Mutation Research, 269 (1992) 141-148 © 1992 Elsevier Science Publishers B.V. All rights reselved 0027-5107/92/$05.00
141
MUT 05143
Forward mutations and DNA-protein crosslinks induced by ammonium metavanadate in cultured mammalian cells M i t c h e l l D. C o h e n , C a t h e r i n e B. Klein a n d M a x C o s t a Institute of Encironmental Medicine, New York Unicersity Medical Center, 550 First Avenue, New York, NY 10016, USA (Accepted 6 March 1992)
Keywords: Ammonium metavanadate; DNA-protein crosslink; Forward mutation; Potassium chromate
Summary Ammonium metavanadate yielded a dose-dependent increase in mutation frequency at the V79 hprt locus following a 24-h exposure period in serum-free FI2 medium. Vanadate also increased the mutation frequency of V79 cells by exposure of cells in salts-glucose medium, but these effects were not as striking, or as dose-dependent as they were in serum-free F12 medium. Ammonium metavanadate enhanced the mutation frequency in a V79 variant containing a transfected bacterial gpt gene. These cells are known to be more responsive to oxidative type mutations, and to mutations involving deletions. Although the absolute level of mutations was greater in these cells with ammonium metavanadate, so was the background, and these cells did not exhibit an enhanced mutagenic response to vanadate when compared to the wild-type V79 cells. The vanadate results were compared to a positive control potassium chromate, which exhibited a dose-dependent increase in mutation frequency. Ammonium metavanadate induced DNA-protein crosslinks formation in both Chinese hamster ovary and human MOLT4 cells, and the role of these relatively unrepaired genetic lesions in the mutations produced by vanadate and chromate are discussed.
Vanadium, a ubiquitous element found in air, water and soil, is both a physiologically and pharmacologically active substance (Jandhyala and Horn, 1983; Nechay et al., 1986a). It is widely distributed as an environmental pollutant in the form of oxides resulting from combustion of fossil fuels. Epidemiological studies have suggested a correlation between exposure to vanadiumbearing combustion particles and the incidence of
Correspondence: Mitchell D. Cohen, Ph.D., New York University, Department of Environmental Medicine, Tuxedo, NY 10987, USA.
human cancers (Stock, 1960; Hickey et al., 1967). However, unlike certain other metals (i.e. chromium, nickel), the potential genotoxic and carcinogenic effects of vanadium have not been as well studied. Data on the mutagenic potential of pentavalent vanadium in bacterial systems has been inconclusive (Hansen and Stern, 1984). Early studies in prokaryotic systems demonstrated that ammonium metavanadate (NH4VO3) and vanadium pentoxide (V205) were more genotoxic in recombination-repair-deficient (rec-) strains of Bacillus subtilis than in the wild-type rec + bacteria (Kanematsu et al., 1980). However, these same compounds were not mutagenic in several strains
142
of Escherichia coli (WP 2 or B / r W P 2) or Sabnonella typhimurium. Metavanadate was recently shown to increase both the convertant and revertant frequencies of the D7 trp- strain of Saccharomyces cerevisiae (Bronzetti et al., 1990). Much less is known about the genotoxic effects of vanadium in mammalian cells (NIOSH, 1977; Hansen and Stem, 1984). Nucleotide incorporation studies showed that pentavalent vanadium was mitogenic in isolated human T-lymphocytes (Ramanadham and Kern, 1981; Marini et al., 1987) and stimulated DNA synthesis in quiescent human or mouse fibroblasts (Carpenter, 1981; Smith, 1983). Later studies with human lymphocytes indicated that VO 3 treatment resulted in increased numbers of polyploid cells but decreased mitotic indices (Sharma and Talukder, 1987; Roldan and Altamirano, 1990). The appearance of mitogen-like proliferation was due to elevations in cellular DNA content without parallel increases in cell division (Bracken and Sharma, 1985) and was thought to arise from VOa-induced destabilization of cytoplasmic microtubules/ microfilaments (Wang and Choppin, 1981; Smith, 1983). A recent study in hamster cells indicated that in the absence of metabolic activation, NH4VO~ and other vanadium oxides were direct-acting clastogens and increased the levels of sister-ehromatid exchanges (SCE) (Owusu-Yaw et al., 1990). However, studies with cultured human lympho. cytes failed to demonstrate V(5+).induced DNA strandbreaks or increased SCE or chromosomal aberrations (McLean et al., 1982; Roldan and Altamirano, 1990). This ability to damage DNA suggests that in certain cells, V(5+) compounds might be mutagenic. We evaluated the mutagenic activity of NH4VO~ in two hamster-cell lines. Transgenic G12 cells have the capacity to detect both point mutations and deletions at the stably integrated bacterial xanthine-guanine phosphoribosyltransferase (gpt) locus, and are much more sensitive than V79 cells to X-ray mutagenesis (Klein and Rossman, 1990). V79 cells, from which GI2 ceils are derived, may be limited in their ability to detect intergenic deletions at the hprt locus. We also examined the ability of NH4VO 3 to induce DNA-protein crosslink complexes (DPC) in both hamster and human cells. As the
predominant anionic form of V(5+), VO 3 possesses similarities in both cellular uptake mechanisms (Cantley and Alsen, 1979) and intracellular redox behavior (Macara et al., 1980) with hexavalent chromate (Cr042-; Cr[VI]), which was selected as the positive control. Materials and methods
Cells and culture Chinese hamster ovary (CHO) cells were maintained in complete minimal essential medium (10% fetal bovine serum (FBS), 1 mM L-glutamine, and 100 U / m l penicillin and 100 ~ g / m l streptomycin). Hamster V79 fetal lung fibroblasts and the V79-derived hprt-/gpt ÷ mutant G12 line were maintained in complete Ham's F12 medium supplemented with 5% FBS. Cell monolayers were passaged by trypsinization every 3-5 d. Human leukemic T-lymphocyte (MOLT4) suspensions were maintained in complete RPMI 1640 plus 10% heat-inactivated FBS. All lines were maintained at 37°C in 5% CO2-95% air. Chemicals K2CrO 4 and NH4VO~ were obtained from J.T. Baker Chemicals (Philipsburg, NJ). Cell culture media, FBS, and supplements were from Hazelton Research Products (Denver, PA) and Gibco (Grand Island, NY). 6-Thioguanine (6-TG) and chemicals for assessing DPC were purchased from Sigma Chemical (St. Louis, MO). [3H]Deoxy. thymidine (TdR) and [aSS]methioninc were from NEN Products (Boston, MA). Cytotoxicity assay Cell numbers were routinely determined with a Coulter counter. Trypsinized cells were seeded into five 60-ram dishes at 300 cells/dish. After a minimum 3-h attachment period, the medium was removed and replaced with serum-free medium or salts-glucose medium (SGM: 50 mM HEPES/100 mM NaCI/5 mM KCI/2 mM CaCl2/5 mM dextrose) containing freshly prepared dissolved metal solutions. V79-derived G12 cells were incubated in a Mg2+-containing Earle's balanced salt solution (EBSS) instead of SGM. After 2 or 24 h, the medium was removed, cells were washed gently with saline A and then placed
143
in complete medium. After 7 d, the cells were fLxed, and stained with 0.5% crystal violet solution. Cytotoxicity was calculated as the percentage of surviving colonies in metal-treated dishes compared with colonies present in plates receiving only medium.
Mutagenicity assay Triplicate 100-mm dishes were each seeded with 4 × 105 cells and exposed (vide supra). To assure an adequate cell survival at higher doses, additional plates were utilized. After 7 d, the monolayers were trypsinized, cells were pooled, and fifteen 100-mm dishes/treatment were reseeded at 2 x 105 cells/dish with 10-ml complete medium containing 6-TG (10/~g/ml). Five 60-mm dishes containing complete medium without 6-TG were also seeded with 300 treated cells to assess plating efficiency (PE). PE dishes or 6-TG selection plates were incubated for 7 or 15 d, respectively, before colonies were fixed and stained. The mutant frequency was derived from the number of resistant colonies obtained after accounting for PE, and was expressed as the number of induced mutants per 106 colony-forming cells.
DNA-protein complexes CHO cells seeded in 150-ram dishes were allowed to grow to 50% confluency. The medium was replaced with medium containing 1 /zCi/ml [3H]TdR and [35S]Met for 18 h, followed by 2-h treatment medium without radiolabel. Cells (circa 80% frequency), were rinsed with saline A and then exposed with metaI-SGM medium. After 2 or 24 h exposure, plates were rinsed and cells collected by scraping. Pooled cells were rinsed twice prior to DPC isolation to avoid additional metal complexation after the cells were ruptured. MOLT4 cells seeded into T-225 culture flasks at 4 x 105 cells/ml RPMI 1640 and incubated for 3 d with [aH]TdR and [35S]Met (0.01 /.~Ci/mi) prior to metal treatment. Labelled cells were placed in T-150 flasks (8 x 105 cells/ml) containing SGM or SGM-metal solution and incubated for 2 or 24 h. Cell viability was estimated by the exclusion of trypan blue dye. The isolation of DNA-protein complexes from intact cells was carried out as previously described (Miller and Costa, 1988). Cells were re-
suspended in a buffer containing 10 mM Tris, 10 mM NaCI, 1.5 mM MgCI2, 1 mM PMSF for 15 min at ice-bath temperatures. Cells were collected by centrifugation and resuspended in buffer containing 0.5% NP-40 for homogenization and nuclei isolation. Nuclei were placed in 5 mM Tris/1 mM PMSF/2% SDS solution, homogenized, and centrifuged at 100000 x g for 16 h. Isolated DPC were resuspended and processed as above in 5 M urea-Tris-PMSF buffer to further purify DPC from free protein. Total DPC recovery (based on the protein/DNA ratio) was estimated from the radioactivity recovered in the purified DPC pellet after acetone precipitation at -20°C. The two-tailed Student's t-test was used to analyze the statistical significance of formation of DPC. Results
Fig. 1 shows the cytotoxicity of ammonium metavanadate in salts-glucose medium and in serum-free culture media to V79 cells. Note that vanadate is more toxic in the salts-glucose medium, and that a longer exposure time yields enhanced cytotoxicity in this medium. Fig. 2 shows the induction of mutations by ammonium metavanadate in V79 cells exposed in serum-free F12 or SGM. Ammonium metavanadate gave a dose120
~]~
loo ~
6O
4O 2O 0 0
I
i
i
~
I
I
I
~.
5
10
1S
20
25
30
35
no
,
I
45
Melal Concentration ( , M )
Fig. I. Vanadate-induced cytotoxicity. V79 cells were exposed to the various levels of NH4VO3 in SGM Iopen symbols) or serum-free F12 (closed symbols) for 2 h ( ~ ) or 24 h (e, O) as described in the Materials and methods. Cultures were fixed and stained on day 8; the numbers of colonies/dish were counted. Data are the mean +_SD of at least 2 independent Expts.
144 10
la 16 w
8
=
)4
8
"o
T
s
i. j
, =w
4
1
r
. ,
0 0
I 10
i
I
,
20
i
t
SO
40
Metal Concentration
i
I
i
50
I 60
0
I O
i
I
10
(pM)
i
20 Metal
Concentration
I
t
I
30
40
50
(pM)
Fig. 2. (A) Mutation induction in V79 cells by NH4VO 3 in serum-free F12 medium. Cells were exposed for 24 h to the indicated doses of NH4VO 3 in serum-free F12 medium as described in Materials and methods. Cells were harvested and replated in 6-TG-containing F12 and colonies counted after fixation and staining on day 15. Each point is the mean MF ( + SD) from at least 2 separate Expts. calculated from the total number of mutant colonies in 15 plates/dose. (B) Mutation induction in V79 cells by NH4VO 3 in SGM. Cells were exposed for 2 h (solid line) or 24 h (dashed line) to the indicated levels of NH4VO 3 in SGM as described in the Materials and Methods.
dependent incre',se in mutation frequency at concentrations tltat were relatively non-cytotoxic in serum-free F12 (Fig. 2A) but the mutation frequency decreased at cytotoxic concentrations. As a companion to Fig. 2A, Table 1 contains the actrtal numbers of mutant colonies and the survival relative to the mutation for one representative experiment, Fig. 2B shows the mutation frequency in salts-glucose media as a function of increasing metal concentrations with representa. tire data outlined in Tables 2A and 2B. While
some mutations could be detected in SGM, in general the optimal mutation responses (i.e. 6-7fold above background) were seen with vanadium in serum-free F12 medium. To assess whether the V79 cells may be limited in their ability to detect certain types of mutations that may be caused by ammonium metavanadate, a cell line more sensitive to oxidation and deletion mutagenesis was utilized. Fig. 3 illustrates the mutagenicity and cytotoxicity of chromate, the positive control in the V79 cell
TABLE 1 PLATING EFFICIENCIES AND TOTAL COLONIES RECOVERED IN V79 CELLS TREATED FOR 24 h WITH NH4VO 3 1N SERUM.FREE MEDIUM NH4VO 3 (pM)
CE"
Survival (% control)
Number of mutant colonies b (PE)
MF c
0 5 10 20 25 50
83.0 81.7 75.0 31.7 16.3 0.5
I00 98.4 90.4 38.2 19.6 < O.l
7 (0,92) 38 (0.74) 22 (0.82) 25 [0,90) 3 (0.86)
2.1:1:0.7 13.1 +5.7 8.0 + 1.3 9.7 + 4.2 1.2 :l: 0.02
a Average cloning efficiency immediately after metal treatment (from 3 Expts,), b Total colonies obtained from 15 plates per treatment in 1 Expt. Number in parentheses is the reseeding plating efficiency of the treated cells at the time of seeding in selection, c Mutant frequency (calculated as described in Materials and methods) :I:SD of the values obtained from the separate experiments. Data is graphically displayed in Fig. 2A,
145 TABLE 2 PLATING EFFICIENCIES AND TOTAL COLONIES RECOVERED IN V79 CELLS TREATED FOR 2 OR 24 h WITH NH4VO 3 IN SERUM-FREE SGM MEDIUM
(A) 2-h treatment NH4VO 3 (p.M)
CE a
Survival (% control)
Number of mutant colonies b (PE)
0 5 10 20 40
85.5 82.5 71.3 67.9 55.2
100 96.5 83.4 79.4 64.6
NH4VO 3 (~M)
CE a
Su~ival (% control)
Number of mutant colonies b (PE)
MF c
0 5 10 20 40
86.6 79.9 11.9 0.5 0
100 92.3 13.7 < 0.1 < 0.1
11 (0.92) 14 (0.89) 11 (0.82) 16 (0.84) 0 (0.76)
4.6 +_0.9 5.7 +_0.6 5.2_+ 1.1 6.7 +_0.5 0
5 (0.88) 2 (0.94) 5 (0.91) 9 (0.96) 7 (0.83)
MF c 2.2-+ 0.5 0.7 + 0.02 1.9 -+0.1 3.6-+ 0.6 3.6 __-4-1.1
(B) 24-h treatment
a Average cloning efficiency immediately after metal treatment (from 2 Expts.). b Total colonies obtained from 15 plates per treatment in 1 Expt. Number in parentheses is the reseeding plating efficiency of the treated cells at the time of selection. c Mutant frequency (calculated as described in Materials and methods) + SD of the values obtained from the separate experiments. This data is graphically displayed in Fig. 2B.
line, while Fig. 4 contrasts the cytotoxicity and mutant frequency of ammonium metavanadate in this G12 cell line. While the actual absolute levels of mutations were higher in this cell line, the uninduced mutant background was also higher, and the vanadate-induced increase in mutant frequency was not as large as in V79 cells. However,
these results support the mutagenic activity of vanadate compounds. A predominant genetic lesion induced by chromate is the DNA-protein crosslink. To assess whether ammonium metavanadate would induce similar types of damage, the formation of this lesion was examined in cells treated with this
TABLE 3 PLATING EFFICIENCIES AND TOTAL COLONIES RECOVERED IN GI2 CELLS TREATED FOR 24 h WITH NH4VO~ IN SERUM-FREE EBSS MEDIUM NH4VO 3 (/~M)
CE a
Survival (% control)
Number of mutant colonies b (PE)
MF ~
0 5 10 20 40
99.2 67.8 12.4 1.7 0.5
100 68.3 12.5 1.7 < 0.1
58 (0.92) 59 (0.89) 62 (0.87) 105 (0.84) 107 (0.93)
24.74- 5.3 27.3 4- 7.4 38.9+ 12.4 55.7 + 19.8 53.5 4. 21.4
a Average cloning efficiency immediately after metal treatment (from 2 Expts.). b Total colonies obtained from 15 plates per treatment in I Expt. Number in parentheses is the reseeding plating efficiency of the treated cells at the time of selection. c Mutant frequency (calculated as described in Materials and methods) + S D of the values obtained from the separate experiments, This data is graphically displayed in Fig. 4.
146 100,
120
LlS
100 T
80
e0
~
60
~ ~
o
w
o
60
o o
~.
~ O
80
g
60
~" 40
40
-
2O
'°I
0 0.00
:E '
~
'
2.50
5.00
7.50
0
0
Discussion The purpose of the present study was to determine whether vanadate (VO;) ions could induce
'
20
Y
30
40
~
0
50
Metal Concentration (pM)
Metal Concentration (pM)
agent. Fig. 5 compares the induction of DNA protein-crosslinks at several exposure time intervals, and at several dosage levels in CHO (Fig. 5A) and in MOLT4 cells (Fig. 5B). The CHO cells readily exhibited a significant time- and dose-dependent increase in DNA-protein crosslink formation by ammonium metavanadate, while DPC formation in MOLT4 cells was less apparent. In both cell lines, the response was again not as large as that exhibited by chromate.
•
10
10.00
Fig. 3. Chromate genotoxicity in V79 cells. Relative cytotoxicity (z~, mean + SD of 5 dishes/point) and mutation induction ( A ) in cells treated with K2CrO 4 in SGM (results are from 2 separate Expts.).
20
20
Fig. 4. Genotoxicity of NHAVO~ in 1312 cells. Comparative cytotoxicity of NH4VO 3 !n V79 ( + ) or G12 ( o ) cell lines following 24-h exposures in serum-free EBSS. 6-TG mutation induction (e) in G12 cells treated for 24 h with VO3-EBSS. Toxicity data points are the average + SD of 5 plates/dose and MF values are the average from 2 separate Expts.
forward mutations, possibly as a result of DPC, or other genetic lesions, in cultured mammalian cells. Our results demonstrate that VO;, like CrO4z-, is potently cytotoxic, although the kinetics of toxicity for each metal anion are quite different. Vanadate requires 24 h to achieve dose-dependent cytotoxicity whereas CrO~- exhibits dose-dependency after only a 2-h exposure. This may be due to a more rapid uptake and/or
O.nO
1,so, ~ 1,40 t
0.00
I
g¢
:
(P
141
MUT 05143
Forward mutations and DNA-protein crosslinks induced by ammonium metavanadate in cultured mammalian cells M i t c h e l l D. C o h e n , C a t h e r i n e B. Klein a n d M a x C o s t a Institute of Encironmental Medicine, New York Unicersity Medical Center, 550 First Avenue, New York, NY 10016, USA (Accepted 6 March 1992)
Keywords: Ammonium metavanadate; DNA-protein crosslink; Forward mutation; Potassium chromate
Summary Ammonium metavanadate yielded a dose-dependent increase in mutation frequency at the V79 hprt locus following a 24-h exposure period in serum-free FI2 medium. Vanadate also increased the mutation frequency of V79 cells by exposure of cells in salts-glucose medium, but these effects were not as striking, or as dose-dependent as they were in serum-free F12 medium. Ammonium metavanadate enhanced the mutation frequency in a V79 variant containing a transfected bacterial gpt gene. These cells are known to be more responsive to oxidative type mutations, and to mutations involving deletions. Although the absolute level of mutations was greater in these cells with ammonium metavanadate, so was the background, and these cells did not exhibit an enhanced mutagenic response to vanadate when compared to the wild-type V79 cells. The vanadate results were compared to a positive control potassium chromate, which exhibited a dose-dependent increase in mutation frequency. Ammonium metavanadate induced DNA-protein crosslinks formation in both Chinese hamster ovary and human MOLT4 cells, and the role of these relatively unrepaired genetic lesions in the mutations produced by vanadate and chromate are discussed.
Vanadium, a ubiquitous element found in air, water and soil, is both a physiologically and pharmacologically active substance (Jandhyala and Horn, 1983; Nechay et al., 1986a). It is widely distributed as an environmental pollutant in the form of oxides resulting from combustion of fossil fuels. Epidemiological studies have suggested a correlation between exposure to vanadiumbearing combustion particles and the incidence of
Correspondence: Mitchell D. Cohen, Ph.D., New York University, Department of Environmental Medicine, Tuxedo, NY 10987, USA.
human cancers (Stock, 1960; Hickey et al., 1967). However, unlike certain other metals (i.e. chromium, nickel), the potential genotoxic and carcinogenic effects of vanadium have not been as well studied. Data on the mutagenic potential of pentavalent vanadium in bacterial systems has been inconclusive (Hansen and Stern, 1984). Early studies in prokaryotic systems demonstrated that ammonium metavanadate (NH4VO3) and vanadium pentoxide (V205) were more genotoxic in recombination-repair-deficient (rec-) strains of Bacillus subtilis than in the wild-type rec + bacteria (Kanematsu et al., 1980). However, these same compounds were not mutagenic in several strains
142
of Escherichia coli (WP 2 or B / r W P 2) or Sabnonella typhimurium. Metavanadate was recently shown to increase both the convertant and revertant frequencies of the D7 trp- strain of Saccharomyces cerevisiae (Bronzetti et al., 1990). Much less is known about the genotoxic effects of vanadium in mammalian cells (NIOSH, 1977; Hansen and Stem, 1984). Nucleotide incorporation studies showed that pentavalent vanadium was mitogenic in isolated human T-lymphocytes (Ramanadham and Kern, 1981; Marini et al., 1987) and stimulated DNA synthesis in quiescent human or mouse fibroblasts (Carpenter, 1981; Smith, 1983). Later studies with human lymphocytes indicated that VO 3 treatment resulted in increased numbers of polyploid cells but decreased mitotic indices (Sharma and Talukder, 1987; Roldan and Altamirano, 1990). The appearance of mitogen-like proliferation was due to elevations in cellular DNA content without parallel increases in cell division (Bracken and Sharma, 1985) and was thought to arise from VOa-induced destabilization of cytoplasmic microtubules/ microfilaments (Wang and Choppin, 1981; Smith, 1983). A recent study in hamster cells indicated that in the absence of metabolic activation, NH4VO~ and other vanadium oxides were direct-acting clastogens and increased the levels of sister-ehromatid exchanges (SCE) (Owusu-Yaw et al., 1990). However, studies with cultured human lympho. cytes failed to demonstrate V(5+).induced DNA strandbreaks or increased SCE or chromosomal aberrations (McLean et al., 1982; Roldan and Altamirano, 1990). This ability to damage DNA suggests that in certain cells, V(5+) compounds might be mutagenic. We evaluated the mutagenic activity of NH4VO~ in two hamster-cell lines. Transgenic G12 cells have the capacity to detect both point mutations and deletions at the stably integrated bacterial xanthine-guanine phosphoribosyltransferase (gpt) locus, and are much more sensitive than V79 cells to X-ray mutagenesis (Klein and Rossman, 1990). V79 cells, from which GI2 ceils are derived, may be limited in their ability to detect intergenic deletions at the hprt locus. We also examined the ability of NH4VO 3 to induce DNA-protein crosslink complexes (DPC) in both hamster and human cells. As the
predominant anionic form of V(5+), VO 3 possesses similarities in both cellular uptake mechanisms (Cantley and Alsen, 1979) and intracellular redox behavior (Macara et al., 1980) with hexavalent chromate (Cr042-; Cr[VI]), which was selected as the positive control. Materials and methods
Cells and culture Chinese hamster ovary (CHO) cells were maintained in complete minimal essential medium (10% fetal bovine serum (FBS), 1 mM L-glutamine, and 100 U / m l penicillin and 100 ~ g / m l streptomycin). Hamster V79 fetal lung fibroblasts and the V79-derived hprt-/gpt ÷ mutant G12 line were maintained in complete Ham's F12 medium supplemented with 5% FBS. Cell monolayers were passaged by trypsinization every 3-5 d. Human leukemic T-lymphocyte (MOLT4) suspensions were maintained in complete RPMI 1640 plus 10% heat-inactivated FBS. All lines were maintained at 37°C in 5% CO2-95% air. Chemicals K2CrO 4 and NH4VO~ were obtained from J.T. Baker Chemicals (Philipsburg, NJ). Cell culture media, FBS, and supplements were from Hazelton Research Products (Denver, PA) and Gibco (Grand Island, NY). 6-Thioguanine (6-TG) and chemicals for assessing DPC were purchased from Sigma Chemical (St. Louis, MO). [3H]Deoxy. thymidine (TdR) and [aSS]methioninc were from NEN Products (Boston, MA). Cytotoxicity assay Cell numbers were routinely determined with a Coulter counter. Trypsinized cells were seeded into five 60-ram dishes at 300 cells/dish. After a minimum 3-h attachment period, the medium was removed and replaced with serum-free medium or salts-glucose medium (SGM: 50 mM HEPES/100 mM NaCI/5 mM KCI/2 mM CaCl2/5 mM dextrose) containing freshly prepared dissolved metal solutions. V79-derived G12 cells were incubated in a Mg2+-containing Earle's balanced salt solution (EBSS) instead of SGM. After 2 or 24 h, the medium was removed, cells were washed gently with saline A and then placed
143
in complete medium. After 7 d, the cells were fLxed, and stained with 0.5% crystal violet solution. Cytotoxicity was calculated as the percentage of surviving colonies in metal-treated dishes compared with colonies present in plates receiving only medium.
Mutagenicity assay Triplicate 100-mm dishes were each seeded with 4 × 105 cells and exposed (vide supra). To assure an adequate cell survival at higher doses, additional plates were utilized. After 7 d, the monolayers were trypsinized, cells were pooled, and fifteen 100-mm dishes/treatment were reseeded at 2 x 105 cells/dish with 10-ml complete medium containing 6-TG (10/~g/ml). Five 60-mm dishes containing complete medium without 6-TG were also seeded with 300 treated cells to assess plating efficiency (PE). PE dishes or 6-TG selection plates were incubated for 7 or 15 d, respectively, before colonies were fixed and stained. The mutant frequency was derived from the number of resistant colonies obtained after accounting for PE, and was expressed as the number of induced mutants per 106 colony-forming cells.
DNA-protein complexes CHO cells seeded in 150-ram dishes were allowed to grow to 50% confluency. The medium was replaced with medium containing 1 /zCi/ml [3H]TdR and [35S]Met for 18 h, followed by 2-h treatment medium without radiolabel. Cells (circa 80% frequency), were rinsed with saline A and then exposed with metaI-SGM medium. After 2 or 24 h exposure, plates were rinsed and cells collected by scraping. Pooled cells were rinsed twice prior to DPC isolation to avoid additional metal complexation after the cells were ruptured. MOLT4 cells seeded into T-225 culture flasks at 4 x 105 cells/ml RPMI 1640 and incubated for 3 d with [aH]TdR and [35S]Met (0.01 /.~Ci/mi) prior to metal treatment. Labelled cells were placed in T-150 flasks (8 x 105 cells/ml) containing SGM or SGM-metal solution and incubated for 2 or 24 h. Cell viability was estimated by the exclusion of trypan blue dye. The isolation of DNA-protein complexes from intact cells was carried out as previously described (Miller and Costa, 1988). Cells were re-
suspended in a buffer containing 10 mM Tris, 10 mM NaCI, 1.5 mM MgCI2, 1 mM PMSF for 15 min at ice-bath temperatures. Cells were collected by centrifugation and resuspended in buffer containing 0.5% NP-40 for homogenization and nuclei isolation. Nuclei were placed in 5 mM Tris/1 mM PMSF/2% SDS solution, homogenized, and centrifuged at 100000 x g for 16 h. Isolated DPC were resuspended and processed as above in 5 M urea-Tris-PMSF buffer to further purify DPC from free protein. Total DPC recovery (based on the protein/DNA ratio) was estimated from the radioactivity recovered in the purified DPC pellet after acetone precipitation at -20°C. The two-tailed Student's t-test was used to analyze the statistical significance of formation of DPC. Results
Fig. 1 shows the cytotoxicity of ammonium metavanadate in salts-glucose medium and in serum-free culture media to V79 cells. Note that vanadate is more toxic in the salts-glucose medium, and that a longer exposure time yields enhanced cytotoxicity in this medium. Fig. 2 shows the induction of mutations by ammonium metavanadate in V79 cells exposed in serum-free F12 or SGM. Ammonium metavanadate gave a dose120
~]~
loo ~
6O
4O 2O 0 0
I
i
i
~
I
I
I
~.
5
10
1S
20
25
30
35
no
,
I
45
Melal Concentration ( , M )
Fig. I. Vanadate-induced cytotoxicity. V79 cells were exposed to the various levels of NH4VO3 in SGM Iopen symbols) or serum-free F12 (closed symbols) for 2 h ( ~ ) or 24 h (e, O) as described in the Materials and methods. Cultures were fixed and stained on day 8; the numbers of colonies/dish were counted. Data are the mean +_SD of at least 2 independent Expts.
144 10
la 16 w
8
=
)4
8
"o
T
s
i. j
, =w
4
1
r
. ,
0 0
I 10
i
I
,
20
i
t
SO
40
Metal Concentration
i
I
i
50
I 60
0
I O
i
I
10
(pM)
i
20 Metal
Concentration
I
t
I
30
40
50
(pM)
Fig. 2. (A) Mutation induction in V79 cells by NH4VO 3 in serum-free F12 medium. Cells were exposed for 24 h to the indicated doses of NH4VO 3 in serum-free F12 medium as described in Materials and methods. Cells were harvested and replated in 6-TG-containing F12 and colonies counted after fixation and staining on day 15. Each point is the mean MF ( + SD) from at least 2 separate Expts. calculated from the total number of mutant colonies in 15 plates/dose. (B) Mutation induction in V79 cells by NH4VO 3 in SGM. Cells were exposed for 2 h (solid line) or 24 h (dashed line) to the indicated levels of NH4VO 3 in SGM as described in the Materials and Methods.
dependent incre',se in mutation frequency at concentrations tltat were relatively non-cytotoxic in serum-free F12 (Fig. 2A) but the mutation frequency decreased at cytotoxic concentrations. As a companion to Fig. 2A, Table 1 contains the actrtal numbers of mutant colonies and the survival relative to the mutation for one representative experiment, Fig. 2B shows the mutation frequency in salts-glucose media as a function of increasing metal concentrations with representa. tire data outlined in Tables 2A and 2B. While
some mutations could be detected in SGM, in general the optimal mutation responses (i.e. 6-7fold above background) were seen with vanadium in serum-free F12 medium. To assess whether the V79 cells may be limited in their ability to detect certain types of mutations that may be caused by ammonium metavanadate, a cell line more sensitive to oxidation and deletion mutagenesis was utilized. Fig. 3 illustrates the mutagenicity and cytotoxicity of chromate, the positive control in the V79 cell
TABLE 1 PLATING EFFICIENCIES AND TOTAL COLONIES RECOVERED IN V79 CELLS TREATED FOR 24 h WITH NH4VO 3 1N SERUM.FREE MEDIUM NH4VO 3 (pM)
CE"
Survival (% control)
Number of mutant colonies b (PE)
MF c
0 5 10 20 25 50
83.0 81.7 75.0 31.7 16.3 0.5
I00 98.4 90.4 38.2 19.6 < O.l
7 (0,92) 38 (0.74) 22 (0.82) 25 [0,90) 3 (0.86)
2.1:1:0.7 13.1 +5.7 8.0 + 1.3 9.7 + 4.2 1.2 :l: 0.02
a Average cloning efficiency immediately after metal treatment (from 3 Expts,), b Total colonies obtained from 15 plates per treatment in 1 Expt. Number in parentheses is the reseeding plating efficiency of the treated cells at the time of seeding in selection, c Mutant frequency (calculated as described in Materials and methods) :I:SD of the values obtained from the separate experiments. Data is graphically displayed in Fig. 2A,
145 TABLE 2 PLATING EFFICIENCIES AND TOTAL COLONIES RECOVERED IN V79 CELLS TREATED FOR 2 OR 24 h WITH NH4VO 3 IN SERUM-FREE SGM MEDIUM
(A) 2-h treatment NH4VO 3 (p.M)
CE a
Survival (% control)
Number of mutant colonies b (PE)
0 5 10 20 40
85.5 82.5 71.3 67.9 55.2
100 96.5 83.4 79.4 64.6
NH4VO 3 (~M)
CE a
Su~ival (% control)
Number of mutant colonies b (PE)
MF c
0 5 10 20 40
86.6 79.9 11.9 0.5 0
100 92.3 13.7 < 0.1 < 0.1
11 (0.92) 14 (0.89) 11 (0.82) 16 (0.84) 0 (0.76)
4.6 +_0.9 5.7 +_0.6 5.2_+ 1.1 6.7 +_0.5 0
5 (0.88) 2 (0.94) 5 (0.91) 9 (0.96) 7 (0.83)
MF c 2.2-+ 0.5 0.7 + 0.02 1.9 -+0.1 3.6-+ 0.6 3.6 __-4-1.1
(B) 24-h treatment
a Average cloning efficiency immediately after metal treatment (from 2 Expts.). b Total colonies obtained from 15 plates per treatment in 1 Expt. Number in parentheses is the reseeding plating efficiency of the treated cells at the time of selection. c Mutant frequency (calculated as described in Materials and methods) + SD of the values obtained from the separate experiments. This data is graphically displayed in Fig. 2B.
line, while Fig. 4 contrasts the cytotoxicity and mutant frequency of ammonium metavanadate in this G12 cell line. While the actual absolute levels of mutations were higher in this cell line, the uninduced mutant background was also higher, and the vanadate-induced increase in mutant frequency was not as large as in V79 cells. However,
these results support the mutagenic activity of vanadate compounds. A predominant genetic lesion induced by chromate is the DNA-protein crosslink. To assess whether ammonium metavanadate would induce similar types of damage, the formation of this lesion was examined in cells treated with this
TABLE 3 PLATING EFFICIENCIES AND TOTAL COLONIES RECOVERED IN GI2 CELLS TREATED FOR 24 h WITH NH4VO~ IN SERUM-FREE EBSS MEDIUM NH4VO 3 (/~M)
CE a
Survival (% control)
Number of mutant colonies b (PE)
MF ~
0 5 10 20 40
99.2 67.8 12.4 1.7 0.5
100 68.3 12.5 1.7 < 0.1
58 (0.92) 59 (0.89) 62 (0.87) 105 (0.84) 107 (0.93)
24.74- 5.3 27.3 4- 7.4 38.9+ 12.4 55.7 + 19.8 53.5 4. 21.4
a Average cloning efficiency immediately after metal treatment (from 2 Expts.). b Total colonies obtained from 15 plates per treatment in I Expt. Number in parentheses is the reseeding plating efficiency of the treated cells at the time of selection. c Mutant frequency (calculated as described in Materials and methods) + S D of the values obtained from the separate experiments, This data is graphically displayed in Fig. 4.
146 100,
120
LlS
100 T
80
e0
~
60
~ ~
o
w
o
60
o o
~.
~ O
80
g
60
~" 40
40
-
2O
'°I
0 0.00
:E '
~
'
2.50
5.00
7.50
0
0
Discussion The purpose of the present study was to determine whether vanadate (VO;) ions could induce
'
20
Y
30
40
~
0
50
Metal Concentration (pM)
Metal Concentration (pM)
agent. Fig. 5 compares the induction of DNA protein-crosslinks at several exposure time intervals, and at several dosage levels in CHO (Fig. 5A) and in MOLT4 cells (Fig. 5B). The CHO cells readily exhibited a significant time- and dose-dependent increase in DNA-protein crosslink formation by ammonium metavanadate, while DPC formation in MOLT4 cells was less apparent. In both cell lines, the response was again not as large as that exhibited by chromate.
•
10
10.00
Fig. 3. Chromate genotoxicity in V79 cells. Relative cytotoxicity (z~, mean + SD of 5 dishes/point) and mutation induction ( A ) in cells treated with K2CrO 4 in SGM (results are from 2 separate Expts.).
20
20
Fig. 4. Genotoxicity of NHAVO~ in 1312 cells. Comparative cytotoxicity of NH4VO 3 !n V79 ( + ) or G12 ( o ) cell lines following 24-h exposures in serum-free EBSS. 6-TG mutation induction (e) in G12 cells treated for 24 h with VO3-EBSS. Toxicity data points are the average + SD of 5 plates/dose and MF values are the average from 2 separate Expts.
forward mutations, possibly as a result of DPC, or other genetic lesions, in cultured mammalian cells. Our results demonstrate that VO;, like CrO4z-, is potently cytotoxic, although the kinetics of toxicity for each metal anion are quite different. Vanadate requires 24 h to achieve dose-dependent cytotoxicity whereas CrO~- exhibits dose-dependency after only a 2-h exposure. This may be due to a more rapid uptake and/or
O.nO
1,so, ~ 1,40 t
0.00
I
g¢
:
(P
