Two Simple Methods for the Detection of Mutagens' · Two Simple Methods for the Detection of Chemical Mutagens' V. N. IYER2 AND W. SZYBALSKI Institute ofMicrobiology, Rutgers, The

METHODS FOR DETECTION OF CHEMICAL MUTAGENS

O'NEIL, J. 1956 Master's Thesis, Department of FoodTechnology, Massachusetts Institute of Technology,Cambridge, Mass.

PALADINO, S., UGOLINI, F., AND CHAIN, E. B. 1954 Fer-mentors of 90 and 300 L capacity for vortex and spargeraeration. Rend. Ist. Super. Sanith. 17, 87-120. (Eng. Ed.).

PFEIFER, V. F., VOJNOVICH, C., AND HEGER, E. N. 1952Itaconic acid by fermentation with Aspergillus terreus.Ind. Eng. Chem., 44, 2975-2980.

RAKE, G. AND DONOVICK, R. 1946 Studies on nutritionalrequirements of Streptomyces griseus for formation ofstreptomycin. J. Bacteriol., 52, 223-226.

RIVETT, R. W., JOHNSON, M. J., AND PETERSON, W. H. 1950Laboratory fermentor for aerobic fermentations. Ind.Eng. Chem., 42, 188-190.

STEFANIAK, J. J., GAILEY, F. B., BROWN, C. S., AND JOHNSON,M. J. 1946 Pilot plant equipment for submerged pro-duction of penicillin. Ind. Eng. Chem., 38, 666-671.

Two Simple Methods for the Detection of Chemical Mutagens'V. N. IYER2 AND W. SZYBALSKI

Institute of Microbiology, Rutgers, The State University, New Brunswick, New Jersey

Received for publication Jupe 20, 1957

The demonstration by Muller (1927) and Stadler(1928) that mutations could be experimentally pro-duced by physical agents (X-rays) was one of the mostfundamental biological discoveries. Definite proof of themutagenic action evoked by chemical agents awaitedthe 1941 studies of Auerbach, published belatedly dueto circumstances associated with World War II (Auer-bach, 1951).Although Drosophila was the organism of choice in

the original investigations, microorganisms were soonrecognized as convenient tools for the study of chemicalmutagenicity (Horowitz et al., 1946; Tatum, 1946;Witkin, 1947; Bryson, 1948). Development of themutational system employing the streptomycin-depend-ence locus by Bertani (1951) gave further impetus tosystematic surveys of large numbers of chemicals formutagenic activity on bacteria (Demerec et al., 1951;Hemmerly and Demerec, 1955). The use of bacteria inmutational studies on carcinogens was reported inseveral papers (cf Scherr et al., 1954). These studiesdemonstrated a close correlation between the carcino-genic effect in mammals and the mutagenic effect onbacteria, stimulating a wide interest in this field. Otherfruits of this approach comprised the discovery ofsimple chemical substances capable of producing muta-tion within a wide range of concentrations in theabsence of significant lethal effects (Demerec andHanson, 1951), and the demonstration of a rathersignificant correlation between the antineoplastic ac-tivity observed in animal tumor experiments andmutational effects as assayed with the bacterial system(Gellhorn and Hirschberg, 1955). The subject is com-

1 Partially supported by a grant (#3240) from the UnitedStates Public Health Service.

2 Postdoctoral Fellow (United States Government, Ful-bright and Smith-Mundt Act). Present address: S. B. GardaCollege, Navsari, Surat Dt., Bombay, India.

prehensively reviewed by Boyland (1954) and Burdette(1955).The present paper describes further simplification of

the methods used for detection of mutagenic substancesamong pure chemicals and in complex natural sub-strates. The modifications introduced should pave theway for efficient large-scale surveys for these remark-able agents, which show pronounced direct or indirecteffects on the hereditary constitution of the cell nucleusat relatively nontoxic concentrations.

MATERIALS AND METHODS

Media. Difco nutrient broth (8 g per L) was em-ployed in most of these studies. Solid media contained2 per cent agar or 0.7 per cent agar (soft agar).Escherichia coli strain Sd4-73 was grown in nutrientbroth containing 20 ,ug per ml of streptomycin, whileits total count was scored on nutrient agar supple-mented with 100 ,ug per ml of streptomycin. The mini-mal medium (7 g K2HPO4; 2 g KH2PO4; 0.5 g Na3-citrate.5H20; 0.1 g MgSO4*7H20; 2.5 g glucose; 20 gagar; 1 L distilled water) for scoring of reversions in thecystine locus was supplemented with 100 ,ug per ml ofstreptomycin.

Strain. The streptomycin-dependent, cystine-re-quiring strain Sd4-73 of Escherichia coli was obtainedfrom Dr. M. Demerec of the Cold Spring HarborLaboratories. It was reisolated and a subculture selectedon the basis of a low background of spontaneous rever-sion in the streptomycin-dependence locus. This stepwas resorted to occasionally when the stock culturebegan to yield an increasing number of spontaneousrevertants.

Principle of the methods. The methods employ themutational system described by Bertani (1951) andsubsequently used by Demerec et al. (1951) for detection

19581 23

on April 4, 2020 by guest

http://aem.asm

.org/D

ownloaded from

V. N. IYER AND W. SZYBALSKI

of bacterial mutagens. Their procedure consisted oftreatment of a washed bacterial suspension for adefinite period by the mutagen dissolved in distilledwater, and plating the treated bacteria and also anuntreated control suspension, at appropriate dilutions,on selective and nonselective media. If the streptomy-cin-dependent mutant of E. coli was employed, it wasplated on nutrient agar containing streptomycin toascertain the total count of bacteria, before and aftertreatment, and on streptomycin-free nutrient agar toassay the number of streptomycin nondependent mu-tants. This method yielded a quantitative measure ofmutagenicity expressed as an increase in the propor-tion of streptomycin nondependent mutants in thepopulation surviving the treatment, when comparedwith the spontaneous mutation frequency. In severalsubsequent studies, Demerec and others (Demerec etal., 1951; Hemmerly and Demerec, 1955; Glover, 1956)have consistently used the streptomycin-dependencelocus and, upon comparison with the mutations inother loci, have found it to be a very reliable, sensitive,and convenient indicator of mutagenicity. The presentstudies with E. coli strain Sd4-73 confirm this conclu-sion. Furthermore, as will be described, the use of thislocus permits study of the mutagenic properties of agreat variety of products which contain nutrientscapable of interfering with the mutagenicity assayinvolving many other loci. Bertani's (1951) studiesindicated that spontaneous, nondependent mutantsare mainly streptomycin sensitive, and are thus at aselective disadvantage in streptomycin-containingmedia.The method used by Demerec et al. (1951) gives very

reliable quantitative results, but is rather cumbersomewhen used as a preliminary screening procedure for agreat variety of substances. The chief disadvantages inthis respect are: (1) the large amount of individualhandling, (2) the separate assay for each concentrationof the mutagen, and (3) the possible clumping of thecells by the chemical, with subsequent distortion of thequantitative results.

In the present studies, the modification of the stand-ard methods proceeded independently along two lines.The membrane method impinges the treated cells onCellophane or Millipore3 membranes, thus avoiding theclumping error and still allowing a quantitative assay,with the mutagen concentration and the length ofexposure as independent and controlled variables. Thepaper-disc method simplifies the assay still further andpermits measuring on one plate the effect of a widerange of concentrations and exposure times, sacrificing,however, the independent control of variables and thequantitative aspects of the results.The paper-disc method appears to be well-suited to

the preliminary survey of a large number of mutagenic3Millipore Filter Corporation, Watertown, Massachulsetts.

substances and materials suspected of containing muta-gens. The following sections will describe the details ofboth methods and compare the results with a widearray of chemicals and antibiotics.

The Paper-Disc MethodA culture of E. coli strain Sd4-73 was prepared in

nutrient broth containing 20,g per ml streptomycinby incubation overnight at 36 C with aeration. Aerationby shaking was found to be as suitable as forced aera-tion, contrary to the conclusions of Demerec (1954).The culture was centrifuged, washed twice in distilledwater to remove streptomycin (a compulsory step),and resuspended in saline to a concentration of approx-imately 109 cells per ml. The number of cells per platewas found to be critical, the yield of mutant coloniesbeing reduced either by crowding or by insufficientpopulation size. A 0.1 ml amount of this suspension wasinoculated into 2.5 ml of soft nutrient agar (0.7 per centagar) on a base of 20 ml of 2 per cent nutrient agar.The use of an additional plate seeded with a tenfoldlower dilution is recommended. Seeding by layeringwith soft agar was found to give more uniform resultsthan spreading the suspension over the surface of themedium. After the agar had set, but without muchfurther delay, a sterile filter paper disc4 of a standarddiameter of 12.7 mm was placed on the surface of themedium and moistened with 0.1 ml of a solution of thetest chemical. Plates of nutrient agar containing 100,ug per ml of streptomycin could be seeded in parallelto give an indication of the size of the inhibition zone,if any, with the amount of the test chemical employed.Maximum expression of mutagenicity on streptomycin-free test plates was usually obtained when testingchemicals at concentration levels where a narrow zoneof inhibition was demarcated around the disc on strepto-mycin-containing control plates. Aqueous solutions oftest chemicals were used wherever possible. Acetonewas the solvent for water-insoluble compounds, afterit was determined that this solvent neither inhibitedgrowth nor exhibited detectable mutagenesis. Solutionsof unstable compounds were prepared just beforetesting. The large majority of compounds reported inthis paper could also be assayed by placing a few crys-tals on the disc moistened with the solvent. In a largenumber of experiments using minimal agar, the cystinelocus was found to be not sensitive enough to allow itsapplication in the test. The use of minimal agar plus10 ,ug per ml of cystine instead of nutrient agar gaveequally good results, and occasionally even moreclearly defined mutant colonies. However, since noremarkable advantages were gained with this medium,there appeared to be no reason for substituting it for aconvenient routine medium such as nutrient agar. Inany series of tests, it is an essential and reassuring meas-

Schleicher & Schuell Co., Keen, New Hampshire.

[VOL. 624

on April 4, 2020 by guest

http://aem.asm

.org/D

ownloaded from

18ETHODS FOR DETECTION OF CHEMIICAL MUTAGENS

ure to run controls for spontaneous mutation with discsmoistened in sterile water or other solvent employedin the test. Petri dishes were incubated at 36 C for atleast 5 days, at which time they were examined formutant colonies. Positive results were characterized bythe appearance of several colonies around the mutagen-saturated discs on nutrient agar, while parallelmutagen-free plates exhibited no growth or significantlyfewer colonies. The mutant colonies were arrangedaround the disc in a ring, the inner and outer radii ofwhich were determined by the amount of mutagen,toxicity, rate of diffusion, and the range of effectivemutagenic concentration. The above procedure and thedescribed results can be visualized in figure 1.

The MIembrane MIethod

In this method, a solution of the test chemical wasfirst spread over a well-dried surface of plain 2 per centagar. The plates were allowed to stand for about 1 hrto permit diffusion of the chemical into the agar. Asterilized sheet of cellophane membrane, cut out fromdialysis tubing, was next placed on the surface of the

agar and inoculated with approximately 101 washedcells of E. coli Sd4-73 in 0.05 ml of distilled water. Thesuspension was spread evenly over the membrane, andwith a sufficiently dry plate, soaked rapidly inlto themedium. The plates were incubated at 36 C for 2 hrand the membranes were then removed carefully fromthe plain agar surface and transferred to the surface ofa nutrient agar plate, which was incubated for at least5 days to permit the development of mutant colonies. Asecond membrane treated in the same manner wasplaced on the surface of streptomycin agar, to detectthe inhibition, if any, by the test chemical. Likewise, athird similarly treated membrane placed on the surfaceof streptomycin-containing minimal agar served as atest for mutation at the cystine locus. Mutagenicity forthe streptomycin-dependence locus was indicated bythe development of a large number of streptomycin-independent colonies on the first membrane as com-pared with the control (figure 2). Similarly, muta-genicity for the cystine locus was manifested by anincreased colony count on the third membrane, al-though this locus was found to be less sensitive to the

,,........... l.

Figure 1. Paper disc test for mutagenicity at the streptomycin-dependence locus (Escherichia coli strain Sd4-73). A, B, and C:Mutagenic effect of beta-propiolactone (10 mg per disc). A: Control plate showing inhibition zone (Inoculum, 109 cells; nutrientagar + 100 ,ug per ml streptomycin). B: Appearance of mutant colonies with high inoculum (109 cells; nutrient agar). C: Appearanceof mutant colonies with low inoculum (108 cells; nutrient agar).D and E: Mutagenic effect of azaserine (100 ,ug per disc, nutrient agar). D: Formation of mutant colonies in proximity of disc

with an inoculum of 109 cells per plate. E: Formation of mutant colonies in proximity of disc with an inoculum of 108 cells per plate.F: Control plate (nutrient agar) for spontaneous formation of streptomycin-independent colonies. Inoculum, 109 cells per plate;

0.1 ml distilled water added to disc.

1958] 25

on April 4, 2020 by guest

http://aem.asm

.org/D

ownloaded from

V. N. IYER AND W. SZYBALSKI

mutagens tested. The potency of the mutagen at agiven concentration can be expressed as the ratio ofthe number of colonies on the mutagen-treated mem-brane to the number on the untreated control.

In its present version, the membrane method permitstesting only one concentration of the mutagen at onetime. A tenfold range of concentration can be tested onone plate by applying the gradient-plate principle(Szybalski, 1952).

RESULTSEvaluation of the two methods described was made

on the basis of comparison of the results with thoseobtained by an accepted method, in this case thestandard method used extensively by Demerec andco-workers and described by Demerec and Hanson(1951). A series of 36 substances, including knownmutagens and other substances of suspected muta-genicity, were tested. Table 1 summarizes the results ofa comparative survey of these 36 chemicals for muta-genicity by the three assay methods. Results obtainedwith manganous chloride (shown to be mutagenic byDemerec and Hanson, 1951) were not included in thepresent survey because, under the conditions of theexperiments, it often became oxidized and precipitatedin the form of a brown deposit. Some of the tests, how-ever, where oxidation was not pronounced, gave anindication of a positive mutagenic response to manga-nous chloride.Of the chemicals tested, seven proved to be decidedly

mutagenic for E. coli by both the paper disc andmembrane methods. Significantly enough, these sevencompounds were the only ones found to be stronglymutagenic when tested by the method of Demerec et al.(1951). The range of sensitivity of the two methods forthese seven mutagens was determined (table 2). Bothmethods were found to be sensitive to low concentra-

tions of the mutagens and the range of sensitivity waswide. Expression of the mutagenicity of the seven com-pounds was not affected by the presence of complexorganic nutrients (Difco Penassay broth).

DIscuSSIONThe two methods described have their main advan-

tages in ease and simplicity. Neither of the methods intheir present versions are fully quantitative but, withthe use of proper controls, they do permit a semiquanti-tative analysis of results. Furthermore, the membranemethod can be developed into a fully quantitativeassay. Matney (1955) has used Millipore membranesfor the quantitative determination of mutation ratesto streptomycin resistance. It is more important thatwithin a wide range of concentrations they have provedto be dependable methods for the detection of muta-gens. The close correlation obtained in comparativestudies with the method of Demerec et al. (1951) showsthat they are as reliable as the more cumbersomestandard method for screening of potential mutagenicsubstances.Though both methods described yield reliable results,

one may be preferred over the other under certainconditions. The paper-disc method is to be recom-mended when only small quantities of the testcompound are available, and when the substance beingtested is not rapidly bactericidal at the test level andis readily diffusible in agar. For most purposes, this isthe method of choice. It has certain features to recom-mend it apart from simplicity. It permits the exposureof the cells on a single Petri dish to a wide concentra-tion range of the mutagen, within which the establish-ment of mutant colonies takes place. At each point inthe vicinity of the disc, the concentration of themutagen is a function of time, increasing at first as themutagen diffuses through the agar, and subsequently

Figure 2. Membrane method for assay of mutagenicity at the streptomycin-dependence locus (Escherichia coli strain Sd4-73, in-oculum 108 cells per membrane). A and B: Cells pretreated for 2 hr with azaserine (100l,g per 20 ml of plain agar). A: Appearanceof streptomycin-independent mutant colonies on membrane transferred to nutrient agar. B: Confluent growth on membrane trans-ferred to streptomycin-containing nutrient agar, indicating the absence of excessive bactericidal effects for azaserine at concentra-tion and time employed. C: Control plate (nutrient agar) with membrane-supported cells, not exposed to azaserine.

26 [VOL. 6

on April 4, 2020 by guest

http://aem.asm

.org/D

ownloaded from

1958] METHODS FOR DETECTION OF CHEMICAL MUTAGENS 27

decreasing due to further dilution of the limited supply concentrations both inhibitory and mutagenic for aof mutagen furnished by the disc. The time-concentra- limited time only, permitting subsequent expressiontion curve becomes flatter with the maximum shifted of mutation as colonial growth. Inasmuch as the bac-to the right for points farther and farther removed from terial cells are fixed in the agar before the addition ofthe disc. Thus, points may exist where the drug is at the mutagen, clumping of the cells by the chemical

TABLE 1Evaluation of the paper disc and membrane methods-a comparison with the method of Demerec et al.

Mutagenic effect of 36 compounds tested on the streptomycin-dependence locus of Escherichia coli, Sd4-73

Paper Disc Method Membrane Method Method of Demerec et al. (1951)

Amount of Amount of Amount of Number ofCompound compound Relative compound Relative compound in inducedin jAg per disc mutageni- in2g per mutageni- Ag per ml of Per cent mutantsper 20 ml cityt 20 ml city: treated survival pe 0

agar* ~~~agar * bacterial pervi10ragar' ~~~~~~~~~~~suspensionsuivrActinomycin .................................... 100 1 100 1 10 3.2 0Aminopterin .................................... 1,000 1 1,000 1 500 74 0Amethopterin ................................... 1,000 2 1,000 1 100 29 0Ammonium hydroxide....... .................... 1,400 2 1,400 5 100 38 0Alazopeptin ..................................... 1,000 1 1,000 1 500 76 08-Azaguanine ......... 1,000 1 1,000 1 50 0.5 0Azaserine ....................................... 100 100 100 100 5 35 1.5 X 1036-Azathymine ................................... 1,000 1 1,000 1 500 58 06-Azauracil ..................................... 1,000 1 1,000 1 500 82 0Benzimidazole................................... 1,000 2 1,000 2 500 15 0Boric acid ...................................... 1,000 1 1,000 1 1,000 64 0Caffeine....................................... 500 1 500 1 500 88 0Carzinophilin ................................... 500 2 500 3 5 23 4.3 X 10'

units units units6-Chloropurine .................................. 1,000 1 1,000 1 10 60 1.8 X 101Cupric sulfate ................................... 1,000 1 1,000 1 100 0.5 02,4-Diamino-5(p-dichlorophenyl)6-ethylpyrimi-

dine.......................................... 1,000 1 1,000 1 50 12 02,4-Diamino-5(3'4'-dichlorophenyl) -6-ethylpyri-midine....................................... 1,000 2 1,000 1 50 3.4 0

2,6-Diaminopurine........... ................... 1,000 1 1,000 1 100 63 06-Diazo-5-oxo-L-norleucine ....... ................ 1,000 1 1,000 1 1,000 9402,5-bis(ethyleneimino)benzoquinone ............. 50 8 1,000 20 5 38 1.0 X 102Formaldehyde ................................... 500 2 500 2 1,000 92 8.5 X 101D-Glucosamine .................................. 1,000 1 5,000 1 1,000 95 0Hydrogen peroxide........... ................... 3,000 2 3,000 1 3,000 79 2.3 X 101Kinetin....................................... 50 1 100 1 50 172 06-Mercaptopurine . ............................... 1,000 1 1,000 1 500 80 0N-methyl-bisG(6-chloroethyl)amine hydrochloride. 100 30 500 60 50 9 3.9 X 103N-methyl-bis(3-chloroethyl)amine N-oxidehydrochloride ................................. 100 12 500 10 100 35 4.0 X 101

Myleran ........................................ 1,000 1 1,000 1 100 53 0Nucleocidin..................................... 100 1 100 1 100 68 0Proflavine (sulfate) .............................. 500 2 500 5 100 20 0beta-Propiolactone.............. ................. 10,000 100 10,000 58 100 32 3.8 X 102Puromycin...................................... 100 1 500 1 100 5.6 0Sulfocidin ....................................... 50 5 10 5 5 8.3 3.0 X 102Triethylenemelamine ............................ 500 26 1,000 20 5 68 9.7 X 102

Trypaflavine .................................... 500 500 100 58Urethane...........1............................ ],000 1 1,000 1 1,000 95 0

* The amount of substance used was adjusted to give a narrow zone of inhibition (Disc Method) or confluent growth (Mem-

brane Method) on streptomycin-containing control plates. Where no inhibition was observed, the highest amount tested is reported.t An approximate ratio of the number of colonies on the plate containing the mutagen-saturated paper disc to the number

of colonies on the control plate. When induced mutant colonies form a definite ring around the paper disc, the figure for rela-tive mutagenicity is multiplied by the ratio of the total area of the plate to the area of the growth of induced mutant colonies.A relative mutagenicity of 1 indicates a lack of significant mutagenic activity.

t An approximate ratio of the number of colonies on the Cellophane membrane previously exposed to the putative mutagento the number of colonies on the nontreated control Cellophane strip. Both values are assumed to be equal or greater than one.A relative mutagenicity of 1 indicates a lack of significant mutagenic activity.

-- - -

on April 4, 2020 by guest

http://aem.asm

.org/D

ownloaded from

V. N. IYER AND W. SZYBALSKI

TABLE 2Mutagenic response to a wide concentration range of 7 mnutagens

as revealed by the disc and membrane methods

Amount of Relative Mutagenicity*Mutagen Mutagen in

pLg per 20 Paper disc Membraneml Agar method method

Azaserine

2, 5-Bis(ethyleneimino)benzoquinone

N-Methyl-bis (,B chloroethyl)amine hydrochloride

N-Methyl-bis (fi chloroethyl)amine-N-oxide hydrochloride

beta-Propiolactone

Sulfocidin

Triethylenemelamine

151050100500

1,000

15

1050100500

1,000

15

1050100500

1,000

15

1050100500

1,000

110

1001,000

10,000100,000

15

1050

10050

1,000

151050100500

1,000

26

16601008012

566891316

381818303030

2661212108

12

40>100>100

0

1335321

1146

202227

11060

>100>100

10025

118

2060

>100>100

cannot influence the assay. The membrane method,on the other hand, is designed to withdraw the cellsfrom further contact with the mutagen after a definiteperiod of time, thus enabling one to control and limitthe possible toxic action of the chemical. This method,however, permits cells to be exposed to only a singleconcentration of mutagen per plate and generallyrequires a larger amount of the test chemical.Both methods remain flexible and capable of a range

of modifications depending on the exigencies of thesituation. The paper-disc method, for instance, mayconceivably be adapted for the study of synergismamong mutagens, or possibly even antimutagenicactivity. The sensitivity of the methods does not appearto be affected by the presence of complex constituentsof nutrient media. Though this may not be true in allcases, the methods open the way to the examinationof a variety of substances ranging from culture filtratesto chemicals of varying degrees of purity. It is signifi-cant that two of the seven compounds reported to bepotent mutagens in this paper (2,5-bis(ethyleneimino)-benzoquinone and sulfocidin) were first found to bemutagenic by the paper-disc method.

ACKNOWLEDGMENTSWe are gratefully indebted to Merck & Co., Inc. for

the sample of actinomycin D; to Lederle LaboratoriesDiv. for aminopterin, amethopterin, alazopeptin,nucleocidin, puromycin and triethylenemelamine; toDr. G. H. Hitchings of The Wellcome Research Labora-tories for the samples of 8-azaguanine, 6-chloropurine,Daraprim (2,4-diamino-5(p-dichlorophenyl)6-ethylpy-rimidine, (2,4-diamino-5(3'4'-dichlorophenyl)6-ethyl-pyrimidine, 2,6-diaminopurine, 6-mercaptopurine andmyleran (1, 4-dimethanesulfonoxybutane); to Dr. J.Ehrlich of Parke, Davis & Co. for azaserine and6-diazo-5-oxo-L-norleucine; to Dr. R. E. Handschu-macher of Yale University for 6-azathymine and 6-aza-uracil; to Dr. S. Wakaki of Kyowa FermentationIndustry, Tokyo, for carzinophilin; to B. F. GoodrichChemical Co. for beta-propiolactone; to Dr. H.Druckrey of Freiburg University for 2,5-bis(ethylene-imino)benzoquinone, N-methyl-bis(,B-chloroethyl)-am-ine HC1, and N-methyl-bis(,B-chloroethyl)amine-N-oxide-HCl; to E. R. Squibb & Sons, Inc. for a sampleof kinetin; and to J. T. Baker Chemical Co. for sul-focidin.The senior author wishes to express his gratitude to

the United States Government for providing the travelgrant and postdoctoral fellowship. We would also liketo thank Dr. S. A. Waksman and Dr. Vernon Brysonfor continued interest and helpful suggestions. Theskillful technical assistance of Miss Clara E. Biro isgratefully acknowledged.

* Cf footnotes 2 and 3 in table 1. Relative mutagenicity of0 indicates lack of mutant colonies on nutrient agar usuallydue to excessive killing.

28 [VOL. 6

on April 4, 2020 by guest

http://aem.asm

.org/D

ownloaded from

METHODS FOR DETECTION OF CHEMICAL MUTAGENS

SUMMARYTwo simple methods for detecting bacterial mutagens

have been devised with the following application inmind: an efficient, large scale survey of pure chemicalsand natural complex substances for mutagenic activityto serve as a guide for the study of carcinogenicity, andin the search for effective antineoplastic agents.The change in the frequency of streptomycin-

independent mutants in the streptomycin-dependentstrain of Escherichia coli is utilized as a measure ofmutagenicity. The substance is assayed for muta-genicity at a wide range of concentrations on a singleplate employing the paper-disc method. The membranemethod permits controlling the concentration and timeof exposure to the drug. Both simplified methods were

compared with the quantitative assay of Demerec andco-workers and found to yield closely correlated results.Seven out of 36 substances examined were markedlymutagenic; for two of them, 2,5-bis(ethyleneimino)-benzoquinone and the antibiotic sulfocidin, muta-genicity was ascertained for the first time by the paper-

disc method. All of the seven substances, and severalothers shown by the described methods to be weaklymutagenic for E. coli, are known to be endowed withantineoplastic properties. Both methods have the par-

ticular advantage of being selective for mutagens witha high ratio of mutagenicity to toxicity.

REFERENCES

AUERBACH, C. L. 1951 Problems in chemical mutagenesis.Cold Spring Harbor Symposia Quant. Biol., 16, 199-213.

BERTANI, G. 1951 A method for detection of mutations,using streptomycin dependence in Eschericia coli. Ge-netics, 36, 598-611.

BOYLAND, E. 1954 Mutagens. Pharmacol. Rev. 6, 345-364.BRYSON, V. 1948 The effects of nitrogen mustard on Esch-

erichia coli. J. Bacteriol., 56, 423-433.

BURDETTE, W. J. 1955 The significance of mutation inrelation to the origin of tumors: a review. Cancer Re-search, 15, 201-226.

DEMEREC, M. 1954 Genetic action of mutagens. Caryo-logia, 6, Suppl., 201-217.

DEMEREC, M., BERTANI, G., AND FLINT, J. 1951 A surveyof chemicals for mutagenic action on Escherichia coli.Am. Naturalist, 85, 119-136.

DEMEREC, M. AND HANSON, J. 1951 Mutagenic action ofmanganous chloride. Cold Spring Harbor SymposiaQuant. Biol., 16, 215-227.

GELLHORN, A. AND HIRSCHBERG, E., Editors 1955 Investi-gations of diverse systems for cancer chemotherapy screen-ing. I. Summary of results and general correlations. Can-cer Research, Suppl. No. 3, pp. 1-13.

GLOVER, S. W. 1956 A comparative study of induced re-versions in Escherichia coli. Carnegie Inst. WashingtonPubl. 612, Genetic Studies with Bacteria, pp. 121-136.

HEMMERLY, J. AND DEMEREC, M. 1955 Tests of chemicalsfor mutagenicity. Cancer Research, Suppl. No. 3, pp.69-75.

HOROWITZ, N. H., HOULAHAN, M. B., HUNGATE, M. G., ANDWRIGHT, B. 1946 Mustard gas mutations in Neurospora.Science, 104, 233-234.

MATNEY, T. S. 1955 New uses of membrane filters. I. Thedetermination of the spontaneous mutation rate of Esch-erichia coli to streptomycin resistance. J. Bacteriol.,69, 101-102.

MULLER, H. J. 1927 Artificial transmutation of the gene.Science, 66, 84-87.

SCHERR, G. H., FISHMAN, M., AND WEAVER, R. H. 1954 Themutagenicity of some carcinogenic compounds for Esch-erichia coli. Genetics, 39, 141-149.

STADLER, L. J. 1928 Mutation in barley induced by X-raysand radium. Science, 68, 186-187.

SZYBALSKI, W. 1952 Gradient plate technique for study ofbacterial resistance. Science, 116, 46-48.

TATUM, E. L. 1946 Induced biochemical mutations in bac-teria. Cold Spring Harbor Symposia Quant. Biol., 11,278-283.

WITKIN, E. M. 1947 Mutations in Escherichia coli inducedby chemical agents. Cold Spring Harbor Symposia Quant.Biol., 12, 256-264.

19581 29

on April 4, 2020 by guest

http://aem.asm

.org/D

ownloaded from