91
Mutation Research, 253 (1991) 91-95 © 1991 Elsevier Science Publishers B.V. 0165-1161/91/$03.50 ADONIS 016511619100093E
MUT 00009
Mutagenicity of 2-amino-N6-hydroxyadenine to TK6 human lymphoblast cells Howard
L. L i b e r a n d C h e r y l M . D e n a u l t
Department of Cancer Biology, Harvard School of Public Health, Boston, MA 02115 (U.S.A.) (Received 30 October 1989) (Accepted 30 June 1990)
Keywords: 2-Amino-N6-hydroxyadenine; TK6 cells; Lymphoblast cells, human
Summary T K 6 human lymphoblast cells (tk +/ ; hprt +) were treated with various concentrations of 2-amino-N 6hydroxyadenine (AHA) for 24 h. A H A was quite toxic to TK6 cells in the dose range 0-0.05 g g / m l , but additional toxicity was not observed between 0.05 and 0.10 g g / m l . A H A induced mutations at 2 distinct genetic loci: the autosomal thymidine kinase (tk) and the X-linked hypoxanthine-guanine phosphoribosyl transferase (hprt). Significant levels of both tk-NG mutants (normal growth rate of 16-18 h, colonies visible after 10-11 days incubation) and tk-SG mutants (slow growth rate of > 24 h, colonies visible after 18 days incubation) were induced. 15 hprt mutants were isolated and analyzed by Southern blot. 8 of these had normal restriction fragment patterns after digestion with PstI, EcoRI, and HindIII, and were defined as ' p o i n t ' mutations; the remaining 7 had partial deletions of the hprt gene. 32 tk- mutants were also isolated. 3 of 22 normal growth mutants and 6 of 10 slow growth mutants had lost the active tk allele. These data suggest that both point mutations and larger-scale alterations are induced by AHA.
The TK6 human lymphoblast system is a useful tool for studying mutation induced by genotoxic agents. One of its major advantages is that quantitative levels of mutation induced at an autosomal heterozygous locus (thymidine kinase, tk) and an X-linked hemizygous locus (hypoxan-
Correspondence: Dr. Howard L. Liber, Department of Cancer Biology, Harvard School of Public Health, Boston, MA 02115
(U.S.A.). Abbreviations: AHA, 2-amino-N6-hydroxyadenine; hprt, hypoxanthine-guanine phosphoribosyl transferase; TFT, trifluorothymidine; 6TG, 6-thioguanine; tk, thymidine kinase.
thine-guanine phosphoribosyl transferase, hprt) can be compared directly. In addition, it is possible to examine the molecular nature of individual mutants at either of these loci with Southern blot techniques (Yandell et al., 1986; Liber et al., 1987). Several potentially mutagenic mechanisms presumably can operate at either gene locus. 'Point' mutations (base substitutions or insertions or deletions of up to several base pairs) or larger deletions or rearrangements of the entire gene can inactivate either locus (Yandell et al., 1986; Liber et al., 1987). Much larger deletions are also theoretically possible, but those which affect the X chromosome may encompass one or more genes
92 which are essential for cell viability. Such a mutation would not be detected because the cell would die. However, large multilocus deletions of chromosome 17 may result in recoverable mutants, due to the presence of the homologous chromosome and thus the homologous essential gene. Finally, mutagenic mechanisms which require homologous sequences in order to operate (e.g., gene conversion or recombination), would presumably function at tk but not at hprt. Quantitative comparisons of mutations induced at these genetic loci might thus be predicted to reveal higher mutant fractions for tk than for hprt, if indeed additional mutagenic mechanisms were operating at the autosomal locus. Recent studies have shown that more mutants are in fact observed at the tk locus (Yandell et al., 1986; Liber et al., 1989). The additional tk mutants are a phenotypically distinct class, which can be recognized by their slow growth rates (tk-SG mutants). Southern blot analyses revealed that more than 95% of these tk-SG mutants had lost the entire active tk + allele and had become either hemizygous or homozygous for the inactive t k allele (Yandell et al., 1986; Little et al., 1987). However, since many t k - mutants with normal growth rates ( t k - N G mutants) also have lost the entire tk + allele, we have concluded that tk-SG mutants result from some sort of multilocus event, such as large deletion or loss of heterozygosity through recombinational mechanisms. Theoretically then, the TK6 mutation system can be used to categorize the nature of induced mutations. A mutagen which acted predominantly by inducing base pair substitutions would be expected to induce hprt and tk-NG, but few if any tk-SG mutants. Furthermore, Southern blot analyses would reveal restriction fragment patterns which would look like the non-mutant control. In this paper, we have examined the mutagenicity of AHA, an adenine analogue which might be expected to induce only point mutations. However, at the hprt locus, we found that about half the mutants were point mutations and the other half were partial deletions. For t k - N G mutants, point mutations were the predominant class of mutant induced. However, A H A also induced a significant fraction of tk-SG mutants, suggesting
that it also is capable of inducing multilocus events. Materials and methods General cell culture methods TK6 human lymphoblasts were grown exponentially in R P M I 1640 medium supplemented with 10% horse serum (Gibco). Cells were maintained in stationary culture in loosely capped tissue culture flasks at 3 7 ° C in 5% CO 2 at densities of 1-12 × 105 cells/ml. TK6 cells normally grow with a doubling time of 16-18 h and form macroscopic colonies with 50-80% efficiency when incubated for 10-12 days in 96-well microtiter dishes (Furth et al., 1981). Mutation assay Prior to use in a mutation assay, TK6 cells were grown in C H A T medium for 2 days to reduce the level of spontaneous tk - / - and hprt- mutants. C H A T medium was complete medium ( R P M I 1640 plus horse serum) with 10 5 M deoxycytidine, 2 × 10 -4 M hypoxanthine, 2 × 1 0 - 7 M aminopterin, and 1.75 × 10 - s M thymidine. Following C H A T treatment, cells were resuspended in C H T ( C H A T without aminopterin) medium for 1 day. Cells were used in mutation assays within several days of C H A T treatment. Cultures of 100 ml were treated at cell densities of 4 × 10 s ceils/ml. Cells were treated with A H A (from Radian Corp., Austin, TX) from a stock solution of 1 m g / m l in saline, prepared fresh and filter-sterilized before use. At the end of the 24-h treatment, each culture was centrifuged and resuspended in fresh medium. Surviving fractions were determined by seeding cells in microtiter dishes immediately after treatment, and counting colonies after 12 days incubation. After treatment, cultures were grown in nonselective medium to allow phenotypic expression prior to plating for determination of mutant fraction. Expression times were 3 days for T F T R and 7 days for 6TG R (Liber and Thilly, 1982). After expression, cells were seeded into 96-well microtiter plates at a density of 20000-40000 cells/well in 0.2 m l / w e l l in the presence of either 0.5/~g/ml 6-thioguanine (6TG) to select hprt- mutants or 2 . 0 / ~ g / m l trifluorothymidine (TFT) to select t k -
93 mutants. Cells from each culture were also plated at 1 cell/well in the absence of the selective agent to determine plating efficiency (PE). All plates were incubated in a 37°C, 5% CO2, humidified incubator for 11 days prior to scoring colonies. Plates for determination of tk- mutation frequency were then refed with 0.022 ml of fresh medium. The concentration of T F T used for refeeding was increased to 20/.Lg/ml, in order to restore the level of this selective agent to its original concentration. This was required because T F T is unstable in medium, and extended incubation of plates results in the outgrowth of wild-type colonies which escape initial selection (Liber et al., 1985). After refeeding, the plates were incubated for an additional 7 days to observe the appearance of slow-growing mutants. Mutant fractions were calculated from the Poisson distribution, as described by Furth et al. (1981).
Isolation and characterization of mutants Mutant colonies were removed from microtiter dishes for further analysis. Each mutant was recloned at a cell density of 0.1 cell/well to insure that the resulting mutant was monoclonal in origin. Large cultures of each mutant were then grown and characterized. To insure that all mutants had the expected phenotype, aliquots of 105 cells were withdrawn and placed in medium containing either 6TG, T F T or CHAT. These flasks were scored for the presence or absence of growth 4-5 days later. Mutants were expected to grow in 6TG or T F T but not in CHAT. DNAs were isolated from mutant cell lines, quantified and digested with PstI, HindlII or EcoRI (for hprt analysis) or SacI (for tk analysis). DNAs were electrophoresed in 0.8% agarose gels, transferred to nitrocellulose and hybridized to the appropriate 32p-labeled probe (a 950-bp c D N A for hprt, obtained from C.T. Caskey, or a l l 0 0 - b p c D N A for tk, obtained from P. Deineger). A SacI polymorphism for tk allowed the distinction of the active and inactive alleles of this gene (Yandell et al., 1986). Results and discussion
Toxicity of A HA TK6 human lymphoblast cells were treated with various concentrations of AHA (0-0.1/~g/ml) for
~3
1.000~
2 ',
0100 t
i~
'N
•
0.007/~ 0.000
I 0.020
,
I 0.040
,
AHA c o n c e n t r a t i o n ,
I 0.060
,
/~9/ml
I 0.080
,
I -0.100
f o r 2 4 hr
Fig. 1. Toxicity of A H A to T K 6 h u m a n lymphoblast cells. Each point is the average of 2 independent Expts. done in duplicate.
24 h. Surviving fractions were measured immediately after treatment. Fig. 1 shows that A H A was toxic in the dose range 0-0.05 /~g/ml; however, little additional toxicity was observed between 0.05 and 0.10 ktg/ml.
Mutagenicity of AlIA Fig. 2 is a plot of the mutational dose response of TK6 cells exposed to A H A for 24 h. Approximately equal levels of both tk-NG mutants (normal growth rate of 16-18 h, colonies visible after 1 0 - 1 ! days incubation) and tk-SG mutants (slow growth rate of > 24 h, colonies visible after 18 days incubation) were induced. A slightly higher ~D o x
280
u>" 240
g g h
200 160
120 o
~ u -o c
80
400 000
I 0020
I 0.040
,
AHA c o n c e n t r o t i o n ,
I 0060
,
/zg/ml
I
O OBO
0
}lO0
for 24 hr
Fig. 2. Mutagenicity of A H A to T K 6 h u m a n lymphoblast cells. Induced mutation frequencies were determined for hprt (0, 6TGR), tk-NG (ll, T F T R colonies visible after 11 days incubation) and tk-SG ( - , T F T R colonies visible only after 7 days additional incubation), Each point is the average of duplicate determinations, which were independent of the experiments shown in Fig. 1. The background mutation frequencies were 3.6 × 10 -6 (6TG), 2.7 × 10 -6 (tk-NG) and 6.2 × 10 - 6 (tk-SG).
94 TABLE 1 Molecular Nature of Mutants Induced by A H A Locus
Dose (/~g/ml)
Induction level
tk-NG
0.10 0
6-7 ×
tk-SG
0.10 0
hprt
0.10 0.07 0
N u m b e r of mutants Point
Large-scale
19 10
3 41 a
4× -
4 4
6 116 a
60 × 35 X -
4 4 19
1 6 10
-
mutants show allele loss; see Table 1). In the set of 10 tk-SG mutants, there were 4 point mutants and 6 which had lost the active allele. The expectation was that 2 or 3 spontaneous mutants would
Induction level is the observed mutation frequency divided by the background. Point mutations were defined as no change in Southern blot pattern. a Data from Little et al. (1987).
~
Mutation Research, 253 (1991) 91-95 © 1991 Elsevier Science Publishers B.V. 0165-1161/91/$03.50 ADONIS 016511619100093E
MUT 00009
Mutagenicity of 2-amino-N6-hydroxyadenine to TK6 human lymphoblast cells Howard
L. L i b e r a n d C h e r y l M . D e n a u l t
Department of Cancer Biology, Harvard School of Public Health, Boston, MA 02115 (U.S.A.) (Received 30 October 1989) (Accepted 30 June 1990)
Keywords: 2-Amino-N6-hydroxyadenine; TK6 cells; Lymphoblast cells, human
Summary T K 6 human lymphoblast cells (tk +/ ; hprt +) were treated with various concentrations of 2-amino-N 6hydroxyadenine (AHA) for 24 h. A H A was quite toxic to TK6 cells in the dose range 0-0.05 g g / m l , but additional toxicity was not observed between 0.05 and 0.10 g g / m l . A H A induced mutations at 2 distinct genetic loci: the autosomal thymidine kinase (tk) and the X-linked hypoxanthine-guanine phosphoribosyl transferase (hprt). Significant levels of both tk-NG mutants (normal growth rate of 16-18 h, colonies visible after 10-11 days incubation) and tk-SG mutants (slow growth rate of > 24 h, colonies visible after 18 days incubation) were induced. 15 hprt mutants were isolated and analyzed by Southern blot. 8 of these had normal restriction fragment patterns after digestion with PstI, EcoRI, and HindIII, and were defined as ' p o i n t ' mutations; the remaining 7 had partial deletions of the hprt gene. 32 tk- mutants were also isolated. 3 of 22 normal growth mutants and 6 of 10 slow growth mutants had lost the active tk allele. These data suggest that both point mutations and larger-scale alterations are induced by AHA.
The TK6 human lymphoblast system is a useful tool for studying mutation induced by genotoxic agents. One of its major advantages is that quantitative levels of mutation induced at an autosomal heterozygous locus (thymidine kinase, tk) and an X-linked hemizygous locus (hypoxan-
Correspondence: Dr. Howard L. Liber, Department of Cancer Biology, Harvard School of Public Health, Boston, MA 02115
(U.S.A.). Abbreviations: AHA, 2-amino-N6-hydroxyadenine; hprt, hypoxanthine-guanine phosphoribosyl transferase; TFT, trifluorothymidine; 6TG, 6-thioguanine; tk, thymidine kinase.
thine-guanine phosphoribosyl transferase, hprt) can be compared directly. In addition, it is possible to examine the molecular nature of individual mutants at either of these loci with Southern blot techniques (Yandell et al., 1986; Liber et al., 1987). Several potentially mutagenic mechanisms presumably can operate at either gene locus. 'Point' mutations (base substitutions or insertions or deletions of up to several base pairs) or larger deletions or rearrangements of the entire gene can inactivate either locus (Yandell et al., 1986; Liber et al., 1987). Much larger deletions are also theoretically possible, but those which affect the X chromosome may encompass one or more genes
92 which are essential for cell viability. Such a mutation would not be detected because the cell would die. However, large multilocus deletions of chromosome 17 may result in recoverable mutants, due to the presence of the homologous chromosome and thus the homologous essential gene. Finally, mutagenic mechanisms which require homologous sequences in order to operate (e.g., gene conversion or recombination), would presumably function at tk but not at hprt. Quantitative comparisons of mutations induced at these genetic loci might thus be predicted to reveal higher mutant fractions for tk than for hprt, if indeed additional mutagenic mechanisms were operating at the autosomal locus. Recent studies have shown that more mutants are in fact observed at the tk locus (Yandell et al., 1986; Liber et al., 1989). The additional tk mutants are a phenotypically distinct class, which can be recognized by their slow growth rates (tk-SG mutants). Southern blot analyses revealed that more than 95% of these tk-SG mutants had lost the entire active tk + allele and had become either hemizygous or homozygous for the inactive t k allele (Yandell et al., 1986; Little et al., 1987). However, since many t k - mutants with normal growth rates ( t k - N G mutants) also have lost the entire tk + allele, we have concluded that tk-SG mutants result from some sort of multilocus event, such as large deletion or loss of heterozygosity through recombinational mechanisms. Theoretically then, the TK6 mutation system can be used to categorize the nature of induced mutations. A mutagen which acted predominantly by inducing base pair substitutions would be expected to induce hprt and tk-NG, but few if any tk-SG mutants. Furthermore, Southern blot analyses would reveal restriction fragment patterns which would look like the non-mutant control. In this paper, we have examined the mutagenicity of AHA, an adenine analogue which might be expected to induce only point mutations. However, at the hprt locus, we found that about half the mutants were point mutations and the other half were partial deletions. For t k - N G mutants, point mutations were the predominant class of mutant induced. However, A H A also induced a significant fraction of tk-SG mutants, suggesting
that it also is capable of inducing multilocus events. Materials and methods General cell culture methods TK6 human lymphoblasts were grown exponentially in R P M I 1640 medium supplemented with 10% horse serum (Gibco). Cells were maintained in stationary culture in loosely capped tissue culture flasks at 3 7 ° C in 5% CO 2 at densities of 1-12 × 105 cells/ml. TK6 cells normally grow with a doubling time of 16-18 h and form macroscopic colonies with 50-80% efficiency when incubated for 10-12 days in 96-well microtiter dishes (Furth et al., 1981). Mutation assay Prior to use in a mutation assay, TK6 cells were grown in C H A T medium for 2 days to reduce the level of spontaneous tk - / - and hprt- mutants. C H A T medium was complete medium ( R P M I 1640 plus horse serum) with 10 5 M deoxycytidine, 2 × 10 -4 M hypoxanthine, 2 × 1 0 - 7 M aminopterin, and 1.75 × 10 - s M thymidine. Following C H A T treatment, cells were resuspended in C H T ( C H A T without aminopterin) medium for 1 day. Cells were used in mutation assays within several days of C H A T treatment. Cultures of 100 ml were treated at cell densities of 4 × 10 s ceils/ml. Cells were treated with A H A (from Radian Corp., Austin, TX) from a stock solution of 1 m g / m l in saline, prepared fresh and filter-sterilized before use. At the end of the 24-h treatment, each culture was centrifuged and resuspended in fresh medium. Surviving fractions were determined by seeding cells in microtiter dishes immediately after treatment, and counting colonies after 12 days incubation. After treatment, cultures were grown in nonselective medium to allow phenotypic expression prior to plating for determination of mutant fraction. Expression times were 3 days for T F T R and 7 days for 6TG R (Liber and Thilly, 1982). After expression, cells were seeded into 96-well microtiter plates at a density of 20000-40000 cells/well in 0.2 m l / w e l l in the presence of either 0.5/~g/ml 6-thioguanine (6TG) to select hprt- mutants or 2 . 0 / ~ g / m l trifluorothymidine (TFT) to select t k -
93 mutants. Cells from each culture were also plated at 1 cell/well in the absence of the selective agent to determine plating efficiency (PE). All plates were incubated in a 37°C, 5% CO2, humidified incubator for 11 days prior to scoring colonies. Plates for determination of tk- mutation frequency were then refed with 0.022 ml of fresh medium. The concentration of T F T used for refeeding was increased to 20/.Lg/ml, in order to restore the level of this selective agent to its original concentration. This was required because T F T is unstable in medium, and extended incubation of plates results in the outgrowth of wild-type colonies which escape initial selection (Liber et al., 1985). After refeeding, the plates were incubated for an additional 7 days to observe the appearance of slow-growing mutants. Mutant fractions were calculated from the Poisson distribution, as described by Furth et al. (1981).
Isolation and characterization of mutants Mutant colonies were removed from microtiter dishes for further analysis. Each mutant was recloned at a cell density of 0.1 cell/well to insure that the resulting mutant was monoclonal in origin. Large cultures of each mutant were then grown and characterized. To insure that all mutants had the expected phenotype, aliquots of 105 cells were withdrawn and placed in medium containing either 6TG, T F T or CHAT. These flasks were scored for the presence or absence of growth 4-5 days later. Mutants were expected to grow in 6TG or T F T but not in CHAT. DNAs were isolated from mutant cell lines, quantified and digested with PstI, HindlII or EcoRI (for hprt analysis) or SacI (for tk analysis). DNAs were electrophoresed in 0.8% agarose gels, transferred to nitrocellulose and hybridized to the appropriate 32p-labeled probe (a 950-bp c D N A for hprt, obtained from C.T. Caskey, or a l l 0 0 - b p c D N A for tk, obtained from P. Deineger). A SacI polymorphism for tk allowed the distinction of the active and inactive alleles of this gene (Yandell et al., 1986). Results and discussion
Toxicity of A HA TK6 human lymphoblast cells were treated with various concentrations of AHA (0-0.1/~g/ml) for
~3
1.000~
2 ',
0100 t
i~
'N
•
0.007/~ 0.000
I 0.020
,
I 0.040
,
AHA c o n c e n t r a t i o n ,
I 0.060
,
/~9/ml
I 0.080
,
I -0.100
f o r 2 4 hr
Fig. 1. Toxicity of A H A to T K 6 h u m a n lymphoblast cells. Each point is the average of 2 independent Expts. done in duplicate.
24 h. Surviving fractions were measured immediately after treatment. Fig. 1 shows that A H A was toxic in the dose range 0-0.05 /~g/ml; however, little additional toxicity was observed between 0.05 and 0.10 ktg/ml.
Mutagenicity of AlIA Fig. 2 is a plot of the mutational dose response of TK6 cells exposed to A H A for 24 h. Approximately equal levels of both tk-NG mutants (normal growth rate of 16-18 h, colonies visible after 1 0 - 1 ! days incubation) and tk-SG mutants (slow growth rate of > 24 h, colonies visible after 18 days incubation) were induced. A slightly higher ~D o x
280
u>" 240
g g h
200 160
120 o
~ u -o c
80
400 000
I 0020
I 0.040
,
AHA c o n c e n t r o t i o n ,
I 0060
,
/zg/ml
I
O OBO
0
}lO0
for 24 hr
Fig. 2. Mutagenicity of A H A to T K 6 h u m a n lymphoblast cells. Induced mutation frequencies were determined for hprt (0, 6TGR), tk-NG (ll, T F T R colonies visible after 11 days incubation) and tk-SG ( - , T F T R colonies visible only after 7 days additional incubation), Each point is the average of duplicate determinations, which were independent of the experiments shown in Fig. 1. The background mutation frequencies were 3.6 × 10 -6 (6TG), 2.7 × 10 -6 (tk-NG) and 6.2 × 10 - 6 (tk-SG).
94 TABLE 1 Molecular Nature of Mutants Induced by A H A Locus
Dose (/~g/ml)
Induction level
tk-NG
0.10 0
6-7 ×
tk-SG
0.10 0
hprt
0.10 0.07 0
N u m b e r of mutants Point
Large-scale
19 10
3 41 a
4× -
4 4
6 116 a
60 × 35 X -
4 4 19
1 6 10
-
mutants show allele loss; see Table 1). In the set of 10 tk-SG mutants, there were 4 point mutants and 6 which had lost the active allele. The expectation was that 2 or 3 spontaneous mutants would
Induction level is the observed mutation frequency divided by the background. Point mutations were defined as no change in Southern blot pattern. a Data from Little et al. (1987).
~
