytes negative with 3- and 24-h treatments.
20.0–38.0% cells with aberrations in the range from 5 to 10 M, with mitotic indices of 49% down to 39% of controls respectively (data not shown). 4. Discussion Our results (summarized in Table 1) show seven cases where aberrations were induced in CHO cells, but negative or equivocal results were obtained in lymphocytes.
It is known that the presence of red blood cells or other blood components can affect the results of cytogenetics assays [11,12]. We examined whether purified lymphocytes might give different results from whole blood cultures for several of our test compounds, and did not find a difference in these cases. There appears to be an intrinsic difference between the cell types. With two compounds that induced aberrations in CHO cells and were weakly positive in the Ames test,
Table 1 Summary of genotoxicity profiles on the test compounds CHO Abs
HL Abs
DNA synthesis inhibition
Ames test
AE rat hepatocytes vitro/vivo
DNA adducts vitro/vivo
In vivo rodent micronucleus (bone marrow)
Compound G (Class I) Compound H (Class I) Compound I (Class I) Compound J (Class I) Compound K (Class II) Compound L (Class II) Compound M (Class II) Compound N (Class III)
Pos
Equivocal
Pos
Neg
Neg
Not done
Not done
Pos
Neg
Pos
Neg
Neg (neg TK6 assay)
Not done
Neg
Pos
Neg
Pos
Neg
Neg
Not done
Not done
Pos (20 h treatment–S9)
Pos
Pos
Weak Pos
Neg vitro
Negative in Salmonella
Not done
Pos
Neg
Pos
Weak Pos
Not done
Not done
Equivocal
Weak Pos (↑↑PP)
Neg (↑PP)
Not done
Not done
Not done
Not done
Not done
Pos (↑↑PP)
Neg (↑PP)
Not done
Not done
Not done
Not done
Not done
Pos
Neg
Neg
Neg
Neg in CHO cells
Neg
Compound O (Class IV)
Pos
Pos
Neg
Neg
Not done
Pos
Compound P (Class V)
Pos+− S9 Also pos human S9
Pos
Little or no DNA synthesis inhibition Little or no DNA synthesis inhibition Pos
Neg
Pos vitro, Neg vivo
Neg-
Aphidicolin
Pos
Pos
Pos
Not done
Not done
Pos, calf thymus DNA +S9, rat hepatocytes in vitro; in vivo liver pos but bone marrow neg Negative in CHO cells Not done
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
Compound (therapeutic class)
Not done
AE = alkaline elution assay for DNA strand breaks; PP = polyploidy.
115
116
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
Fig. 9. The percentage of cells with chromosomal aberrations in cultures treated with Compound O (concurrent controls for CHO: 0.5–1.5; concurrent controls for lymphocytes: 0.0–1.0). Numbers on tops of columns are population doublings taken at harvest, as percentages of controls (CHO cells) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. Note different scale of graph due to substantially higher increase in aberrations. ** Statistically significant increase over concurrent control (P ≤ 0.01). * Statistically significant increase over concurrent control (P ≤ 0.05). CHO positive with 3- and 20-h treatments. Lymphocytes positive with 24-h treatment.
one was negative in the lymphocyte aberration assay (K) and the other was weakly positive (J) when the mitotic index was ≥50% of controls (see below). Compound K, is known to affect tumor cells differently from normal cells, and the differential aberration induction in CHO cells and HL may be related to effects on cell cycle checkpoints and apoptosis.
Several of our candidates inhibited DNA synthesis strongly in CHO cells. Early studies that demonstrated the induction of aberrations by inhibitors of DNA synthesis used plants, but this was then demonstrated in normal human cells including lymphocytes [13]. Bender and Moore [5] pointed out that HL were relatively resistant to inhibition of DNA synthesis by aphidicolin compared
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
117
Fig. 10. The percentage of cells with chromosomal aberrations in cultures treated with Compound P (concurrent controls for CHO: 1.0–2.5; concurrent controls for lymphocytes: 1.0). Numbers on tops of columns are cell counts, taken at harvest, as percentages of controls (CHO cells, graph a) population doublings taken at harvest, as percentages of controls (CHO cells, graph b) or mitotic indices, as percentages of controls (lymphocytes). Numbers inside columns are frequency of aberrations per 100 cells as a cell may have more than one aberration. No precipitate observed in CHO culture medium at doses scored. Precipitate observed in whole blood culture medium at 600 M for 3-h treatment. ** Statistically significant increase over concurrent control (P ≤ 0.01). CHO positive at similar doses with cell counts or population doubling (PD) as cytotoxicity assessment but toxicity seen to be much greater by PD method. Lymphocytes positive with 3- and 24-h treatments.
with transformed cells such as mouse L1210 and HeLa, requiring a 100 times higher aphidicolin concentration to achieve the same degree of inhibition of 3 HTdR uptake. However, we found here that aberrations were induced in lymphocytes in the same dose range of aphidicolin that induces aberrations in CHO cells. It is likely that other factors play into the differential aberration induction with some compounds in lymphocytes compared with CHO cells, including more stringent checkpoint controls in human cells that may be more likely to prevent damaged cells reaching mitosis compared with rodent cells [8,9]. Differences in checkpoint control may also be related to the greater sensitivity to induction of aneuploidy and polyploidy of rodent compared with human cells. Tsutsui et al. [14] showed that normal early passage Syrian hamster fibroblasts displayed more aneuploidy and polyploidy than normal human fibroblasts when treated with classical spindle poisons or with the estrogenic compounds 17- estradiol and diethylstilbestrol. We have
also found previously that two compounds that induced high levels of polyploidy (but not structural aberrations) in CHO cells were completely negative in early passage normal human fibroblasts (unpublished data). In the current experiments, two candidates that strongly induced polyploidy in CHO cells, L and M, gave weaker (but still significant) increases in polyploidy in the human lymphocytes. Lymphocytes are sensitive to aberration induction by classical DNA damaging agents ranging from ionizing radiation to alkylating agents, and to clastogens such as topisomerase inhibitors. We found here that a potent aberration inducer (Compound O), which also induced micronuclei in vivo (although it was negative in the Ames test) was also clearly positive in the lymphocyte aberration assay. Two compounds were found to reach convincing aberration levels in the lymphocyte assay only when the mitotic indices were suppressed below 50% of control, Compound J and Compound P. One of these, Compound
118
C. Hilliard et al. / Mutation Research 616 (2007) 103–118
J, is weakly positive in the Ames test and the second, Compound P, although Ames test negative, is capable of inducing DNA strand breaks and DNA adducts, but it appears likely that aberration induction is related more to toxicity and inhibition of DNA synthesis than it is to primary DNA damage. The studies here indicate that for this limited set of compounds, lymphocytes are more likely to give negative or equivocal results for compounds that are positive in the CHO assay associated with toxicity. We note that two compounds that are weakly positive in the Ames test (Compound K and Compound J) and one that induced DNA adducts were negative or equivocal in the lymphocyte aberration assay; it is not clear what the overall mutagenic potential in vivo is for these, and no carcinogenicity data are available. When “pushed” to doses that decrease the mitotic index by more than half compared with controls, additional positive results may also occur in lymphocytes. These studies again pinpoint the unsatisfactory nature of mitotic index as a measure of cyotoxicity/cell growth. The present experiments also demonstrate once again that mitotic index can actually increase at a certain times after treatment. The need to find suitable donors, variability among individuals, and increased time required for the assay along with the inadequacy of the MI as a toxicity measure make the lymphocyte assay less attractive for routine screening. However, studies in lymphocytes or other normal human cells may play a useful part in the follow-up strategy to assess weight of evidence about genotoxic risk for humans for compounds that are positive in the Chinese hamster cell aberration assays. Acknowledgement The blood lymphocyte assays on Compound K were conducted at Covance Laboratories.
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
References [14] [1] S. Galloway, Cytotoxicity and chromosome aberrations in vitro: experience in industry and the case for an upper limit on toxicity
in the aberration assay, Environ. Mol. Mutag. 35 (2000) 191– 201. S. Galloway, J. Miller, M. Armstrong, C. Bean, T. Skopek, W. Nichols, DNA synthesis inhibition as an indirect mechanism of chromosome aberrations: comparison of DNA-reactive and non-DNA-reactive clastogens, Mutat. Res. 400 (1998) 169– 186. C. Hilliard, M. Armstrong, C. Bradt, R. Hill, S. Greenwood, S. Galloway, Chromosome aberrations in vitro related to cytotoxicity of nonmutagenic chemicals and metabolic poisons, Environ. Mol. Mutag. 31 (1998) 316–326. M. Bender, J. Preston, Role of base damage in aberration formation interactions of aphidicolin and X-rays, Progr. Mutat. Res. 4 (1982) 37–46. M. Bender, R. Moore, Dose relationships for different effects of aphidicolin in human peripheral blood leukocytes, Mutat. Res. 198 (1988) 227–231. S. Greenwood, R. Hill, J. Sun, M. Armstrong, T. Johnson, J. Gara, S. Galloway, Population doubling: a simple and more accurate estimation of cell growth suppression in the in vitro assay for chromosomal aberrations that reduces irrelevant positive results, Environ. Mol. Mutag. 43 (1) (2004) 36–44. C. Dunnett, A multiple comparison procedure for comparing several treatments with a control, J. Am. Stat. Assoc. 50 (1955) 1096–1121. A. Kung, S. Sherwood, R. Schimke, Cell line-specific differences in the control of cell cycle progression in the absence of mitosis, Proc. Natl. Acad. Sci. 87 (1990) 9553–9557. R. Schimke, A. Kung, D. Rush, S. Sherwood, Differences in Mitotic Control Among Mammalian Cells, in: Cold Spring Harbor Symposia on Quantitative Biology Vol LVI, 1991, p. 417. V. Reddy, R. Storer, G. Laws, M. Armstrong, J. Barnum, J. Gara, C. McKnight, T. Skopek, J. Sina, J. DeLuca, S. Galloway, Genotoxicity of naturally occurring indole compounds: correlation between covalent DNA binding and other genotoxicity tests, Environ. Mol. Mutag. 40 (1) (2002) 1–17. J. Wilmer, G. Erexson, A. Kligerman, Implication of an elevated sister-chromatid exchange frequency in rat lymphocytes cultured in the absences of erythrocytes, Mutat. Res. 109 (1983) 231–248. K. Mehnert, R. During, W. Vogel, G. Speit, Differences in the induction of SCEs between human whole blood cultures and purified lymphocyte cultures and the effect of an S-9 mix, Mutat. Res. 130 (1984) 403–410. B. Kihlman, W. Nichols, A. Levan, The effect of deoxyadenosine and cytosine arabinoside on the chromosomes of human leukocytes in vitro, Hereditas 50 (1963) 139–143. T. Tsutsui, N. Suzuki, H. Maizumi, C. Carrett, Aneuploidy induction in human fibroblasts: comparison with results in Syrian hamster fibroblasts, Mutat. Res. 240 (1990) 241–249.