Method for the Preferential Isolation of Actinomycetesfrom Soils

J. N. PORTER, J. J. WILHELM, AND H. D. TRESNER

Biochemical Research Section, Lederle Laboratories, American Cyananttid Company, Pearl River, New York

Received for publication October 21, 1959

Species of the genus Streptomyces and of relatedgenera of the Actinomycetales, at one time consideredto be microbiological curiosities of no great economicimportance, have become the subjects of intensivesearches for sources of new, biologically active com-pounds. The demonstrated ability of these micro-organisms to produce useful antibiotics and to carryout other transformations of commercial interest hasfocused attention on factors bearing on their isolationfrom their natural habitat, the soil. It would be de-sirable, therefore, to be able to isolate aerobic soil-inhabiting actinomycetes with a minimum of inter-ference from associated bacteria and higher fungi.

There have been a number of reports in the literaturedealing with selective isolation of the higher fungi atthe expense of bacteria or of bacteria at the expense ofmolds. Several investigators used antibiotics as theselective agents. For example, Beech and Carr (1955)found cycloheximide, gliotoxin, and frequentin to beeffective selective yeast and mold inhibitors in applejuices and ciders. Phillips and Hanel (1950) foundcycloheximide to be inactive against representatives of27 species of bacteria and suggested that this anti-fungal antibiotic might be used to rid bacterial culturesof contaminating molds. On the other hand, Butler andHine (1958) selectively isolated fungi from soil platedon potato dextrose agar adjusted to pH 5.6 to 6.1 andto which were added 100 ,g per ml of the antibioticnovobiocin.

Turning to preferential actinomycete isolation, Crooket al. (1950) found sodium propionate to be an effectivefungal inhibitor, useful for their Streptomyces program.Corke and Chase (1956), in their microbiological in-vestigations of heavily mold-infested acid forest soils,studied the use of both sodium propionate and cyclo-heximide as medium additives. They reportedcycloheximide to be the more effective compound ofthe two and recommended its addition to actinomyceteisolation media at a level of 40 ,ug per ml of agar. Intheir hands, cycloheximide at 100 jl.g per ml showed no

suppression of any of 85 actinomycete cultures.Dulaney et al. (1955) isolated Streptomyces selectivelyon a nutrient agar medium containing, per ml: cyclo-heximide, 20 units; polymyxin, 20 units; subtilin, 20units; and penicillin, 20 units. They stated that the

amounts of antibiotics in the medium were criticaland that, whereas the selective inhibition of molds couldbe accomplished, the separation of streptomnycetes andbacteria was a difficult achievement.The fact that the development of Streptomyces

colonies on agar plates could be favored over bacteriaby selection of the nitrogen source in the medium wasreported by Benedict et al. (1955). They noted thatL-arginine is readily attacked by most streptomycetesand they recommended the use of this amino acid as areplacement for the glycine of the glycerol-glycinemedium of Lindenbein (1952).We have achieved a set of conditions very favorable

to the preferential isolation of actinomycetes fromsoils by combining the principle of the selective inhibi-tion of molds by means of an antifungal antibiotic withthat of the use of Benedict's modification of the Linden-bein medium to limit bacterial development. A surface-layering technique enhances the effectiveness of theisolation procedure.

METHODS AND RESULTS

Plating procedure. A plating method described byKelner (1948) was employed in carrying out most ofthese studies. The technique is one which ensures thedevelopment of most microorganisms as surfacecolonies, thus greatly expediting their tentative identi-fication and comparison. Fifteen-ml aliquots of anappropriate medium were poured into Petri dishes andallowed to harden to form a basal layer. Soil sampleswere suspended in sterile water, appropriately dilutedand 1 ml of each desired dilution pipetted into 14-mlportions of agar medium maintained at 48 C in a waterbath. Five-ml quantities were then pipetted onto thehardened surfaces of two basal layers, thus providingduplicate plates for each dilution. The composition ofthe agar, including additives, of both basal and surfacelayers was identical. An exception to this procedure wasthe plating out by the conventional nonlayering methodof cultures contaminated with molds by design.

Nutrient media. A number of nutrient media, forexample, asparagine dextrose, Emerson, Czapek, yeastextract, and Bennett agars, are commonly used toinvestigate the growth and development of aerobicactinomycetes. Where bacteria are numerous in soils,

174

on Septem

ber 27, 2018 by guesthttp://aem

.asm.org/

Dow

nloaded from

ISOLATION OF ACTINOMYCETES FROM SOILS

none of these media are particularly satisfactory forstreptomycete isolation and some, such as yeast extractagar, actually potentiate spreading bacterial colonies.The writers, however, have obtained consistently goodresults with Benedict's modification of the Lindenbeinmedium.' That it is possible to obtain plates with thismedium on which streptomycetes are growing almostexclusively is shown in figures 1 and 2. These figuresalso demonstrate the virtual impossibility of isolatingactinomycetes from the same soil sample on someother media.

Antibiotic additives to culture media. Antibacterialantibiotics such as the tetracyclines, polymyxin, neo-

' Composition (per L): glycerol, 20 g; L-arginine, 2.5 g;NaCl, 1 g; CaCO3, 0.1 g; FeSO4*7H20, 0.1 g; MgSO4-7H20, 0.1 g;agar, 20 g.

mycin, and streptomycin alone or in combination wereadded to several agar media at levels ranging from0.5 to 100 ,g per ml. Selective inhibition of bacteriawas not satisfactorily achieved. However, since theBenedict medium modification achieved the purposeof favoring the actinomycete to bacterium colony ratio,attention was turned to the use of antifungal agents inthe medium.

In our antibiotic screening program a polyene anti-biotic, designated A-5283, with a broad spectrum ofactivity against fungi was isolated. This antibiotic,subsequently identified as pimaricin, was found to beeminently suitable for eliminating molds from soildilution plates at a level of 50 ,g per ml. Moreover, at100 ,g per ml, A-5283 was without observable effect onStreptomyces populations.

Figure 1. Isolation plates from a New Jersey soil, 3 X 10t dilution. Benedict agar (upper left), yeast agar (upper right), asparaginedextrose agar (lower left), Bennett agar (lower right).

1751960]

on Septem

ber 27, 2018 by guesthttp://aem

.asm.org/

Dow

nloaded from

Figure 2. Isolation plates from an Indiana soil, 3 X 10-8 dilution. Benedict agar (upper left), yeast agar (upper right), asparaginedextrose agar (lower left), Bennett agar (lower right).

Figure S. Isolation plates from a Mississippi soil. Top row, left to right: Benedict agar plates containing 0, 5, 25 and 50 iAg of anti-biotic A-5283 per ml of agar, respectively. Bottom row, left to right: Plates containing asparagine dextrose, Emerson, Czapek, yeast,and Pennett agar, respectively. No A-5283.

on Septem

ber 27, 2018 by guesthttp://aem

.asm.org/

Dow

nloaded from

ISOLATION OF ACTINOMYCETES FROM SOILS

Antibiotic A-5283 is water-insoluble and can with-stand autoclaving at 15 lb pressure for 15 min withless than 10 per cent loss of activity. A stock solution ofA-5283 at 1 mg per ml is prepared in N,N-dimethyl-formamide and dispensed in agar media at desiredlevels before autoclaving. It has been our experience,however, that agar media containing A-5283 cannot bekept for more than 3 days, even under refrigerationconditions, without further loss of antibiotic activity.Figures 3 and 4 illustrate dilution plates made from a

Mississippi soil and from a local greenhouse soil. Com-plete mold suppression is shown on Benedict agar platescontaining 50 ,g per ml of A-5283. Less completeinhibition is indicated at lower levels. Heavy moldgrowth and complete suppression of actinomycetes fromthe same soils occurred on asparagine dextrose,Emerson, Czapek, yeast, and Bennett agars lackingA-5283.The facility with which streptomycetes may be

selectively isolated from soil on the Benedict agar-A-

Figure 4. Isolation plates from a greenhouse soil. Arrangement of plates the same as in figure 3

Figuire 5. Plating on yeast-malt agar of Streptomyces lavendulae cultures contaminated with (from left to right) Neurospora crassa,Penicillitmi chrysogenurnt, Saccharomyces cerevisiae, Rhizopuis nigricans. Top Row: plates containing 50 Ag per ml of A-5283. Bottom row:plates withouit antibiotic A-5283.

1960] 177

on Septem

ber 27, 2018 by guesthttp://aem

.asm.org/

Dow

nloaded from

J. N. PORTER, J. J. WILHELM, AND H. D. TRESNER

5283 combination has led us to incorporate the tech-nique in our routine isolation program. In 521 soilsamples collected on a world-wide basis and studiedby this technique, the numbers of streptomycetes per gwere highly variable, normally ranging from 104 to 108.Bacteria were a serious handicap to isolation in the caseof 49 soils, and yeasts or molds were a problem in 11soils. In only one soil, however, were contaminantspresent in such numbers that the isolation of Strepto-myces by our method was impossible. For practicalreasons, actual comparisons with other isolation mediaon the same soils was not feasible. However, in an earlierphase of the program, unamended asparagine dextroseagar was used to isolate Streptomyces from 332 soils ofvarious origins. In the case of 92, contaminants were aserious problem and from 52 no streptomycete isolatescould be obtained.

Culture decontamination. Since A-5283 is such anefficacious antifungal agent for soil isolation work, theuse of this antibiotic in decontaminating mold-infestedStreptonyces cultures also suggested itself. To test theidea, a spore suspension of a strain of Streptomyceslavendulae was prepared in sterile saline and portionsmixed separately with each of the following fungi:Rhizopus nigricans, Penicillium chrysogenum, Saccharo-myces cerevisiae, and Neurospora crassa. Approxirfmately0.5-ml amounts of each mixture and of S. lavendulaespores alone were incorporated in yeast-malt agarcontaining 50 ,ug per ml of A-5283 in poured Petriplates. Plates without A-5283 were included. The anti-biotic A-5283 had no visible effect on the devrelopmentof the Streptomyces cultures but the growth of thefungal contaminant was completely inhibited in allcases (figure 5). Several mold-contaminated Strepto-myces cultures in our collection have been successfullyfreed of the fungus contaminant by the use of thisantibiotic.At appropriate concentrations, other antifungal anti-

biotics can be employed effectively both in soil isolationplates and in decontamination work. We have found

both nystatin and cycloheximide at 50 ,ug per ml to besatisfactory in these respects.

SUMMARYA method is described of combining an antifungal

antibiotic with a glycerol-arginine medium whichselectively encourages the development of aerobic, soil-inhabiting actinomycetes on isolation plates. Surfacelayering of the inoculated agar ensures the growth ofmost organisms as surface colonies. Such antifungalantibiotics as pimaricin (A-5283 from our antibioticprogram) nystatin, and cycloheximide were used suc-cessfully to suppress molds in isolation plates. Theseantibiotics were also found to be useful in decontaminat-ing mold-infested Streptomyces cultures.

REFERENCESBEECH, F. W. AND CARR, J. G. 1955 A survey of inhibitory

compounds for the separation of yeasts and bacteria inapple juices and ciders. J. Gen. Microbiol., 12, 85-94.

BENEDICT, R. G., PRIDHAM, T. G., LINDENFELSER, L. A., HALL,H. H., AND JACKSON, R. W. 1955 Further studies in theevaluation of carbohydrate utilization tests as aids in thedifferentiation of species of Streptomyces. Appl. Micro-biol., 3, 1-6.

BUTLER, E. E. AND HINE, R. B. 1958 Use of novobiocin forisolation of fungi from soil. Soil Sci., 85, 250-254.

CORKE, C. T. AND CHASE, F. E. 1956 The selective enumera-tion of actinomycetes in the presence of large numbers offungi. Can. J. Microbiol., 1, 12-16.

CROOK, P., CARPENTER, C. C., AND KLENS, P. F. 1950 Theuse of sodium propionate in isolating actinomycetes fromsoils. Science, 112, 656.

DULANEY, E. L., LARSEN, A. H., AND STAPLEY, E. 0. 1955 Anote on the isolation of microorganisms from naturalsources. Mycologia, 47, 420-422.

KELNER, A. 1948 A method for investigating large microbialpopulations for antibiotic activity. J. Bacteriol., 56,157-162.

LINDENBEIN, W. 1952 Uber einige chemisch interessanteAktinomycetenstamme und ihre Klassifizierung. Arch.Mikrobiol., 17, 361-383.

PHILLIPS, G. B. AND HANEL, E. 1950 Control of mold con-taminants on solid media by the use of Actidione. J.Bacteriol., 60, 104-105.

178 [VOL. 8

on Septem

ber 27, 2018 by guesthttp://aem

.asm.org/

Dow

nloaded from