Molecular dosimetry studies of forward mutation induced at the yg2 locus in maize by ethyl methanesulfonate

Mutation Research, 211 (1989) 231-241

231

Elsevier MTR 04726

Molecular dosimetry studies of forward mutation induced at the yg2 locus in maize by ethyl methanesulfonate W i l l i a m E. S c h y * a n d M i c h a e l J. P l e w a Institute for Environmental Studies, University of Illinois at Urbana-Champaign, Urbana, IL 61801 (U.S.A.)

(Received 3 March 1988) (Revision received11 October 1988) (Accepted 13 October 1988)

Keywords: Moleculardosirnetry; Forward mutation; Zea mays, yg2; Ethyl methanesulphonate

Summary The yg2 assay in Zea mays detects forward mutation in somatic cells within leaf primordia of embryos and it was used in an analysis of the molecular dosimetry of ethyl methanesulfonate (EMS). Parallel genetic and molecular dosimetry experiments were conducted in which the frequency of forward mutation and the level of covalently bound ethyl D N A adducts were determined. Prepared kernels were treated for 8 h at 2 0 ° C with 1-10 mM EMS. EMS induced a direct concentration-dependent increase in mutation induction proportional to the exposure concentration (slope = 0.93). The kinetics of mutation induction demonstrated in the intact maize system were consistent with the kinetics observed earlier in in vitro model systems using cultured mammalian cells, and contrasted with the exponential increase in mutation induction characteristic of microbial species. Parallel molecular dosimetry experiments were conducted using [3H]EMS. D N A was extracted and purified from embryonic tissues containing the leaf primordia, the target tissue of the yg2 assay. A linear increase in the molecular dose was observed as a function of EMS concentration. Using concentration as a common parameter between the parallel genetic and dosimetry studies, mutation induction appeared to increase nearly in a direct proportion to the molecular dose. However, studies in other genetic systems indicate that the levels of specific D N A adducts, such as O6-ethylguanine (Ot-EtGua) show a better correlation with mutation induction kinetics than molecular dose. Neither molecular dose, nor O6-EtGua levels account for differences in the absolute frequencies of mutation induction observed in different genetic systems. Therefore, reliable assessment of health risks posed to humans by chemical mutagens appears to require consideration of other factors in addition to D N A dose or adduct formation, including differences in repair capabilities and in the size of the genetic targets in humans relative to the model genetic systems under study.

Correspondence: Dr. Michael J. Plewa, University of Illinois, 305 Environmental Research Laboratory, 1005 W. Western Avenue, Urbana, IL 61801 (U.S.A.). * Present address: York University, Department of Biology, 4700 Keele Street, Toronto, Ont., M3J 1P3 (Canada).

The chemical ethyl methanesulfonate (EMS) (CAS No. 62-50-0) has been widely used to induce high frequencies of mutation in a variety of genetic systems (Sega, 1984). Because of its frequent use in mutagenicity studies and as a representative direct-acting alkylating agent, EMS was selected

0027-5107/89/$03.50 © 1989 ElsevierSciencePublishers B.V. (BiomedicalDivision)

232

for use in quantitative comparative mutagenesis studies (Aaron and Lee, 1978). Such studies are essential for genetic risk assessment, since in order to extrapolate data obtained in experimental organisms to humans, it must be shown that reliable predictions of genetic effects in one organism can be made on the basis of data obtained in others (Lee, 1976). The variation in metabolism and physiology between the different genetic systems makes it desirable to find a common parameter to use as a basis of comparison among the different systems. It was suggested that comparisons be made on the basis of the molecular dose (Lee, 1976; Aaron, 1976). The molecular dose is defined as the amount of agent reaching the DNA of the target cells of the genetic assay. For ionizing radiation the exposure and molecular dose are nearly identical, while for chemical mutagens, these may differ quite substantially (Lee, 1976). Studies of the molecular dosimetry of EMS were conducted in several different organisms, including Escherichia coli, V79 Chinese hamster ovary (CHO) cells, L5178Y mouse lymphoma cells (Aaron et al., 1980), Drosophila melanogaster (Aaron and Lee, 1978), Neurospora crassa and Saccharomyces cerevisiae (van Zeeland et al., 1983). The genetic endpoint was limited to forward mutation and parallel experiments were conducted in which [3H]EMS was used and the amount of covalently bound ethyl groups to D N A was determined. Previous molecular dosimetry studies in microbial organisms ( E. coli, N. crassa and S. cerevisiae ) all demonstrated exponential increases in mutation induction that could not be accounted for on the basis of molecular dose. Cultured V79 and mouse lymphoma cell lines were used as model systems to study the molecular dosimetry of EMS in higher eukaryotic cells. These cell lines exhibited a linear increase in mutation induction proportional to the molecular dose, though the highest concentrations tested in these organisms were significantly lower than in the microbes, due to toxicity. In addition to the contrast in mutation-induction kinetics, the absolute frequencies of mutation per unit ethylation, measured at exposures between 1 and 10 mM EMS, appeared to differ between the V79 cells and the microbial species. However, the V79 cells demonstrated a

mutation frequency per unit ethylation approximately an order of magnitude greater than even the L5178Y mouse lymphoma cell line. Because of the difference in the results obtained between the two mammalian cell lines, it was uncertain whether the sensitivity to alkylation, expressed as the induced mutation frequency per unit alkylation, might be similar in all organisms or distinctly increased in higher eukaryotic cells. This is an important theoretical question, both in terms of understanding the factors that ultimately determine the observed mutation frequency, and in terms of risk assessment. The sensitivity of higher eukaryotic cells to mutagens relative to some genetic systems routinely used in screening assays must be determined to accurately estimate the hazard posed to human health from exposure to mutagenic agents. To test the genetic response of an intact higher eukaryotic system in molecular dosimetry studies, we used the yellow-green 2 (yg2) assay that measures forward mutation in maize. The yg2 locus is located on the short arm of chromosome 9 and this mutant phenotype is expressed in the leaves of the sporophyte. A homozygous recessive plant has pale, yellow-green leaves and heterozygous plants have normal, green-colored leaves. In a heterozygous kernel, when the dominant allele is lost within a leaf primordial cell due to a gene mutation, a chromosome break between the centromere and the yg2 locus, or a loss of the chromosome carrying the dominant allele, the recessive allele is expressed and the cell appears deficient in chlorophyll production. When a mutant cell divides, it gives rise to a clone of mutant progeny yielding a yg2 sector in a mature leaf. Such a leaf is a chimera or mosaic, since it is composed of genetically different (mutant and non-mutant) cell lines and may be used as a macroscopic indicator of mutation induction. The advantages inherent in the yg2 assay, as well as the assumptions made, were described previously (Plewa et al., 1984; Schy and Plewa, 1985). The objectives of this study were to analyze the molecular dosimetry of EMS in maize and to compare the results obtained in this intact higher eukaryote with results obtained in molecular dosimetry studies with EMS in other genetic systems, on the basis of the observed mutation kinet-

233 ics and the mutation frequencies induced per unit ethylation. We examine the issue of whether molecular dose can be used as an effective parameter to predict mutation frequencies both within and among genetic systems. Materials and methods

Chemicals EMS was purchased from Eastman Kodak Company, Rochester, NY, and was stored at - 2 0 ° C until ready for use. [1-3H]EMS (3 Ci/mmole) was obtained from Amersham Nederland in 5-mCi aliquots as a neat liquid in borosilicate glass ampoules sealed under vacuum and stored at - 2 0 °C until ready for use. Calf-thymus DNA was obtained from Cooper Biomedical Inc., Malvern, PA. 3,5-Diaminobenzoic acid dihydrochloride (DABA), 99% pure was purchased from Aldrich Chemical Co. Milwaukee, WI. Hexadecyltrimethylammonium bromide (CTAB) was obtained from Sigma Chemical Co., St. Louis, MO. Biofluor scintillation cocktail and a 3H20 standard were purchased from New England Nuclear, Boston, MA. Nuclease P1 from Penicillium citrinum was purchased from Pharmacia Inc., Piscataway, NJ. Alkylated and unalkylated base standards Both alkylated and unalkylated purines were used as UV marker standards to identify chromatographic products following DNA hydrolysis. 7-Ethylguanine (7-EtGua) was synthesized by the method of Jones and Robins (1963), adapted for ethylation rather than methylation and additional 7-EtGua was generously provided by Dr. Joseph Guttenplan of the department of Biochemistry at New York University. Adenine and guanine were purchased from Sigma. Production of maize lines Heterozygous (Yg2 /yg2) kernels of Zea mays were produced from crosses of inbred yg2/yg2 × inbred Early Early Synthetic Yg2/Yg2 made in the Department of Agronomy genetics nursery at the South Farms of the University of Illinois during the summers of 1985 and 1986. The crossing of two inbred lines ensured the resulting F 1 heterozygous kernels were highly isogeneic. These

kernels were used in both the genetic assays and the molecular dosimetry experiments.

Genetic experiments using the yg2 assay Heterozygous (Yg2/yg2) maize kernels were prepared for chemical treatment as previously described (Plewa et al., 1984), with the following modifications. For each treatment group, the number of prepared kernels treated was increased from 12 to 40. Kernels were treated in a 50-ml erlenmeyer flask in 12 ml of potassium phosphate buffered EMS solution (pH 7.2) or buffer alone (control). Molecular dosimetry experiments Kernels were prepared and treated under identical conditions as described for the genetic experiments. However, for the dosimetry treatments, 35 ml of [3H]EMS solution, ranging in specific activity from 90 mCi/mmole for the 1-mM treatment to 12.1 mCi/mmole for the 10-raM treatment, was added to 125-ml erlenmeyer flasks containing 200 prepared kernels. A flask containing 200 prepared kernels in 35 ml of potassium phosphate buffer was included as a control. Sufficient [3H]EMS solution was prepared to assure that 0.1 ml of solution was left over to determine the precise specific activity of each treatment concentration. Following the 8-h treatment, kernels were rinsed repeatedly with cold distilled water and were frozen at - 8 0 °C until dissection. The specific activity of each treatment concentration was determined by preparing replicate dilutions (1 : 1000) for each concentration and adding 20 + 0.2 gl to scintillation vials containing 10 ml of Biofluor scintillation cocktail. Dissection of kernels Embryonic tissue, containing the cone of leaf primordia minus the coleoptile, was dissected from treated kernels. This fraction contained the precise target tissue of the yg2 assay, the leaf 4 and leaf 5 primordia. Immediately upon dissection, this tissue was placed in test tubes on dry ice and subsequently stored at - 8 0 °C until DNA extraction. Extraction of DNA DNA was extracted from the small quantities of tissue obtained using the cetyltrimethylam-

234

monium bromide (CTAB) method of D N A extraction (Murray and Thompson, 1980), adapted to a micro scale. Frozen tissue was lyophilized overnight and ground into a fine powder by adding glass beads and grinding in a sterilized mortar with a pestle. Ground tissue was added to one or more 1.5-ml microfuge tubes containing 0.4 ml CTAB buffer (0.7 M NaC1, 1% CTAB, 50 m M Tris-HC1 p H 8, 10 m M EDTA, and 1% 2mercaptoethanol) to yield approximately a 1 : 3 tissue to buffer ratio. The tubes were incubated at 6 0 ° C for 30 min with occasional agitation. The tissue/buffer suspension was then extracted with 0.4 ml PCA (25 : 24 : 1, phenol : chloroform : isoamyl alcohol), centrifuged for 2 rain and the aqueous (top) layer transferred to a clean microfuge tube. The PCA layer was saved for back-extraction with CTAB buffer (described below). PCA extraction was repeated until the aqueous layer was nearly clear. At this point, 0.08 volumes of 3 M sodium acetate (NaAc) were added to the aqueous fraction and 2 volumes of cold 100% ethanol (EtOH) added. The microfuge tubes were cooled for 10-15 min at - 8 0 ° C and the precipitate was collected by centrifugation in a microcentrifuge for 25 rain. The supernatant was decanted into another microfuge tube and centrifuged for 25 rain. Another D N A - C T A B pellet was obtained and the supernatant discarded. The first PCA fraction was back-extracted with 0.4 ml of CTAB buffer, the aqueous fraction transferred, and another pellet recovered. The CTAB was removed from the D N A - C T A B pellets obtained in all 3 fractions by rinsing the pellets in 0.5 ml of 95% EtOH, centrifuging for 2 min and decanting the supernatant. The pellets were dried for 3 rain under vacuum and were dissolved in 50/~1 of TE (10 mM Tris-HC1, 1 mM EDTA) buffer (pH 8).

Purification of DNA DNA was purified by loading PCA-extracted fractions on a CsC1 density gradient. D N A was added to CsC1 solutions at densities of between 1.48 and 1.56 g/mi. In addition, 340 /~1 of a 10-mg/ml solution of ethidium bromide (EtBr) was added to a final concentration of 300 # g / m l . The DNA-CsC1 solutions were transferred into ll.3-ml polyallomer tubes and centrifuged for 30 h at 53 000 rpm (191000 X g) in a T-1270 (Sorvall)

rotor at 10 ° C. The D N A band was visualized under UV light and was harvested and transferred to 15-ml sterile Corex centrifuge tubes. The EtBr was removed by 3 successive extractions with an equal volume of NaCl-saturated isopropanol. Nensorb 20 columns (Dupont, Wilmington, DE) were used to recover the DNA and remove the CsC1 salt by adding several rinses (total volume, 10 ml) with buffer (100 mM Tris, 10 mM triethylamine, and 1 m M disodium EDTA, pH 7.7). CsCl-free D N A was eluted in 200-#1 fractions from the column with a solution of 1 : 1 methanol:water, and the solvent was subsequently removed in a speed evaporator. DNA pellets in each fraction, if present, were dissolved in 40/~1 of TE buffer (pH 6.5) and fractions from each treatment group were pooled.

Quantitation of DNA DNA was quantified using the DABA-fluorescence assay (Kissane and Robins, 1958; Setaro and Morley, 1976). Fluorescence assays were conducted by adding known volumes of DNA extracts to 13 x 100 mm sterile glass test tubes. Simultaneously, a standard curve was constructed by adding known quantities of calf-thymus DNA, ranging from 200 ng to 4 #g to sterile glass test tubes and placing both the unknown and standard tubes in a forced air-drying oven at 60 ° C overnight. The following day, the tubes were removed and 0.1 ml of freshly prepared 300 m g / m l DABA in distilled water was added to each tube and allowed to incubate for 30 rain at 60 ° C in the dark. Subsequently, 1.9 ml of 0.6 N perchloric acid was added to each tube and the fluorescence of each sample was determined on a Perkin-Elmer Model MPF-44B fluorescence spectrofluorimeter. The excitation and emission wavelengths employed were 420 and 520 nm, respectively, with a 10-nm slit width. A standard curve was constructed for each assay and the quantity of DNA present in unknown fractions was determined from this curve. This value for the quantity of DNA was used in the calculation of the molecular dose for each treatment group.

Determination of molecular dose Prior to scintillation counting, D N A in each fraction was incubated with nuclease P1 at 50 ° C

235

for 8 h by adding 4 units/ml at a DNA concentration of 25 btg/ml. Nuclease P1 acts as both an endonuclease and an exonuclease, cutting DNA non-specifically and yielding 3'-nucleoside monophosphates, thereby preventing the formation of a precipitate upon addition of the DNA to scintillation cocktail. Following digestion with nuclease P1, DNA was added to scintillation vials containing 10 ml of scintillation cocktail and counted on a Packard 2524 scintillation counter. The windows of the analysis and ratio channels spanned the range from 50 to 1000 and 50 to 200, respectively, with a 50% gain set for each channel. Background counts were determined by adding TE buffer to Treatment concentration: 10 mM EMS Amount of D N A isolated: 5.65/~g Specific activity (spec. act.): 12.11 m C i / m m o l e Counting efficiency (C.E.): 40.2% Net clam: 39.2

Molecular dose ( E / N ) = (dpm//,t g) ( F ) a where F =

F=

1.48 × 1 0 - 4

spec. act.

1.48 x 1 0 - 4 12.11 m C i / m m o l e

= 1.22 x 1 0 - 5 net dpm =

net cpm C.E.

39.2 cpm 0.402

= 97.4 dpm 97.4 dpm (1.22 × 10 - 5 ) 5.65 ~g

Molecular dose

= 2.11 ×10 -4 E / N

a F is a conversion factor which is derived from: 1 F= - dpm/mCi

1 nucleotide 1 m C i / m m o l e " mmole " nucleotide/#g

1 1 .6.023×102o. 1 2.22 × 109 " spec. act. 1.83 × 1015 1.48 X 1 0 - 4

spec. act. Fig. 1. Calculation of the molecular dose to the proximate target tissue of the yg2 assay following treatment with 10 mM EMS.

scintillation cocktail in a vial and counting it with the other samples. Counting efficiency was determined at the time of counting using the channels ratio method (Kobayashi and Maudsley, 1974), calibrated by counting acetone-quenched 3H20 standards, prepared as described above for the DNA samples. The calculation of the molecular dose to the DNA of the target tissue of the yg2 assay was carried out as shown in Fig. 1. From the net dpm of the DNA fraction, the amount of DNA present, and the specific activity of the mutagen solution determined for each concentration during treatment, the molecular dose, expressed as ethylations per nucleotide (E/N), was calculated for each treatment group.

Determination of radiolabelling due to incorporation of [3H]EMS degradation products during DNA synthesis In genetic systems that may be undergoing DNA replication during treatment, radiolabelling may, in part, be due to incorporation of degradation products of the [3H]EMS during DNA synthesis (2C labelling). If this effect were significant, an overestimate for the molecular dose would result. Since some percentage of maize kernels undergo DNA synthesis during treatment (Stein and Quastler, 1963; Plewa et al., 1984), it was important to estimate the upper limits of apparent ethylation derived from [3H]2C incorporation during DNA synthesis. This was accomplished by comparing the level of labelling per base of an unalkylated product to that of an alkylated product. The unalkylated base adenine was compared to 7-EtGua, the most prevalent alkylated product. The bases were separated by HPLC using a Whatman (Clifton, N J) 25 × 0.46 cm Partisil-10 strong cation exchange (SCX) column for both the neutral hydrolysate and partially apurinic fractions of DNA according to the method of Beranek et al. (1980). Fractions (1 ml) were collected in scintillation vials following the chromatographic separation. Scintillation cocktail was added to each scintillation vial and the fractions counted under the same conditions described for the dosimetry experiments. From the net cpm obtained for each fraction and by knowing the specific activity of the treat-

236 ment solution, the amount of base hydrolyzed, and the counting efficiency, the amount of ethylations (7-EtGua) or apparent ethylations (adenine) per base was determined for both bases. Comparison of the radioactivity per base between adenine, which should only have had the background level of radioactivity had no incorporation of radioactive breakdown products occurred, and 7-EtGua yielded an estimate of apparent binding that resulted from incorporation during D N A synthesis. Results and discussion

Genetic experiments The kinetics of mutation induction in leaf 4 and 5 primordia were determined for prepared kernels treated with concentrations of EMS ranging from 0 to 10 m M in 8 separate experiments. The yg2 mutation frequency per locus was obtained for both leaf 4 and leaf 5 by dividing the total number of mutant yg2 sectors observed per treatment group by the estimated number of targets (the Yg2 allele). The estimated number of targets is the product of the number of leaves (4 or 5) scored, the estimated number of target cells per leaf primordia, and the number of Yg2 alleles per cell (one for a Yg2/yg2 kernel) (Plewa et al., 1984). The mean induced yg2 mutation frequency per locus was obtained by subtracting the mean mutation frequency per locus of the control from the mean mutation frequency per locus for each treatment group. The data from each experiment were combined and plotted as the mean induced yg2 mutation frequency per locus as a function of EMS concentration for the combined leaf 4 and 5 data. A concentration-dependent increase in the induction of yg2 sectors was observed (slope = 0.93, Fig. 2). In earlier studies with cultured m a m malian cells, both V79 C H O cells and L5178Y mouse lymphoma cells exhibited linear mutation induction kinetics (Aaron et al., 1980). These kinetics differed markedly from the exponential mutation kinetics observed previously in both prokaryotic (E. coli) (Aaron et al., 1980) and eukaryotic (Neurospora and yeast) (van Zeeland et al., 1983) microbial systems. The genetic response observed in maize at the yg2 locus corroborates this distinct difference in the mutation induction kinetics observed between microbial organisms and higher eukaryotes.

o Combined L e o f Ooto

~

~o

× • Moleculoc [3ose o

o

/~/~/.

!

x

co LO CD c2~ t..A [12 U_

Z"

o

T

~

3

o

4

./.=-~

5

6

7

8

g

I0

EHS CONCENTRRTION (mM EHS)

Fig. 2. EMS concentration used as a common parameter to relate the molecular dose (dashed line) and the mean induced mutation frequency per locus (solid line) in the parallel genetic and dosimetry experiments that were conducted. The plot shows the nearly parallel increase in mutation induction and molecular dose obtained in the maize yg2 assay.

Most of the genetic experiments were conducted by treating 40 prepared kernels in 12 ml of treatment solution. However, concurrent studies conducted on the efficiency of D N A extraction from leaf primordial tissue indicated that significantly more than 40 kernels per treatment group were required to extract a sufficient quantity of D N A to determine the molecular dose. Consequently, a genetic experiment was conducted in which 200 prepared kernels were treated in 50 ml of 1 and 3 m M EMS solutions and the results compared to a previous experiment where 40 kernels were treated in 12 ml at the same EMS concentrations. A t-test was conducted and no significant difference was observed in the mutational response between either of the two concentrations under the different treatment conditions. F r o m these experiments, it appeared that the mutational response was not significantly affected by changes in the relative kernel number and solution volume. Thus, the genetic data obtained could be directly compared with a dosimetry protocol that treated 200 kernels in a larger volume of treatment solution.

Molecular dosimetry experiments Molecular dosimetry analyses were conducted in 3 separate experiments where Yg2/yg2 kernels were treated with [3H]EMS concentrations of 0, 1, 3, 5, 7.5 and 10 mM. The data obtained in these experiments are plotted along with the genetic

237 data in Fig. 2. A linear increase in molecular dose was observed as a function of EMS concentration, with a slope of 0.22. In Fig. 2, concentration was used as a common parameter to relate the molecular dose to the mean induced yg2 mutation frequency per locus. From the regression lines illustrated, the value of the dependent variables (the mean induced mutation frequency per locus and molecular dose, respectively) for each concentration examined in the genetic experiments were calculated, and together these values were plotted as the mean induced yg2 mutation frequency per locus as a function of molecular dose (Fig. 3). From this plot, it is apparent that the frequency of forward mutation in both leaves 4 and 5 increased in proportion to the molecular dose to the D N A of the proximate target cells of the y g 2 assay. The induced yg2 mutation frequency per locus per unit ethylation was also calculated for each concentration. This value was obtained by dividing the mean induced yg2 mutation frequency per locus by the molecular dose. A comparison of the induced frequency of mutation per unit ethylation, obtained in this study for the combined leaf 4 and 5 data in maize and for other organisms previously analyzed in molecular dosimetry experiments with EMS, is shown in Table 1. The o m

~r (3

o Combined

~7.5

Leof

% = 4,25×

-

o/

Doto

0,853

o/ Cz

O3 t_.) C3

J

~5.0 c~ 0c2 kL

o/

¥

o/

o.J

o/~9/ o"

E21 Z

o/

o/

,/OOj , 0.0 0.0

i

0.5

1.0

1.5

2.0

2.5

MOLECULRR DOSE (E/N x 104)

Fig. 3. Plot of the mean induced mutation frequency as a function of the molecular dose. This plot was derived from the regression lines in Fig. 2 from the parallel genetic and dosimetry experiments. Using concentration as a commonparameter, the mean induced mutation frequencyper locus and molecular dose were calculated for each concentration assayed in the genetic experiments. Forward mutation induction at the yg2 locus was found to be proportional to the molecular dose.

TABLE 1 COMPARISON OF MUTATION INDUCTION PER UNIT DOSE BETWEEN DIFFERENT ORGANISMS SUBJECTED TO MOLECULAR DOSIMETRY ANALYSIS WITH EMS Organism

Genetic marker

Exposure Ind. Murat. Freq. conc. Unit Ethylation (mM)

E. coli a

gal+

10

0.10

Neurospora b

ad-3a,b

10

0.01

Saccharomycesb ade-l,2

2.5

0.38

V79 CHO a

hprt-

5

2.70

L5178Y a

tk -, hprt -

5

0.46, 0.55

Maize

yg2

1-10 (range)

3.21 ± 0.34 c

a These data were taken from Aaron et al. (1980). b These data were taken from van Zeeland et al. (1983). c The mean value (+ the standard error of the mean) was calculated from data obtained from both leaves 4 and 5.

maize data obtained in this ~tudy established that, in addition to the contrast I n mutation induction kinetics, the absolute frequencies of mutation per unit ethylation, measured at exposures between 1 and 10 m M EMS, differed between the higher eukaryotic systems and the microbial species. Previous work in V79 cells demonstrated a mutation frequency per unit ethylation approximately an order of magnitude greater than that of the mouse lymphoma cell line (Table 1). The maize data indicated that of the two mammalian in vitro cell lines, the measurement of mutation at hprt in the V79 cell line was more representative of the absolute mutation frequencies that might be obtained in an intact higher eukaryote.

Distinguishing between covalent binding and incorporation during DNA synthesis Using H P L C analyses, it was determined that incorporation of labelled 2C degradation products of EMS during D N A synthesis did not account for a significant fraction of the apparent chemical binding of the ethyl adduct to DNA. The radioactivity associated with an unalkylated base, adenine, was compared to the radioactivity associated with an alkylated base, 7-EtGua, in both the neutral

238 TABLE 2 COMPARATIVE RADIOACTIVITY ASSOCIATED WITH THE ALKYLATED BASE (7-EtGua) vs. THE UNALKYLATED BASE (ADENINE) IN THE NEUTRAL HYDROLYSATE AND PARTIALLY APURINIC FRACTIONS OF DNA AND ESTIMATE (UPPER LIMIT) OF PERCENTAGE OF APPARENT BINDING DUE TO INCORPORATION a.b OF BREAKDOWN PRODUCTS OF [ 3H]EMS Fraction

Base

Amount of base (/~g)

dpm per /~g base

Apparent Et/base c

Neutral Adenine Hydrolysate 7-EtGua

0.205 0.055

262 4072

8.73 ×10-5 1.80×10-3

Partially Apurinic

0.595 0.054

47 1660

1.58× 10-5 7.35 ×10-4

Adenine 7-EtGua

a % Incorporation (NH)

8.73 × 10-5

4.8% 1.80×10 -3 1.58×10 -5 b % Incorporation (PA) 2.1% 7.35×10 -4 c Apparent ethylations/base (Et/base) were calculated according to the equation: Apparent Et/base = (dpm/yg base) × ( F ) where F= (1/(dpm/mCi))x(1/specific activity (mCi/ mmole)) x (MW of base (yg/mmole)). hydrolysate and the partially apurinic fractions of D N A . The resulting ratio reflects the percentage of apparent binding due to incorporation of radioactivity. The upper limit of apparent binding due to incorporation during D N A synthesis was f o u n d to be, at most, 4.8 and 2.1% in the neutral hydrolysate and partially apurinic fractions, respectively (Table 2). These values represent the upper limit estimations of 2C incorporation since the adenine fraction was not completely resolved chromatographically from the 7 - E t G u a fraction. Thus, some of the radioactivity collected with the adenine fraction m a y have corresponded to the tail of the 7-EtGua fraction. F r o m these experiments, it was clear that the overwhelming majority of the radioactivity associated with the D N A resulted from covalent binding to the D N A , and incorporation of labelled 2C [3H]EMS degradation products during D N A synthesis, was not significant.

The utility of the molecular dosimetry approach One of the questions that molecular dosimetry studies were designed to answer was whether the

observed difference in mutation induction kinetics between higher eukaryotes and microbial species could be accounted for on the basis of the molecular dose. In this study, maize kernels treated with [3H]EMS solutions demonstrated a linear increase in molecular dose with increased E M S concentration (slope = 0.22). This relationship between the exposure concentration of E M S and the resulting molecular dose was surprisingly similar in all of the genetic systems subjected to molecular dosimetry studies, despite significant differences in metabolism and physiology between the different systems (Fig. 4). In maize, the induction of forward mutation at the yg2 locus was proportional to the molecular dose (Fig. 3). In contrast, the exponential increase in m u t a t i o n induction observed in the microbial systems (Aaron et al., 1980; van Zeeland et al., 1983) cannot be accounted for on the basis of molecular dose (Fig. 4). E. coli cells treated at concentrations of E M S between 1 and 10 m M showed no significant increase in m u t a t i o n induction at two of three loci examined, though the measured dose at these concentrations was not abnormally low (Aaron et al., 1980; Fig. 4). A l t h o u g h the molecular dose showed a direct proportionality with mutation induction

3 x

£ ~z

O.O

oE. coli V7g c e l ] s o + Neurosporo • Moize ~ o Yeost ~6

Z.5

5.0

/ / 7 ~ / .j

jl-Jj

7.5

i j_-~ ..... ,'

10.O

EXPOSURECONCENTRRTION(mM)

Fig. 4. Comparison of the molecular dose measured as a function of treatment concentration between 1 and 10 mM EMS in all of the organisms in which molecular dosimetry experiments have been conducted (with the exception of Drosophila where such data were not available). Despite marked differences in physiology and metabolism among these different systems and a different treatment protocol for maize, the observed molecular dose measured in these organisms over this range of concentrations was quite consistent.

239 in higher eukaryotic systems, the data in the microbial species demonstrated that this relationship is not universal. In fact, one could argue that the apparent linear relationship between molecular dose and mutation induction observed in the higher eukaryotic systems may result from cytotoxicity in the higher forms that masks any nonlinear component of the concentration-response curve that might be present. The maize data obtained in this study established that, in addition to the contrast in mutation induction kinetics, the absolute frequencies of mutation per unit ethylation, measured at exposures between 1 and 10 mM EMS, differed between the higher eukaryotic systems and the microbial species. Thus, the molecular dosimetry approach in comparative mutagenesis studies has clearly demonstrated that differences in the genetic response, both in terms of mutation kinetics and mutation frequencies, result from factors other than the level of chemical-DNA binding per se. Evidence for the importance of some of these other factors is briefly reviewed below. Correlation between the level of 06-EtGua and mutation-induction kinetics Molecular dose may be further resolved by quantifying the level of specific chemical-DNA adducts, and analysis at this level has demonstrated strong correlations with mutation induction that extend across a wide variety of genetic systems. Mohn and van Zeeland (1985) compared the induction of mutation by a number of different ethylating agents, including EMS, in E. coli and V79 cells, and, in parallel studies, measured the molecular dose and the level of two different DNA adducts, 7-EtGua and Ot-EtGua. Like Aaron et al. (1980), they found a nearly proportional increase in molecular dose with increasing EMS concentrations in both E. coli and V79 cells. However, the exponential increase in mutation induction that was observed in E. coli correlated well with an exponential increase in O6-EtGua formation. The formation of 7-EtGua, like the molecular dose, increased linearly with exposure concentration and was not correlated with mutation induction in E. coll. Likewise, the linear increase in mutation induction in V79 cells observed previously (Aaron et al., 1980) was con-

firmed in this study and was correlated with a linear increase in O6-EtGua. The significance of these results was greatly strengthened when mutation induction by EMS, ENU and diethyl sulfate (DES) in both E. coli and V79 cells was found to be highly correlated with the level of O6-EtGua, irrespective of the molecular dose or the level of 7-EtGua measured (Mohn and van Zeeland, 1985). These data indicated that the level of Ot-EtGua may be the factor that best accounts for the difference in the kinetics of mutation induction between the higher eukaryotic and microbial systems. They also observed, as in earlier studies based on mutation frequency per unit ethylation, that the absolute mutation frequencies per 0 6EtGua adduct were higher in the V79 cells than in E. coli. Factors that may account for differences in mutation frequencies among organisms Several factors have been suggested to explain the wide variation in mutation frequencies observed between different organisms. Three of the most plausible explanations include, (1) differences in repair of alkylation-induced damage between different organisms, (2) the location of the genes used as genetic markers and the genetic endpoint measured, and (3) differences in the target size of genes or genomes. Effect of DNA repair of alkylation-induced damage The study by Mohn and van Zeeland (1985) supported the importance of DNA repair in influencing both the kinetics of mutation induction and the absolute mutation frequency. They monitored the ratio of O6-EtGua to 7-EtGua in both V79 cells and E. coli at different levels of EMS exposure. A constant ratio of O6:N-7 adducts was observed over the range of concentrations examined in the mammalian cells, while the ratio was lower at low concentrations of EMS in the bacteria than at high concentrations. Such a change in the ratio of these adducts is indicative of either the action of some constitutive component of the adaptive response to alkylation or the uvr + excision-repair pathway. Another possible explanation for the elevated mutation frequency in higher eukaryotic cells is the absence or a decreased efficiency of some of

240 the repair enzymes. For example, Warren et al. (1979) monitored the removal of different D N A adducts using HPLC analysis following treatment of V79 cells with MNU. The cells did not appear to remove O6-MeGua adducts or phosphotriesters, perhaps reflecting the absence of a functional alkyltransferase enzyme. In other mammalian cells that can remove these adducts, the rate of removal appears to be 100-fold slower than that observed in bacteria (Pegg et al., 1984). In addition, the mammalian alkyltransferase enzyme cannot remove the potentially mutagenic adduct O4-Alk Thy, a substrate that is readily removed by the bacterial alkyltransferase (Pegg et al., 1985). Thus, variations in repair rates, capacities, and substrate specificities in different genetic systems all may contribute to different mutation kinetics and frequencies.

Influence of gene location and the number of genetic targets on the detection of gene mutation The location and number of genetic targets, or its 'genetic context', can impact the observed kinetics of mutation induction. Molecular dosimetry studies at the homologous genes ad-3 in Neurospora and adel or ade2 in yeast displayed exponential mutation kinetics of 1.5 and 2.6 for yeast and Neurospora, respectively (van Zeeland et al., 1983). The difference in slopes was ascribed to the fact that, in contrast to the haploid yeast, the Neurospora strain employed is a heterokaryon, a functional diploid, and thus can detect any multi-locus deletions induced by EMS, as well as point mutations. Like the heterokaryon strain of N. crassa, maize kernels heterozygous at the yg2 locus can detect both point mutation and chromosome aberration. EMS is thought to primarily induce point mutations, the vast majority consisting of G : C ~ A : T transitions (Burns et al., 1986). In kernels treated under conditions nearly identical to those employed in this study, at concentrations between 1 and 10 mM EMS, no significant increase in the induction of micronuclei were observed in meristematic root tip cells (Wagner and Plewa, 1985). Thus, it is likely that the vast majority of EMS-induced mutation measured in this maize molecular dosimetry study resulted from point mutational events.

Effect of gene size Although the size of the structural genes for proteins should be the same in organisms from E. coli to humans, the presence of intervening sequences (introns) in eukaryotes provides a basis for increasing the effective target size in higher organisms. A mutation affecting a site essential for proper splicing of the transcript or one which activates a cryptic splice site may provide additional sites for mutation not present in prokaryotes. In addition, certain genetic loci, such as that coding for the Na + / K + ATPase, are referred to as small marker genetic loci. At this locus, only mutations affecting a select number of bases that disrupt the site to which the toxin ouabain binds are detectable by conferring ouabain resistance (Liber et al., 1986). The frequency of mutation at such small marker loci is typically much lower than at other genes in the same organism and the genetic endpoint of such loci is limited to point mutation. Such differences in the genetic endpoints detected at different loci, whether due to their size or context, are responsible for some of the variation in mutation frequency observed between organisms as well as within a single organism. Molecular dosimetry studies indicate that the level of chemical binding to D N A alone is not sufficient to explain differences observed in mutation kinetics and frequencies between different organisms. Instead, the level of specific DNA adducts and cellular factors characteristic of each genetic system appear to influence the observed kinetics and frequency of mutation induction, respectively. Additional study of these factors will be required to reliably extrapolate results between organisms and ultimately lead to the capability of assessing risk posed by mutagens to human health.

Acknowledgements This research was conducted in partial fulfillment of the Ph.D. degree by W.E.S. We wish to thank Drs. Frits Sobels and Albert van Zeeland for their assistance in obtaining the [3H]EMS required in this study in conjunction with a consortium of the European Economic Community. We appreciated the helpful suggestions included by Dr. van Zeeland in his correspondence. The

241 I n s t i t u t e for E n v i r o n m e n t a l S t u d i e s p r o v i d e d t h e f u n d s u s e d to p u r c h a s e t h e [ 3 H ] E M S in its a w a r d o f a n E n v i r o n m e n t a l T o x i c o l o g y S c h o l a r s h i p to W . E . S . I n a d d i t i o n , this r e s e a r c h was f u n d e d b y NIEHS grant No. RO1 ESO1895 GEN and by a G r a n t - i n - A i d o f R e s e a r c h f r o m S i g m a Xi, as w e l l as a G r a n t for T h e s i s / P r o j e c t s u p p o r t f r o m t h e G r a d u a t e C o l l e g e of t h e U n i v e r s i t y o f Illinois, Urbana-Champaign.

References Aaron, C.S. (1976) Molecular dosimetry of chemical mutagens: Selection of appropriate target molecules for determining molecular dose to the germ lines, Mutation Res., 38, 303-310. Aaron, C.S., and W.R. Lee (1978) Molecular dosimetry of the mutagen ethyl methanesulfonate in Drosophila melanogaster spermatozoa. Linear relation of DNA alkylation per sperm cell (dose) to sex-linked recessive lethals, Mutation Res., 49, 27-44. Aaron, C.S., A.A. van Zeeland, G.R. Mohn, A.T. Natarajan, Ada G.A.C. Knapp, A.D. Tates and B.W. Glickman (1980) Molecular dosimetry of the chemical mutagen ethyl methanesulfonate: Quantitative comparison of mutation induction in Escherichia coli, V79 Chinese hamster cells and L5178Y mouse lymphoma cells, and some cytological results in vitro and in vivo, Mutation Res., 69, 201-216. Beranek, D.T., C.C. Weis and D.H. Swenson (1980) A comprehensive quantitative analysis of methylated and ethylated DNA using high pressure liquid chromatography, Carcinogenesis, 1, 595-606. Burns, P.A., F.L. Allen and B.W. Glickman (1986) DNA sequence analysis of mutagenicity and site specificity of ethyl methanesulfonate in uvr + and uvrB- strains of Escherichia coli, Genetics, 113, 811-819. Jones, J.W., and R.K. Robins (1963) Purine nucleosides, III. Methylation studies of certain naturally occurring purine nucleosides, J. Am. Chem. Soc., 85, 193-201. Kissane, J.M., and E. Robins (1958) The fluorometric measurement of deoxyribonucleic acid in animal tissues with special reference to the central nervous system, J. Biol. Chem., 233, 184-188. Kobayashi, Y., and D.V. Maudsley (1974) Biological Applications of Liquid Scintillation Counting, Academic Press, New York. Lee, W.R. (1976) Molecular dosimetry of chemical mutagens: Determination of molecular dose to the germ line, Mutation Res., 38, 311-316.

Liber, H.L., P.M. Leong, V.H. Terry and J.B. Little (1986) X-Rays mutate human lymphoblast cells at genetic loci that should respond only to point mutation, Mutation Res., 163, 91-97. Mohn, G.R., and A.A. van Zeeland (1985) Quantitative comparative mutagenesis in bacteria, mammalian cells, and animal-mediated assays, A convenient way of estimating genotoxic activity in vivo?, Mutation Res., 150, 159-175. Murray, M.G., and W.F. Thompson (1980) Rapid isolation of high molecular weight plant DNA, Nucleic Acids Res., 8, 4321-4324. Pegg, A.E., M.E. Dolan, K. Marimoto and D. Scicchitano (1984) Repair of alkylated DNA by alkyltransferases from mammalian cells and from bacteria, abstract from the Symposium on DNA adducts: Dosimeters to monitor human exposure to environmental mutagens and carcinogens, 24-26 Sept. 1984, Research Triangle Park, NC. Pegg, A.E., M.E. Dolan, D. Scicchitano and K. Morimoto (1985) Studies of the repair of O6-alkylgnanine and O4-al kylthymine in DNA by alkyltransferases from mammalian cells and bacteria, Environ. Health Perspect., 62, 109-114. Plewa, M.J., P.A. Dowd and E.D. Wagner (1984) Calibration of the maize yg2 assay using gamma radiation and ethylmethanesulfonate, Environ. Mutagen., 6, 781-795. Schy, W.E., and M.J. Plewa (1985) Induction of forward mutation at the yg2 locus in maize by ethylnitrosourea, Environ. Mutagen., 7, 155-162. Sega, G.A. (1984) A review of the genetic effects of ethyl methanesulfonate, Mutation Res., 134, 113-142. Setaro, F., and C.G.D. Morley (1976) A modified fluorometric method for the determination of microgram quantities of DNA from cell or tissue cultures, Anal. Biochem., 71, 313-317. Stein, O.L. and H. Quastler (1963) The use of tritiated thymidine in the study of tissue activation during germination in Zea mays, Am. J. Bot., 50, 1006-1011. van Zeeland, A.A., G.R. Mohn, C.S. Aaron, B.W. Glickman, M. Brendel, F.J. de Serres, C.Y. Hung and H.E. Brockman (1983) Molecular dosimetry of the chemical mutagen ethyl methanesulfonate, Quantitative comparison of the mutagenic potency in Neurospora crassa and Saccharomyces cerevisiae, Mutation Res., 119, 45-54. Wagner, E.D., and M.J. Plewa (1985) Induction of micronuclei in maize root-tip cells and a correlation with forward mutation at the yg2 locus, Environ. Mutagen., 7, 821-832. Warren, W., A.R. Crathorn and K.V. Shooter (1979) The stability of methylated purines and of methylphosphotriesters in the DNA of V79 cells after treatment with Nmethyl-N-nitrosourea, Biochim. Biophys. Acta, 563, 82-88.