Vol. 42, No. 1APPLIED AND ENVIRONMENTAL MICROBIOLOGY, July 1981, p. 135-1410099-2240/81/070135-07$02.00/0

Specific Lysogenicity in Streptomyces azureusSEIYA OGATA,* SADAZO YOSHINO, HIKARU SUENAGA, KOJI AOYAMA,t NAKAO KITAJIMA,t

AND SHINSAKU HAYASHIDALaboratory ofApplied Microbiology, Department ofAgricultural Chemistry, Kyushu University, Fukuoka,

812, Japan

Received 30 March 1981/Accepted 6 April 1981

Slope (or plate) cultures of thiostrepton-producing Streptomyces azureus(ATCC 14921) often showed spontaneously developing plaques. Plaques increasedin number during serial subcultures. The production of aerial mycelia andsporulating aerial hyphae was interrupted by the overlapping plaques, whereasthe growth of substrate mycelia continued in the plaques. These abnormal(eroded) cultures were easily restored to their normal conditions once they werepassed through liquid cultures under shaking conditions. A few phage particleswere found in the plaques, together with some headless tails and numerous tailtips which formed a hexagonal crystal or a large crystal mass when viewed in anelectron microscope. No lytic phenomenon and no phage production were foundin the liquid cultures, although all mycelia and spores harbored phage-producingabilities. It was also found that the propagation of phages was successful in solidculture, but not in liquid culture. The whole phage was named SAt2, whichbelongs to group B of Bradley's morphological classification. From these results,it is considered that S. azureus is lysogenic with temperate phage SAt2, of whichvirulent mutants are able to infect the aerial mycelia and sporulating hyphae oftheir lysogenic host.

Thiostrepton-producing Streptomyces azur-eus forms unique colonies or lawns with plaquesin solid medium. The plaques never show clearmorphology, and the growth of substrate myce-lia continues in them. When plaque-carryingcultures were used for the fermentation of drugas seed cultures, the growth of mycelia wasremarkably delayed, and drug production de-creased. Therefore, this study was undertakento understand this unique lytic phenomenon andto clarify the lytic factor.

It was established that the plaques are notdue to the contamination of miscellaneousphages into the cultures, but appear spontane-ously. Thus, it is supposed that this strain pro-duces a phage to which it is itself susceptible.Many reports on various aspects of lysogeny inthe genus Streptomyces have been published.However, up to now, we fail to find a similarlysogenicity described in these reports. However,a similar phenomenon is often seen in someother bacteria such as Bacillus cereus (10), Ba-cillus entomocidus (11), and Staphylococcus au-reus (3). It is probable that these strains carryspecific temperate phages, which frequently mu-tate to virulent forms capable of attacking theirlysogenic host. It is reported that spontaneouslysis or eroded growth of the cultures is due to

t Present address: Food and Fine Chemical Plant, AsahiChemical Industry Ltd., Nobeoka, 882, Japan.

the virulent mutants. However, S. azureus willattract attention because lytic phenomenon andphage production occur only in solid cultures,unlike other bacteria.

MATERIALS AND METHODSOrganisms. S. azureus ATCC 14921 was used

throughout this work. About 60 strains of typicalknown Streptomyces species were supplied by A. Seino(KCC Culture Collection of Actinomycetes, KakenChemical Ltd.). These strains were used for the detec-tion of phage host range.Media and cultural conditions. Bennetti broth

and agar (glucose, 1%; meat extract, 0.1%; yeast ex-tract, 0.1%; NZ-amine type A, 0.2%; agar 1.5% (hard)or 0.6% (soft); pH 7.2) and Rye flakes agar (rye flakes,1%; glucose, 0.2%; yeast extract, 0.1%; CaCO3, 0.2%;agar, 1.5%; pH 7.2) were used for the growth of orga-nisms and plaque formation of phages. Rye flakes agarwas more suitable than other media for the productionof aerial mycelium and sporulating aerial hypha of S.azureus. The temperature of incubation was 28°C.

Preparation of spore suspension. Spores werescratched from 7-day-old cultured slants or plates ofRye flakes agar and suspended in Bennetti broth.Chains of spores were broken down with a mixer or

homogenizer (PCU-2, Kinematica). The mycelia andremaining chains of spores were removed by repeatedfiltration through filter paper (Toyo roshi no. 2). Allexperiments were started from this spore suspension.The spore suspensions of other strains were performedby the method for S. azureus.Colony formation. Plates were inoculated with a

135

on March 23, 2021 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

diluted suspension of spores to make about 40 coloniesat the highest dilution and incubated for 7 days.Among the colonies developed, there were 5 to 8%abnormal colonies in sporulation. About 2,000 normalcolonies were isolated from about 60 plates, and thesewere used for the following process. Half of them weretransferred one by one to 15 ml of fresh Rye flakesagar in plates (diameter, 9 cm) to make nine coloniesper plate, and the remainder was transferred one byone to 4 ml of fresh Rye flakes agar in small vials (size,27 by 55 mm; volume, 20 ml) with a cotton plug tomake a single colony per vial. They were incubatedfor 7 days at 28°C. Then, these colonies were subcul-tured every 7 days in the same type of apparatus withRye flakes agar.On the other hand, single-colony isolation (single-

spore isolation) was repeated five times. At each time,about 200 colonies were isolated and subcultured insmall vials, as described above.

Preparation of phages. The semiconfluent platelysis method was adapted for the phage preparation(2). The spores obtained from abnormal (eroded) cul-tures were inoculated on top agar of Bennetti mediumor Rye flakes agar and then incubated for 3 to 5 days.Occasionally, mixtures of spores from normal and ab-normal cultures were used instead of a single kind ofspore. To the overlapping plaques of each plate, 5 mlof 0.05 M tris(hydroxymethyl)aminomethane-hydro-chloride buffer (pH 7.2) or ammonium acetate buffer(pH 6.0) was added, and the plates were allowed to sitat room temperature for 5 h. The phage extract washarvested by gentle pouring and then centrifuged toremove the mycelia and their debris. This extractusually showed a titer of from 3,000 to 5,000 plaque-forming units per ml by the spot test described below.Occasionally, the phage suspension was concentratedby high-speed centrifugation, and the resulting pelletwas resuspended in ammonium acetate buffer for elec-tron microscopy. Also, the phage extract was filtratedthrough a membrane filter (0.45 ,um).Overlapping invasion test. This test was per-

formed to investigate the infectious abilities of abnor-

mal (eroded) colonies. The film of spores of myceliaderived from an abnormal colony was partially over-lapped on that of a normal colony on the surface ofRye flakes agar plates by use of a loop and the cross-brush technique. The plates were incubated until twocolonies became confluent, giving a lawn of myceliumgrowth.Stamp test. This test was performed to investigate

the infectious abilities of plaques. A single plaque wasplaced on the plate early in the growth of normalmycelia by use of a loop. The plate was also incubatedas described above.Spot test (assay method for phage). The phage

extract had a low titer, and the indicator employedfrequently produced spontaneous plaques. Therefore,the phage activity was assayed according to the dou-ble-layer method for bacteriocins (5). Twofold serialdilutions of the phage suspension were prepared andexamined by the spot test on Bennetti agar seededwith the spores obtained from a normal culture. Theactivity was defined at the highest dilution clearlyshowing several plaques or an inhibitory zone. Thehost range of phage was also examined by this method.Photography of colony morphology. Photo-

graphs of colony morphology were taken with a NikonFE camera (Nippon Kogaku Kogyo Ltd.), using Fujifilm F.

Electron microscopy. The above preparation ofphages in ammonium acetate buffer (pH 6.0) wasnegatively stained with potassium phosphotungstate(pH 6.0) and placed on grids coated with collodion-carbon. Electron micrographs were taken with a JEM-100 B electron microscope (Japan Electron OpticusLaboratory Ltd.). Also, the samples made directlyfrom a single plaque or overlapping plaques weredirectly suspended in potassium phosphotungstate.

RESULTSMorphology of colonies. The subcultured

colonies were morphologically divided into threegroups, A, B and C, as shown in Fig. 1. Nornal

FIG. 1. Various colonies which appeared during serial subcultures. A, Normal colony; B and C, abnormalcolonies. A type colony is rich in aerial mycelia and spores. Bl and B2 type colonies have one to three plaques.B3 and B4 type colonies have scores of plaques. B5 and B6 type colonies show overlapping plaques orsemiconfluent lysis. The C type colony has few aerial mycelia and no spores. Bar, 1 cm.

136 OGATA ET AL.

on March 23, 2021 by guest

http://aem.asm

.org/D

ownloaded from

SPECIFIC LYSOGENICITY IN S. AZUREUS 137

(healand:B andefirseriaerodtatiosomEcolorvelolplaqiopedmycbut i

Ficolocultumalin tlabncthos

100

90

80U,

z 700

60

50

Z 400p 30

i 20

10

a

Ithy) colonies were designated as A type, tions. The colonies derived from the repeatedabnormal (eroded) colonies were designated single-spore isolation also displayed the sameid C types. Bi and B2 type colonies were behavior as did those from an original slantitely derivative of A type colonies during culture. The higher frequency of appearance ofI subcultures. The number of plaques in the abnormal colonies in the plates may be due toed colonies increased by repeated transplan- inner infection by infectious factors in the*n in numerical order of the B group, with plaques, because nine colonies grew together ina exceptions; occasionally, Bi and B2 type a plate; on the other hand, in a vial, only a singlenies with one to three plaques directly de- colony grew. No genetically healthy coloniesped to B5 type colonies with overlapping were isolated among 3,000 colonies used in thisues. Ultimately, all B group colonies devel- experiment.I to C type colonies, in which few aerial We could temporarily cure their illness by theelia and no sporulating aerial hypha grew, following treatment. The spores or mycelia orin which substrate mycelium did grow. both of eroded colonies were inoculated intogure 2 shows the appearance of abnormal Bennetti broth and incubated for more than 24nies from normal colonies during serial sub- h under shaking conditions. After that, the grow-ires. The frequency of appearance of abnor- ing mycelia and flakes were reinoculated on acolonies in the plates was higher than that slant or plate of Rye flakes agar. The lawnie vials. All colonies in the plates became developed was characterized by the productionrmal within six transplantations, whereas of typical aerial mycelia, sporulating hypae, ande in the vials did so within 10 transplanta- numerous spores superimposed upon the sub-

strate growth. In the lawn, there were no plaquesor occasionally a few plaques. However, thesecultures again became abnormal by serial sub-cultures.

nfiectious ability of plaques. Infectious ac-tivities of plaques in B type and C type colonieswere determined by the overlapping invasiontest and stamp test. Figures 3A and B show theresults of the overlapping invasion test betweenA type colony and B4 type or C type colony,respectively. The abnormal colonies invadedinto the A type colony and inhibited its growth.In the invaded area and on the border line,plaques developed. Although there were obser-vationally no plaques in the C type colony, ithad plaque-producing activity. Figure 3C showsthe result of the stamp test. The stamped areamade a zone of inhibition in the growth of aerialmycelium and sporulating hypha. From these

0 1 2 3 4 s 6 7 8 9 lo results, the plaques had infectious activities or

SUBCULTURED TIMES factors.

[G. 2. Appearance of plaque-carrying colonies A spot test of the filtrated extract obtainedng serial subcultures. The cultures of each line from overlapping plaques was made on the sur-started from 1,000 colonies. 0, Colonies grown face of a top agar layer seeded with normal

etri dishes; 0, colonies grown in small vials. spores. Figure 3D shows the plaques on the lawn

FIG. 3. Infectious activity ofplaque. A and B show the results of the overlapping invasion test. C shows theresult of the stamp test. D shows the result of the spot test. Bar, 1 cm.

Fidurinwerein pt

VOL. 42, 1981

on March 23, 2021 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

of a normal culture caused by a spot drop of 16-fold in a series of twofold dilutions.From these results, it is clear that the eroded

colonies and plaques contained some infectiousand filterable agents. It is supposed that theagents are able to self-multiply in the specificmycelia.Detection ofinfectious agents by electron

microscopy. Electron microscopy confirnedthe presence of phages and their related parti-cles. Whole phages and their ghosts were foundin very small amounts only in a concentratedpreparation (Fig. 4). This phage was namedSAt2 and belongs to group B of Bradley's mor-phology (1). The head had a hexagonal outlineabout 60 nm in diameter, and the tail was non-contractile and about 210 nm long and 11 nmwide. The headless tails were more easily found

FIG. 4. Electron micrographs ofphage SAt2 of S.azureus. Bar, 50 nm.

even in nonconcentrated preparations (Fig. 5).Also, there were tailless heads in some prepara-tions (Fig. 5A). Thus, the headless tail probablyremained after the head was broken off fromwhole phage or its ghost. Tail tip-like particleswere found in numerous amounts. Their distalends formed hexagonal crystals, as shown in Fig.6; a single hexagonal crystal consisted of sevenbasic units (Fig. 7A-C, and E). When a prepa-ration was made directly from a plaque or over-lapping plaques, the characteristic crystallinemass was somewhat larger, as shown in Fig. 6B.In the crystals, three types of basic units couldbe seen: one had no discernible internal structure(Fig. 7C and F), the second had a hollow innercore, and the third had a full inner core. Twelve(or more) appendages seemed to be arrangedradially around the outer core of 18 nm (Fig.6A). Some crystals consisted of headless tails, asshown in Fig. 5C. Although most all of the basicunits of crystals showed the distal view, someother basic units showed the lateral view (someparticles of Fig. 6B and arrow particles of Fig.7D-F). Some particles of lateral view were at-tached with short hollow tubes (inner cores).Also, there may be some particles of oppositeview, although they are not discriminated fromthose of distal view. The small pieces of tailinner core with a width of 7 nm were seen asthin hollow tubes in Fig. 7G and H. These par-ticles morphologically resembled the tail tip andtail constituents of phage SAt2. The schematicmorphologies of phage SAt2 and a tail tip-likeparticle are shown in Fig. 8, although the finestructure of the tail tip-like particle is beingstudied with the rotation technique (9).Host range. A spot test indicated that the

plaques could not develop on any strains used(S. caelestis, S. bellus, S. coeruleorubidus, S.

FIG. 5. Electron micrographs of headless tail and tailless head. Bar, 50 nm.

138 OGATA ET AL.

on March 23, 2021 by guest

http://aem.asm

.org/D

ownloaded from

SPECIFIC LYSOGENICITY IN S. AZUREUS

FIG. 6. Electron micrographs ofcrystal mass oftail tip-like particles. Thepreparation ofA was made frombuffer extract and the preparation ofB was made directly from plaques as described in the text.

FIG. 7. Electron micrographs of various tail tip-like particles and pieces of tail inner core. A, B, C and Dshow a hexagonal crystal of tail tip-like particles. In C, three types of basal units are seen. The arrows showthe lateral view of tail tip-particles. G and H show the pieces of tail inner core. Bar, 50 nm.

coerulescens, S. curacoi, S. cyaneus, S. glauces-cens, 5 strains of S. viridochromogenes, 12strains of S. griseus, 9 strains of S. lavendulae,S. niveus, S. spheroides, S. venezuelae, S. vir-giniae, S. purpurascens, S. fradiae, S. alboni-ger, S. albus, S. rimosus, S. achromogenes, S.catenulae, S. echinatus, S. eurythermus, S. al-bogriseolus, S. ambofaciens, S. aureofacience,S. endus, S. griseoluteus, S. humidus, S. oliva-

ceus, S. rochei, S. sparsogenes, S. aureus, S.bottropensis, S. capreolus, S. humifer, S. orien-talis, and S. toyocaensis). However, the growthof S. cyaneus (KCC S-0220) and S. coerulescens(KCC S-0218) were inhibited. These strains be-long to blue spore-forming organisms that arethe same as S. azureus. The inhibition may bedue to the killing action of phage tails or tail tip-like particles. The bactericidal action will be

VOL. 42, 1981 139

on March 23, 2021 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

118 Tm

; ) b C *

*-@~~~~~.W 3.5.3S

1118

FIG. 8. Schematic diagrams ofphage SAt2 and distal view of tail tip particles. A, Lateral view ofphageSAt2; B, distal view of tail tip-like particles; a, particle with a full inner core; b, particle with a hollow innercore; c, particle with no discernible internal structure. Bar, 50 nm.

studied in the future.No production of phage and its related

particles in the liquid culture. The followingexperiment was performed to investigatewhether or not phage SAt2 and related particleswere reproduced or spontaneously produced inthe liquid culture. The spores from a normal or

an abnormal slant were inoculated into Bennettibroth and incubated for 6 h, at which time theculture broth was divided equally into two parts.The phage suspension was added to one culture,and the heated phage suspension was added toa control culture. After 24 and 48 h of shakingcultivation, the reproduction or spontaneousproduction of phage SAt2 and related particleswas investigated by an assay method and elec-tron microscopy. As shown in Table 1, the addedphages could not multiply in the liquid culture.Furthermore, it was also clear that their spon-taneous production did not occur in both theliquid cultures made from normal and abnormalcultures.

DISCUSSIONLysogenic strains of bacteria are usually im-

mune to the phage and more common virulentmutant which they produce, and immunity isconferred on them by the prophage (8). Butsome virulent mutants such as Avir (derived fromphage A) of Escherichia coli (4, 6), W,8 (derivedfrom phage Wa) of Bacillus cereus (10), and vl(derived from phage 4C31) of Streptomyces coe-licolor (7) can infect their producer strains ly-sogenic with their wild-type phages and can formclear plaques; virulent mutants of phages 4C31and A occur with a much lower frequency than

TABLE 1. No reproduction and no spontaneousproduction ofphage SAt2 progenies in liquid

culturePlaque-forming unitsper ml after the fol-

Experimental condition lowing incubation:

24h 48h

Normal culture' (control) 0 0Normal culture" plus phagesc 30 1Abnormal culture' (control) 3 0Abnormal culture' plus phagesc 50 1

a Normal culture was derived from the spores of anormal slant.

b Abnormal culture was derived from the spores ofan abnormal slant.

c Phage SAt2 was added to yield a titer of 2,000plaque-forming units per ml.

those of phage Wa. It is clear that they lose (orreduce) the affinity for their characteristic im-munity substance (repressor). In the case of B.cereus and Bacillus entomocidus, this type ofvariant appears at a high frequency in both solidand broth cultures. So, solid cultures often showspontaneously developing phage plaques byplating from old broth cultures and during sub-cultures. The spontaneous plaque formation ofS. azureus in the solid medium was very similarto that of B. cereus and others. The most prob-able explanation for this phenomenon is thatplaques are formed by a virulent mutant ofphage SAt2 with which S. azureus is lysogenic.However, we can recognize some differences be-tween S. azureus and other bacteria. A greatdifference was that the phages formed uniqueplaques containing the growth of substrate my-

A

B

140 OGATA ET AL.

on March 23, 2021 by guest

http://aem.asm

.org/D

ownloaded from

SPECIFIC LYSOGENICITY IN S. AZUREUS 141

celia. Also, it was confirmed that phage SAt2was not produced in liquid culture. The simplestexplanation for these characteristic behaviors ofphage SAt2 is that it can infect the aerial mycelia(or sporulating hyphae or both) but not sub-strate mycelia, and their induction does not oc-cur in the liquid culture, although all kinds ofmycelia and spores harbor phage genome. Atpresent, we cannot elaborate on these character-istic behaviors. However, it is supposed thatsuch different behaviors between substrate my-celium and aerial mycelium are thought to bedue to the quality and quantity of phage-specificimmunity or the presence of phage receptors oncell surface or both. Also, the mycelia developedin liquid culture may have the same propertiesas do substrate mycelia in solid culture. Hence,we can definitely conclude that such peculiarbehavior of S. azureus, unlike other bacteria, isowing not to the properties of phage, but to theproperties of differentiated mycelia of the hostorganism. We should like to clarify this problemin the near future.Whole phages were liberated after sponta-

neous production or after reproduction in sus-ceptible mycelium, but only in very smallamounts. For instance, one to three particles,including ghost phages, per field of vision werefound in a preparation, whereas there were 20 to30 headless tails in the same preparation. Sincea few tailless heads were also found, the headlesstails must be particles detached from wholephages. On the other hand, the number of tailtip-like particles was numerous, so they oftenformed a large crystal mass. These particles maybe the tail components of phage SAt2. Variousideas about the formation of these particles fallinto two basic categories: degradation and in-completion hypotheses. Degradation from wholephages to tail tip particles is illogical, becausethere are many more tail tip particles than wholephages and detached tails. Probably, tail tip-likeparticles are incomplete substances of phagetails. The logical explanation for the higher pro-

duction of tail tips is that the synthesis or assem-bly of various phage constituents, except tailtips, is remarkably depressed in S. azureus;therefore, more tail tips should accumulate, asdo the precursors of phage tails. The pieces ofinner core must have been formed by the break-down of tail tip particles with inner cores, be-cause there were few of them. A demonstrationof this would require much more informationand data.

ACKNOWLEDGMENTSWe are deeply grateful to H. Eguchi (Biotron Institute,

Kyushu University) for his helpful advice on photographs andto J. R. Norris (Cadbury Schweppes Ltd.) for his helpfulinformation on the phages of B. cereus. Thanks are also givento Y. Minematsu-Yoshino, T. Kono-Fujisawa, and Y. Koyamafor their helpful technical assistance.

This work was partially supported by the Mishima KaiunFoundation.

LITERATURE CITED

1. Bradley, D. E. 1967. Ultrastructure of bacteriophagesand bacteriocins. Bacteriol. Rev. 31:230-314.

2. Eisenstark, A. 1967. Bacteriophage techniques, p. 449-524. In K. Maramorosch and H. Koprowski (ed.), Meth-ods in virology. Academic Press, Inc., New York.

3. Fisk, R. T. 1942. Studies on Staphylococci. 1. Occurrenceof bacteriophage carriers among strains of Staphylococ-cus aureus. J. Infect. Dis. 71:153-160.

4. Jacob, F., and E. C. Wollman. 1953. Induction of phagedevelopment in lysogenic bacteria. Cold Spring HarborSymp. Quant. Biol. 18:101-121.

5. Jacob, F., and E. C. Wollman. 1959. Colicins and otherbacteriocins, p. 381-393. In M. H. Adams (ed.), Bacte-riophages. Wiley-Interscience Publishers, Inc., NewYork.

6. Kaizer, A. D. 1955. A genetic study of the temperatecoliphage X. Virology 1:424-443.

7. Lomovskaya, N. D., K. F. Chater, and N. M. Mkrtu-main. 1980. Genetics and molecular biology of Strep-tomycetes bacteriophages. Microbiol. Rev. 44:206-229.

8. Lwoff, A. 1953. Lysogeny. Bacteriol. Rev. 17:269-337.9. Markham, R., S. Frey, and G. J. Hills. 1963. Methods

for the enhancement of image detail and accentuationof structure in electron microscope. Virology 20:88-102.

10. McCloy, E. M. 1958. Lysogenicity and immunity to Ba-cillus phage W. J. Gen. Microbiol. 18:198-220.

11. Norris, J. R. 1961. Bacteriophages of Bacillus cereus andof crystal-forming insect pathogens related to B. cereus.J. Gen. Microbiol. 26:167-173.

VOL. 42, 1981

on March 23, 2021 by guest

http://aem.asm

.org/D

ownloaded from