|Home | About | Journals | Submit | Contact Us | Français|
The flow cytometry mutation assay (FCMA) uses hybrid CHO AL cells to measure mutations of the cd59 gene located on human chromosome 11 by the absence of fluorochrome-conjugated antibody binding to the CD59 surface antigen. Mutant expression peaks between 6–12 days, then decreases to a stable plateau, instead of a constant mutant fraction obtained by clonogenic assays. To evaluate this variable mutant expression time, cells were treated with radiation, EMS or asbestos and cell proliferation and survival were measured at times leading up to peak mutant expression. Potential doubling time (Tpot) values increased by at least 75% for each agent by 3 hr after treatment but returned to control levels after only three days. Survival returned to 90% of control within a week, close to the peak expression day for all 3 agents. The survival of CD59− cells sorted on the peak day of expression was roughly half that of CD59+ cells. Cloned EMS-treated CD59− cells had a doubling time of 16.7 hr vs. 14.1 hr for CD59+ cells. Triple mutants (CD59−/CD44−/CD90−) were preferentially lost from the population over time, while the proportion of CD59−/CD90− increased. In conclusion, the peak day of mutant expression occurs only when cells recover from the toxic effects of the mutagen. A fraction of cells originally quantified as mutants are lost over time due to lethal deletions and slower growth.
People may be exposed to a wide range of potentially genotoxic substances in the environment. Furthermore, regulatory agencies require many commercial compounds to be tested for mutagenicity prior to production (1). Since it was discovered that many mutagens are also carcinogens, there have been efforts to develop more sensitive methods to detect mutations caused by various agents, including mammalian cell assays such as the mouse lymphoma assay and Chinese hamster ovary–human hybrid (CHO AL) assay (2–4). We and others have developed a flow cytometry mutation assay (FCMA) using the hybrid CHO AL cell line that measures mutant yield in less than two weeks (5–7). Mutations in the cd59 gene located on human chromosome 11 are measured by the absence of fluorochrome-conjugated antibody binding to the CD59 antigen on the cell surface, instead of relying on clonogenic survival.
We have previously characterized the sensitivity of the FCMA to detect mutations induced by a wide range of genotoxic agents, including low responders, while discriminating between true mutagens and false positives (8). Due to the significant differences in mutant detection between clonogenic assays and the FCMA mutations, it is important to determine if mutant fractions generated by the two methods are similar. Significantly larger mutant fractions for the same agents have been measured using the FCMA when compared to clonogenic assays such as the mouse lymphoma assay (MLA) or the HPRT, which is expected since only a small region of the p-arm of human chromosome 11 is essential for survival (9). Also, the FCMA yields much larger mutant fractions than the clonogenic CHO AL assay when comparing the same agent (7,10), even though both assays rely on the loss of CD59 expression on the cell surface for mutant detection.
The peak day of mutant expression for various agents using the FCMA ranges from days 6 to 12 after treatment (8), whereas clonogenic assays measure a stable mutant population 3 to 6 days after treatment that does not vary over time (11). Variations in mutant expression and mutant fraction over time suggest a change in growth and/or survival characteristics of cells with mutations of the cd59 gene. Since cd59 is not an essential gene for the hybrid cells, deletions of the entire gene would not be expected to change cell growth characteristics. Large deletions of chromosome 11 caused by radiation are not lethal and are a significant part of the mutant fraction (12).
To elucidate why the peak day of mutant expression and the mutant fraction vary over time, we investigated growth and survival characteristics of cells treated with a wide range of mutagenic agents, including a clastogen (γ-radiation) (13), an alkylating agent (EMS) (14,15), and asbestos that indirectly acts on DNA by generating reactive oxygen species (16,17). To investigate the variation in the day of peak mutant expression, daily survival and potential doubling time (Tpot) (18,19) were calculated for cells after treatment to determine if mutants are maximally expressed only after cells recover from genotoxic and cellular damage. The variation in mutant fraction over time may be due to loss of cells from the population which are originally scored as CD59 negative. Thus we measured the survival of CD59 negative cells sorted from the mutant region on the peak day of expression, as well as growth characteristics for CD59 negative clones, and compared them to CD59 positive cells.
The CHO AL cells were originally obtained from C.A Waldren (Colorado State University). These cells consist of a standard set of Chinese hamster ovary K1 chromosomes as well as a single copy of human chromosome 11. Cells designated as CHO AL(N), which we used in these experiments, contain a neomycin resistance gene on chromosome 11 which confers resistance to the antibiotic G418. By treating cells periodically with 800 µg/ml G418 antibiotic (Sigma-Aldrich, St. Louis, MO), spontaneous background mutants were reduced. Cells were cultured in Ham’s F-12 nutrient mix supplemented with 10% fetal bovine serum (Gemini Bio-Products, Woodland, CA), penicillin/streptomycin (0.14 and 0.2 g/L, respectively) and 7.5% (w/v) sodium bicarbonate, pH 7.3. Cells were maintained in T75 tissue culture flasks at 37°C in a humidified 5% CO2 incubator. The cells were passed every 3 to 4 days to avoid confluence. After 3 to 4 months of continual use, the flasks of stock cells were discarded and new cells were thawed and maintained as above.
Twenty four hr prior to treatment, 4 × 105 cells were seeded in T-75 flasks so that there would be approximately 1 × 106 cells at treatment time. Cells were then treated with the following: 137Cs γ radiation (4 Gy at 0.93 Gy/min at room temperature), ethyl methane sulfonate (EMS) (6.5 or 8 mM for 3 hr) or chrysotile “A” Rhodesian asbestos (20 µg/ml for 24 hr). Mock treatments were used as controls. Following treatment, cells were washed twice with warm phosphate buffered saline (PBS) and given fresh medium.
Labeling for the presence of CD59 was done using phycoerythrin (PE) – conjugated mouse anti-human CD59 monoclonal antibodies (Caltag Laboratories, Burlingame, CA). The cells were first trypsinized, counted and 1 × 106 cells were transferred to 15 ml conical tubes for processing. With exception of the incubation period, the cells and buffers were kept on ice at all times prior to analysis on the flow cytometer. The tubes were then centrifuged for 3 min at 1500 rpm to pellet the cells. After aspirating the supernatant, the cells were suspended in 1 ml cold staining buffer (1% BSA, 0.1% sodium azide in PBS) and transferred to 1.2 ml microcentrifuge tubes. Following another centrifugation, the supernatant was aspirated and the cells were carefully resuspended in 50 µl staining buffer containing a 1:40 dilution of the anti-CD59 antibody. The tubes were incubated at 37° C for 30 min and subsequently washed once more in 1 ml cold staining buffer. After pelleting the cells, they were resuspended in 0.5 ml staining buffer and passed through a 40 µm nylon mesh filter into flow cytometry analysis tubes.
At time points ranging from 3 hr to 8 days after treatment with EMS, asbestos or radiation, non-confluent dividing cells were pulse labeled with 10 µM bromodeoxyuridine (BrdU) for 20 minutes by adding BrdU directly to the culture medium. Cells were then washed twice with warm PBS and returned to the incubator for 3 hr to allow for uptake of BrdU during DNA synthesis. After 3 hr, cells were trypsinized, counted and 5 × 105 cells were transferred to 15 ml conical tubes for processing. Samples were centrifuged for 3 min at 1500 rpm to pellet cells. The supernatant was aspirated and the cells were resuspended in 2 ml cold PBS. The samples were then fixed by adding 5 ml cold 100% ethanol (70% final concentration) drop wise while vortexing gently. Samples were kept on ice for at least 30 min prior to denaturation and staining.
After 30 min samples were centrifuged, the supernatant was removed and cells were resuspended in 0.1 ml PBS. 2 ml of pepsin/HCl (0.2 mg/ml in 2 N HCl) was slowly added to the samples while gently vortexing and samples were incubated at 37° C for 20 min. After incubating, 3 ml of 1 M Tris was added to samples to terminate pepsin hydrolysis and the samples were centrifuged. The supernatant was aspirated and the cells were resuspended in 2 ml PBS and centrifuged again. The PBS was then aspirated and the pellet was resuspended in 100 µl of a 1:30 dilution of the primary mouse anti-BrdU antibody (Dako, Denmark) in TBFP buffer (0.5% Tween 20, 1% BSA, 1% FBS in PBS). Samples were incubated for 25 min at room temperature and then rinsed with 5 ml PBS and centrifuged. The supernatant was decanted and the pellet was resuspended in 200 µl of a 1:200 dilution of the secondary antibody goat anti-mouse IgG conjugated to Alexa Fluor® 488 (Molecular Probes, Eugene, OR) in deionized water and the samples were incubated for 25 min at room temperature in the dark. The samples were again rinsed with 5 ml PBS and resuspended in 1 ml of propidium iodide (10µl/ml in PBS) containing RNAse (40 KU/ml). Samples were incubated at room temperature for 20 min prior to flow cytometry analysis.
An EPICS V (Beckman Coulter, Miami, FL) and a MoFlo™ (Beckman Coulter, Fort Collins, CO) flow cytometers were used for cell cycle experiments utilizing bromodeoxyuridine (BrdU) and propidium iodide (PI). For the EPICS V a 488 nm laser was used for excitation along with 515 nm longpass, 590 nm dichroic, 590 nm longpass (red photomultiplier tube) and 530 nm shortpass (green photomultiplier tube) filters. Gating of forward scatter and integral red fluorescence versus peak integral red fluorescence was use to select for whole cells with incorporated PI. Histograms of PI and PI versus log integral green fluorescence (LIGFL) for 30,000 events were collected for each sample. Histograms were analyzed using MultiCycle and Multi2D software (Phoenix Flow Systems, San Diego, CA. A method to calculate Relative Movement of green fluorescent cells (S phase cells) was previously described, as well as determining time in S phase (Ts) and Tpot values (18,20). The percentage of cells in G2 phase was determined by analyzing the PI cell cycle distributions with MultiCycle.
Cells were trypsinized, counted and plated for survival in 6-well plates from 2 to 12 days after treatment with EMS, asbestos or radiation. Three wells containing 4 ml medium were plated with 300 to 500 treated cells depending on the time after treatment and three wells were plated with 300 control cells. Cells were grown for 7 to 8 days before they were stained with crystal violet for colony counting. Control wells were used to calculate plating efficiency (PE) and surviving fraction of treated cells was calculated by (colonies / (PE × cells plated)).
CD59 negative cells from the mutant region of the FCMA were isolated using a MoFlo™ Cell Sorter (Beckman Coulter, Fort Collins, CO), using the same filter set-up and gating scheme as previously published (8). Single cells were sorted into individual wells of 96-well culture plates containing 200 µl of medium. Sorted cells were grown for 18 to 21 days, replacing medium twice, until they formed visible colonies. Cloned cells were transferred by pipette to T25 tissue culture flasks, allowed to expand and passed when nearly confluent.
On the day of peak mutant expression after treatment with EMS (9 days) and γ-radiation (6 days), 300 cells were sorted from four regions of the mutant peak into 35 mm culture dishes containing 4 ml of medium using a MoFlo™ cell sorter. Cells were allowed to grow for 7 to 8 days before colonies were stained with crystal violet and counted for surviving fraction.
Single cells sorted from mutant regions 1–3 (Fig. 4A) nine days after treatment with 8 mM EMS were expanded to 11 stable CD59 negative clonal populations. Four clones were isolated from region 1, 5 clones from region 2, and 2 clones from region 3. One million cells were seeded in T75 flasks at a 1:1 ratio of isolated mutant clone cells and sorted CHO AL control cells (CD59 positive). Cells were trypsinized, counted and passed (2 × 105) on days 3, 6 and 10 after mixing. Isolated cells were stained with CD59 specific antibodies and the number of CD59 positive and CD59 negative cells was determined by gating within flow cytometry histograms.
A CyAn™ flow cytometer (Beckman Coulter, Ft. Collins, CO) was used with a 488 nm solid-state laser for excitation and a 545 nm dichroic longpass and 575/25 bandpass filter for PE detection. A total of 1 × 105 cells were analyzed for each sample. A gating region was set on forward scatter versus side scatter to eliminate cellular debris. The photomultiplier tube voltage was set so that control CD59+ cells were approximately in channel 1200.
Multicolor staining for CD59/CD90/CD44 was done as previously described (21). For CD59/CD90/CD44, cells were isolated and washed once, then suspended in 50 µl staining solution (1.3 µl anti-CD59-PE, 10 µl anti-CD44-biotin (Serotec, Raleigh, NC), 10 µl anti-CD90-Alexa647 (Serotec, Raleigh, NC) and 28.7 µl staining buffer). Samples were incubated on ice for 30 min, then 1 ml staining buffer was added to each sample and samples were centrifuged. Samples were resuspended in 100 µl Streptavidin-Alexa488 (1:100 dilution in staining buffer) (Molecular Probes, Eugene, OR) and were incubated on ice for 30 min. The pellet was resuspended in 0.5 ml staining buffer and passed through a 40 µm mesh filter into flow cytometry analysis tubes. The fraction of CD59 mutants was determined by gating on 10% of the CD59+ peak, as discussed by Ross and Fox (21).
The variation in the fraction of mutant cells measured by the FCMA after treatment with EMS, radiation or asbestos shows a peak in mutant expression (6–9 days) followed by a decrease over time (Fig. 1). As previously reported, the peak day of mutant expression varies for different treatments (8).
We first determined whether this variation in peak day of mutant expression was the result of alterations in cell growth or the cell cycle. By using bivariate histograms of DNA content versus BrdU incorporation, the relative movement of cells through S phase and in turn the potential doubling time of treated cells was calculated by standard methods. Tpot values for control cells before treatment with EMS, radiation and asbestos was determined to be 14.1 ± 1.2 hr, which corresponds well to the CHO AL doubling time of 14.5 hr established by our lab using growth curves. Tpot values for cell populations more than doubled 3 hr after treatment with 8 mM EMS, increased 75 % by 6 hr after treatment with asbestos, and the percentage of cells in G2 phase more than tripled after treatment with 4 Gy γ-rays (Fig. 2). Because of the large G2 delay caused by radiation, it was not possible to get an accurate measure of Tpot. By day 3 after treatment normal cycling had resumed, giving Tpot values similar to controls, which is well before the peak day of mutant expression for all three agents. Thus, the peak day of mutant expression cannot be explained simply by alterations in cell growth.
We next evaluated whether the variation in peak mutant expression was due to the difference in the time for the recovery of clonogenic survival. Cells were treated with 6.5 or 8.0 mM EMS, 20 µg/ml asbestos or 4 Gy γ-radiations. At various times up to 12 days after treatment, cells were isolated and plated for clonogenic survival along with controls for each sample. After growing for 7–8 days, colonies were counted and surviving fractions were determined (Fig. 3). The survival on the peak day of mutant expression was ~96% for both 6.5 and 8.0 mM EMS, ~97% for asbestos and ~70% for cells treated with radiation. The survival of cells treated with all agents recovered significantly from times right after treatment to the peak day of mutant expression.
The fraction of CD59 negative cells was reduced substantially after the peak day of mutant expression (Fig. 1), as previously reported (8). Cells were sorted from six regions of the mutant peak of radiation and EMS treated cells and were plated for clonogenic survival (Fig. 4A). These 6 regions were selected because we have previously shown that, while the majority of CD59 mutants are in regions 1 and 2, some mutants are also present in regions 3–6 (21). Cells treated with 4 Gy γ-rays and isolated from region 1 had a survival of ~20% while regions 2–6 were ~50% of control cells (Fig. 4B). Cells sorted from the 6 mutant regions 9 days after treatment with 8 mM EMS had a survival of less than 40% compared to untreated control cells and only 50% of cells sorted from the positive peak, while the treated cells in the CD59+ peak had a survival of 80% compared to the control (Fig. 4B). However the survival of cells sorted from the 6 mutant regions 35 days after treatment with 8 mM EMS was indistinguishable from controls (Fig. 5). This suggests that a large fraction of CD59− cells that are counted on the peak day of mutant expression are not viable, are lost from the population and will not be detected when samples are stained several days later.
To determine if CD59 mutants grew at a slower rate than CD59 positive cells, single cells sorted from mutant regions 1–3 (Fig. 4A) 9 days after treatment with 8 mM EMS were expanded to 11 stable CD59− clonal populations. On days 3, 6 and 10 after mixing with CD59 positive control clones, the percentage of CD59− cells in the total population was determined by flow cytometry. By compiling the results for all 11 clones (Fig. 6) we find CD59− cells are lost from the mixed population at an exponential rate. The average doubling time (Td) of the 11 clonal populations was determined by growth curves at least 30 days after treatment with EMS to be 16.7 ± 0.9 h while the Td for stock CHO AL cells has been measure by our lab to be 14.5 hr. Again, it is important to note that these clones were in culture for over 30 days prior to mixing and were selected as the fastest growing clones from the 96 well plates.
In order to ascertain if complex mutations are lost at a disproportionate rate from the population, we measured the expression of 3 different mutations simultaneously (Fig 7). From days 6 to 24 after treatment with 8 mM EMS, CD59− cells were analyzed for the presence or absence of CD90 and/or CD44 (Fig. 8). Spontaneous background mutant fractions were stable over months as follows: CD59− (13.0 ± 5.4%), CD59−/CD44− (27.7 3.5%), CD59−/CD90− (8.0 ± 2.6%) and triple mutants (51.3 ±8.3 %). These percentages are of a total background CD59 mutant fraction of ~0.3% of the total population. Six days after treatment the fractions of CD59 mutants that were also negative for another marker changed dramatically. CD59 only mutants increased to about 30% of the total CD59− population on day 6 but stabilized at about 10% by day 9. Triple mutants slowly decreased from 35% on day 6 to slightly more than 20% on day 24. The fraction of CD59−/CD44− cells decreased from 28% to 12% over the same time. Finally, only 10% of CD59− cells were also CD90− on day 6 after treatment but rapidly increased to nearly 60% of CD59 mutants by day 24.
The purpose of our study was to characterize the kinetics of mutant expression measured by the FCMA by determining the causes of the variation in the peak day of mutant expression for different agents as well as the subsequent decrease in the mutant fraction over time. Logistically, differences in the peak day of expression can be overcome by sampling and measuring mutants on a range of days following treatment with a novel agent on days 6, 9 and 12 for example, which would include the peak day for mutant expression for every mutagen tested with the FCMA to date. Though this would increase the amount of time required in testing a new compound, the FCMA is already more rapid then standard assays (MLA, CHO AL) by a matter of weeks (8).
To better understand the variation in peak mutant expression between different genotoxic agents we analyzed the clonogenic survival of cells over time after treatment, in conjunction with measuring Tpot and fraction of G2 cells, to determine if peak mutant expression corresponded to a return of treated cells to untreated cell growth and survival characteristics. Tpot analysis for EMS and asbestos treated cells showed that normal cell proliferation resumed within 3 days after treatment. The percentage of cells in G2 phase was calculated for irradiated cells since the well characterized G2 block induced by radiation in mammalian cells (22) made it impossible to calculate Tpot. The percentage of G2 phase cells in the irradiated population returned to control levels within three days, indicating normal cell cycling had resumed. Three days is well before the earliest peak mutant expression of day 6 for radiation, and normal cell cycling has resumed nearly a week before the peak day for cells treated with asbestos and EMS. Thus these changes in cell growth parameters cannot fully explain the delay and variation in peak mutant expression.
Unlike cell proliferation, survival of treated cells did not return to control levels until the peak day of mutant expression (EMS and asbestos) or just after (radiation). Therefore, the time at which survival of the treated population returns to control levels can be considered a rough indicator for the maximal day of mutant expression and is supported by results for other DNA damaging agents like N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) with survival at ~95% on the peak day of expression (day 12) (8). This is most likely due to the loss of cells that are destined to die from the population so the remaining cells are nearly all viable.
After the peak day of mutant expression, the mutant fraction decreased to a stable plateau. The FCMA differs from traditional mammalian mutation assays in that mutants are measured by the lack of CD59 staining instead of selective clonogenic survival, allowing for the possibility of counting dying cells, cells blocked in the cell cycle, or slower growing cells, which would not be scored in the clonogenic assays. To determine if the peak mutant fraction is artificially high due to the inclusion of cells that will ultimately die, we examined the survival of CD59− and CD59+ cells. The survival of CD59− cells on the peak day of expression after irradiation was roughly half that of CD59+ cells, suggesting that CD59− cells would be disproportionately lost from the population over time as we have previously proposed. This would indicate that the plateau level is likely the fraction of mutants that are able to survive and would be similar to the mutant yield that would be measured by a clonogenic assay.
It is worth noting that the level of the plateau relative to the peak is not constant for the three different mutagens analyzed here. The plateau is about 40% of the peak for EMS, less than 20% for gamma radiation and not significantly different for asbestos. The peak levels also vary substantially. This is likely a reflection of the type of mutations caused by the three agents. Ionizing radiation is known to be a clastogen (13) which often causes very large deletions (see Ref 12 for example); EMS is an alkylating agent (14,15) that causes small mutations, though it can also cause some large deletions (manuscript in preparation); and asbestos induces ROS (reactive oxygen species) (16,17) and hence causes point mutations. It is consistent with the mechanisms that radiation would cause greater cell killing than EMS which would be greater than asbestos. The plateau levels would reflect the varying degrees of cell killing, which would explain the different peak/plateau ratios we observed.
To verify that CD59− cells are disproportionately lost from populations due to growth characteristics, we mixed clonal CD59− and CD59+ cell populations and found that mutants were rapidly lost from the population and on average had doubling times over 2 hr longer than control cells. Slowly growing mutants have also been observed in the L5178Y/TK+/− Mouse Lymphoma Assay (23) and TK6 assay (24). Since the cd59 gene is located on a nonessential copy of human chromosome 11, mutations of cd59 and even large deletions of adjacent genes should have no effect on growth, unlike thymidine kinase assays like the MLA where deletions of adjacent genes lead to slowly growing mutants or cell death (25). It is interesting to note that our previous results with γ-radiation did not show a decreased growth rate in the mutant population (21). The difference between EMS-induced mutants and radiation-induced mutants may be that EMS causes smaller mutations that may not kill cells but slows their growth whereas radiation causes large deletions that kill cells if mutations occur in other chromosomes besides chromosome 11.
Another factor leading to the loss of CD59− cells over time may be large deletions or the loss of entire chromosome 11, including the essential region on the short arm at 11p15.5 (26). The loss of this essential region may not be immediately lethal, resulting in cells with this deletion to be lost over several weeks. The FCMA allows for the possibility of measuring multiple mutations on chromosome 11 simultaneously to better understand the mutagenic lesions caused by various genotoxic agents (21). The fraction of cells with mutations only in CD59 slightly increases on day 6 but were stable at ~10% at later times. However, complex mutants and larger deletions such as CD59−/CD44− and triple mutants become a smaller proportion of all CD59− cells over the 24 day period. On the other hand the fraction of CD59−/CD90− mutants increased over time. Most of these mutants are not complex deletions but instead single mutations in glycosylphosphatidylinisotol (GPI) anchor X-linked gene PigA (27). CD59 and CD90 are anchored to the extracellular membrane by GPI so cells with mutated PigA will be negative when stained with anti-CD59/CD90 antibodies. Of 32 EMS generated mutants that lack surface staining for both CD59 and CD90, over 90% lack GPI anchors when stained with FLAER, a bacterial toxin aerolysin conjugated to Alexa-488 (12) (manuscript in preparation).
The variation in the peak day of mutant expression is likely due to the difference in the amount of time needed for cells to recover from genetic insult and express the mutated phenotype. We conclude that the decrease in the mutant fraction after the peak day of expression is caused by the loss of cells destined to die that may have lost chromosome 11 entirely and to slowly growing mutants. We have previously shown that AL cells with triple mutants after ionizing radiation are preferentially lost from the cell population over time (21). Some cells scored as mutants with the FCMA may not be viable and would not be detected using the clonogenic assay. While these CD59− cells may not be considered true mutants since they are inviable, they do give a more accurate measure of DNA damage caused by genotoxic agents since they clearly do show that there is a higher fraction of induced mutations than would be indicated by a clonogenic assay.
This work was supported by NIH SBIR Grant Number R44 CA91566 awarded to Cytomation GTX, Inc. M.H. Fox is the owner of Cytomation GTX, Inc.
This work was presented in part at the XXIV Congress of the International Society for Analytical Cytology in Budapest, Hungary, May 17–21, 2008.