JOURNAL OF BACTERIOLOGY, OCt. 1968, p. 902-908Copyright @ 1968 American Society for Microbiology

Vol. 96, No. 4Printed in U.S.A.

Relation of Mucoid Growth of Staphylococcus aureus

to Clumping Factor Reaction, Morphology in Serum-Soft Agar, and Virulence

KOSAKU YOSHIDA1 AND RICHARD D. EKSTEDTDepartment of Microbiology, Northwestern University Medical School, Chicago, Illinois 60611

Received for publication 8 July 1968

The growth characteristics of several strains of Staphylococcus aureus in BrainHeart Infusion and in a modified Staphylococcus Medium No. 110 were compared.In the latter medium all of the strains studied showed an increased mucoid character.Some of the strains studied showed a greater potential to synthesize excess slime layermaterial than others. The highly mucoid strains grew as diffuse-type colonies inmodified Staphylococcus Medium No. 110 serum-soft agar and reacted as thoughthey were negative in the test for clumping factor. These strains were also found to bemore virulent when used to challenge normal mice intraperitoneally.

One of the most discouraging aspects ofresearch in the field of infection and resistance toStaphylococcus aureus has been the lack of asuitable laboratory experimental system. SinceHunt and Moses (7) described the greater mousevirulence of the diffuse variant of the Smithstrain of S. aureus, several investigators haveused this organism, or a strain with similarproperties, to study various aspects of experi-mental staphylococcal disease (1, 5, 6, 10, 13, 15).Our own studies with the Smith diffuse strain ofS. aureus (2, 3) suggested that, in addition to the"Smith Surface Antigen" described by Morse(11), teichoic acid from this strain also shows invitro immunological properties and is an impor-tant protection-inducing antigen in mice.

In an extension of these studies to other non-Smith-like strains of staphylococci, we wereagain faced with the problem of virulence of theseorganisms for mice. In this paper, we report amethod which has been successful in enhancingthe mouse virulence of several strains of S. aureusand in inducing them to exhibit growth charac-teristics similar in some respects to the Smithdiffuse-type strains.

MATERIALS AND METHODSOrganism. The diffuse and compact variants of the

Smith strain and of the SK4473 strain were obtainedfrom M. G. Koenig, Vanderbilt University School ofMedicine, Nashville, Tenn. The 258, 334, and 418strains were obtained from M. Silverman, Veteran's

1 Present address: Department of Bacteriology,Nippon Medical School, Bunkyo-Ku, Tokyo, Japan.

Administration Research Hospital, Chicago, Ill., asrecent clinical isolates. The MD strain was isolatedfrom an abscess in a dog, and the S. aureus type Istrain was obtained from the American Type CultureCollection. All strains were positive in the test forclumping factor, with the exception of the Smithdiffuse strain. They were all positive for free coagulasein the tube test and were hemolytic on rabbit blood-agar.

Culture medium. In addition to the routine use ofBrain Heart Infusion (BHI) broth (Difco) preparedas suggested by the manufacturer, a modified dialysatemedium similar in its composition to StaphylococcusMedium No. 110 (Difco) was prepared as follows:2.5 g of yeast extract (Oxoid) and 10 g of peptone(Difco) were dissolved in 50 ml of 3.0% NaCI andplaced into a dialysis bag. The bag was dialyzedagainst 950 ml of 3.0% NaCl for 2 days at 4 C with amagnetic stirrer. To the dialysate we then added 10 gof mannitol, 2 g of lactose, and 5 g of K2HPO04. Agar(Difco), to a concentration of 1.5%, was added to aportion of this medium, and the remainder was pre-pared as a broth. In preliminary experiments, we foundthat better growth is obtained when the agar mediumis supplemented with additional buffer. Therefore,before sterilization, phosphate buffer (pH 7.4) wasadded to the modified 110 agar to a final concentra-tion of 0.05 M. The media were autoclaved at 121 Cfor 10 min and were stored at 4 C until used.

Preparation ofcell suspensions and challenge method.The organisms used in these studies were maintainedby monthly transfer on BHI agar slants. To preparestandardized cell suspensions for challenge experi-ments, the organisms were subcultured into modified110 broth and were incubated at 37 C for 18 hr. A5-ml amount of these cultures was then pipetted ontothe surface of medium 110 agar plates containing ap-proximately 30 ml of medium. The plates were incu-

902

on Novem

ber 9, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

MUCOID GROWTH OF STAPHYLOCOCCUS AUREUS

bated for an additional 18 hr at 37 C. After incuba-tion, the excess broth was gently discarded and theorganisms were carefully washed three times with coldfresh media while still on the plates. The organismswere then scraped into small tubes containing coldfresh media and were allowed to settle for 10 to 15min at room temperature. The supernatant fluids werediscarded, and cold fresh medium was added. The or-

ganisms were then dispersed with an inoculating loop.The cell suspensions were again allowed to settle for 15min at 4 C to remove large clumps. The organismsfrom the upper part of suspensions were then adjustedturbidimetrically at 430 nm in a Coleman model 9nephlocolorimeter, with the fresh modified 110 brothas a blank. Plate counts of appropriate dilutions ofthese suspensions showed that an optical density(OD) of 0.5 corresponded to 1.8 X 108 to 3.3 X 108colony-forming units (CFU) per ml. Ten times con-centrated suspensions were prepared by centrifugingthe 0.5 OD cell suspensions and resuspending themin 0.1 volume. Ten-fold dilutions were made from the0.5 OD cell suspensions with fresh modified 110 broth.Organisms grown on BHI agar plates were used as a

control. A 5-ml amount of BHI broth cultures wastransferred onto BHI agar plates containing 30 ml ofBHI agar and was incubated at 37 C for 18 hr. Theorganisms were washed with cold modified 110 broththree times, and the 0.5 OD cell suspensions were

prepared and diluted in cold fresh modified 110medium as described above. Plate counts of these sus-pensions varied from 1.6 X 108 to 4.5 X 108 CFU/ml.A 0.5-ml amount of each cell suspension was injectedintraperitoneally into groups of adult (20 to 25 g) CF1mice (Carworth Farms, New City, N.Y.) immediatelyafter preparation. After challenge, the animals were

observed and deaths were recorded for a period of 1week; then the experiments were terminated.To determine the direct toxicity of the surface slime

material elaborated by the organisms grown on modi-fied 110 agar, a concentrated preparation of this ma-terial was made in the following way. Organismsgrown on modified 110 agar for 18 hr at 37 C wereharvested in 110 broth. The cells were separated bycentrifugation at 15,000 X g for 20 min, resuspendedin 20 ml of distilled water, and stirred in a WaringBlendor for 10 min at room temperature. The cellswere again separated as described above, and thesupernatant fluid was dialyzed against distilled waterand lyophilized. Several concentrations of the lyoph-ilized material, dissolvedin fresh modified 110 medium,were injected intraperitoneally into groups of normalmice.

Serum-soft agar was prepared with either BHI-ormodified 110 broth according to the method of Finkel-stein and Sulkin (4); 1% normal rabbit serum and0.15% agar (Difco) were used.

The clumping factor (CF) reaction was determinedby standard methods with undiluted normal rabbitplasma.

RESULTS

Growth characteristics of the strains ofS. aureus

in modified 110 and BHI media. All of the organ-

isms studied showed a strikingly increased mucoidor viscous type of growth on modified 110 agar ascompared with organisms grown on BHI agar.Figure 1 shows cultures of the MD strain of S.aureus on modified 110 agar and on BHI agar.The viscosity and excess slime production isevident. In modified 110 broth, the organismsalso showed a viscous type of growth, with mostof the organisms settling to the bottom of thetubes during incubation. When the tubes wereswirled, a ropy mass of organisms rose in thetubes but did not disperse. In BHI broth, therewas some settling of the organisms but they couldbe dispersed easily into a homologous cell sus-pension. Figure 2 shows a phase-contrast photo-micrograph of the MD strain grown on modified110 agar. The marked slime layer present aroundthe organisms could be observed either in moist orin dried India-ink preparations, but were moreclearly seen in the latter. Nothing even remotelyresembling this type of material could be observedwith the organisms grown on BHI agar.

All of the organisms studied, with the exceptionof the Smith diffuse strain of S. aureus, showed acharacteristic "compact" type of colonial mor-phology when grown in BHI serum-soft agar. Instriking contrast, however, the Smith compact,SK4473, and MD strains showed "diffuse" typegrowth when cultured in modified 110 serum-softagar. With the S. aureus type I, 258, 334, and 418strains, however, most of the colonies showed anintermediate type of growth, between a typicaldiffuse and typical compact colony.

Figure 3a shows the Smith compact straingrowing in BHI and 110 serum-soft agar, andFig. 3b shows the 258 strain growing under thesame conditions. When the diffuse-growingmutants from the 258 cultures were carefullyisolated in a capillary pipette and subcultured intofresh 110 serum-soft agar, all of the coloniesshowed the typical diffuse type of morphology.On the other hand, when diffuse-growing organ-isms in 110 medium were transferred to BHIserum-soft agar, all grew with the compact type ofcolonial morphology.

Virulence of organisms grown on modified 110media. A comparison was made of the mousevirulence of the staphylococcal strains studiedafter cultivation for 18 hr on BHI or modified 110agar plates. In preliminary experiments, wefound that routine washing of the mucoid organ-isms with saline removed much of the slime layermaterial and reduced the virulence of the organ-isms. Therefore, these organisms were carefullywashed as described in Materials and Methods,and were suspended in fresh 110 medium forinjection. The organisms cultured on BHI agar

903VOL. 96, 1968

on Novem

ber 9, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

904 YOSHIDA AND EKSTEDT J. BACTERIOL.

FiG. 1. MD strain of S. aureus grown on BHI or modified Staphylococcus Medium No. 110 agar. Plates wereinoculated with 3 to 5 ml of broth culture, incubated at 37 C for 18 hr, and agitated before the photograph wastaken.

FIG. 2. India ink phase-contrast photomicrograph ofMD strain ofS. aureus grown on modified StaphylococcusMedium No. 110 agar.

on Novem

ber 9, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

MUCOID GROWTH OF STAPHYLOCOCCUS AUREUS

FIG. 3. Colonial morphology of (a) Smith compact and (b) 258 strains ofS. aureus in BHI andmodifiedStaphylo-coccus Medium no. 110 serum-soft agar.

were handled in a similar manner and were sus-pended in fresh 110 medium for injection. Toconfirm the injection of comparable challengedoses, plate counts of these suspensions werealways done.The three strains of S. aureus (Smith compact,

SK4473, and MD) which could be induced togrow as typical diffuse-type colonies in modified110 serum-soft agar also showed an enhancedvirulence in mice, whereas those strains (258, 334,418, and type I) which could not be induced toexhibit typical diffuse-type growth under theseconditions did not show a significant enhancementof their virulence for mice (Table 1).

Since the organisms used in these studies weregrown in modified 110 medium and were preparedfor injection in 110 broth, it was necessary todetermine whether the media had any adverseeffect on the animals. Since the modified 110 brothcontains 3.0% NaCl, it was considered possiblethat this hypertonic salt solution might decreasethe resistance of the mice by virtue of its effect on

the viability or phagocytic activity of peritonealleukocytes. In control experiments, however, thesame organisms cultured on BHI agar and pre-pared for injection in 110 broth did not show anincrease in virulence. In addition, we found thatnormal mice injected intraperitoneally with asmuch as 25 mg of lyophilized slime materialdissolved in modified 110 broth showed no illeffects and survived indefinitely.

Intraperitoneal phagocytosis of mucoid and non-mucoid strains. Mucoid and nonmucoid organismsof the same strain were injected into separategroups of normal mice at a dose of approximately108 CFU in 0.5 ml of 110 broth. Within 5 min ofthe injection and at 3 hr, groups of three micewere sacrificed and their peritoneal cavities werewashed out with 1.0 ml of sterile saline. With amodified Maaloe procedure as described byKoenig (8), the total number of organisms re-coverable, those extracellular, and those asso-ciated with phagocytic cells were determined bydifferential centrifugation and plating. Figure 4

VOL. 96, 1968 905

on Novem

ber 9, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

906 YOSHIDA AND EKSTEDT J. BACTERIOL.

TABLE 1. Mouse virulence of S. aureus strains grown on modified Staphylococcus mediumNo. 110 or Brain Heart Infusion agar

Challenge Smith compact SK4473 MD Type I 258 334 481(CFU) 110 BHI 110 BHI 110 BHI 110 BHI 110 BHI 110 BHI 110 BHI

109 10/10a 6/10 10/10 10/10 10/10 10/10 9/10 9/10 9/10 10/10 9/10 8/10 10/10 10/10108 15/20 1/15 10/10 7/10 8/10 0/10 8/20 5/20 2/10 0/10 10/10 7/10 8/10 5/10107 1/10 0/10 6/10 0/10 2/10 0/10 0/10 0/10 0/10 0/10 0/10 0/10 0/10 0/10106 0/10 0/10 0/10 0/10 0/10 0/10 0/10 0/10

a No. dead/no. challenged.

107

6

-

------OTAL

-~_4EXTRACELLULAR

\_____ NTRACE LLULAR

s \ SOTAL

NON MUCO ID , NTRACELLULAR____MUCO ID \XTRACELLULAR

Hi4niiR AFTFP IM.lFrnTnM3

FIG. 4. Recovery of the mucoid and nonmucoid MDstrains of S. aureus from the peritoneal cavity ofnormal mice.

shows the results of a typical experiment of thiskind with the MD strain. It is clear that themucoid organisms were much less efficientlyhandled than the nonmucoid organisms underthe conditions of these experiments.

Effect of washing on virulence and reactivity in

the test for clumping factor. Table 2 shows theresults of (CF) tests with organisms grown on

modified 110 and BHI agar, and the effect ofwashing the organisms on their reaction in thesetests. In these experiments, the mucoid organismswere washed by repeatedly centrifuging themfrom homogenous saline suspensions; this pro-

cedure should be distinguished from the carefulpreparation of cell suspensions for the challengeexperiments as described in Materials andMethods. It is clear that organisms grown on

modified 110 agar react as though they are CFnegative before washing, but became CF positiveafter 1 to 2 saline washes. Table 3 shows theeffect of similar washing experiments on thevirulence of the MD and 258 strains grown on

modified 110 agar. After each wash, a sample of theorganisms was adjusted turbidimetrically to con-tain approximately 108 CFU per ml in the case of

TABLE 2. Effect of washing strainis of S. aureusgrown on modified Staphylococcus Medium No.110 or Brain Heart Infusion (BHI) agar on their

reaction in the test for clumping factor

Clumping factor reaction

Strain of ,. aureuis GrowniS Before Washed Washed thWreewash once twice timres

Smith diffuse BHI - a _110 - - - -

Smith compact BHI + + ND ND110 - - + ND

MD BHI + + ND ND110 - - + ND

258 BHI + + ND ND110 - - + ND

aClumping factor reaction absent, -; clumpingfactor reaction present, +; ND, not done.

the MD strain and 109 CFU per ml in the case ofthe less mucoid 258 strain; 0.5 ml was injectedintraperitoneally into groups of mice. It is clearthat repeated washing, presumably by removingthe slime layer material, reduced the virulence ofthe organisms. The Smith diffuse strain of S.aureus remained CF negative and virulent throughrepeated washes.

It was also shown that BHI-grown organismssuspended in a 5 %0 saline solution of crude homol-ogous slime layer material are not more virulentthan the same organisms suspended in saline.These experiments were done to investigate thepossibility that the slime layer was functioningin a manner similar to mucin (3).

DISCUSSIONFor many years, a number of investigators (12)

have engaged in the controversy regarding theexistance of true capsules in the genusStaphylococcus. It is generally conceded that theSmith strains of S. aureus, as described by Koenigand Melly (9), can be shown to possess true

on Novem

ber 9, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

MUCOID GROWTH OF STAPHYLOCOCCUS AUREUS

TABLE 3. Effect of washing the 258 andMD strainsof S. aureus grown on modified Staphylococcus

medium No. 110 agar on theirmouse virulence

Mortality after challengeTimes washed

258a strain MD' strain

None...... 16/25b 10/10Once.... 1/20 7/10Twice................. 0/20 5/10Three times 4/10

a The 258 strain was injected intraperitoneallyat a dose of approximately 109 CFU per mouse,whereas the more mucoid MD strain was injectedat a dose of approximately 108 CFU per mouse.

b No. dead/no. challenged.

capsules; recently, Wiley and Maverakis (17)were able to demonstrate a "specific capsularreaction" with these organisms in homologousantiserum. The "wound strains" of S. aureusdescribed by Wiley (16) also appear to elaboratecapsular material and have been shown to under-go a specific capsular reaction in homologousantiserum; Mudd (12) has suggested that thisphenomenon be designated the extracellularperipheral precipitation reaction.

Whereas the Wiley wound strains were isolatedafter prolonged incubation in a glycerol-brothmedium and probably represent the selection ofmutant organisms, the phenomenon described inthis paper involved the induction of mucoidgrowth in a medium containing lactose, mannitol,and NaCl and similar in its composition toStaphylococcus medium No. 110. The conversionfrom nonmucoid to mucoid organisms occurredin various degrees after the initial growth of allstrains studied in this medium. In addition, sub-culture back into BHI media resulted in a com-plete return to the nonmucoid form. Theseresults are similar to the pseudocapsulationreaction described by Sall, Mudd, and Taubler(14), except that gelatin was not required in ourmedium and at least 3% NaCl was essential.The three strains, Smith compact, SK4473, and

MD, showed the most striking alteration tomucoid growth on modified 110 agar, and thesestrains converted completely to diffuse-typegrowth in 110 serum-soft agar. On the other hand,the type 1, 258, 334, and 481 strains were lessmucoid on 110 agar and showed only a few diffusecolonies in 110 serum-soft agar; these strainswere intermediate in character or remained ascompact colonies under these conditions.

While the mucoid strains resembled the typicalSmith diffuse strain in reacting as though they

were negative for CF, growing as diffuse col-onies in 110 serum-soft agar, and showing in-creased mouse virulence, it should be madeclear that the similarity was related solely to thesynthesis of excess slime layer by these organismsin response to the nutritional conditions providedby the modified 110 medium. The slime layercould be removed by several saline washes; in sodoing, the organisms became CF positive and lessvirulent, suggesting that the slime layer merelycovered the CF on the cell surface and also madethe organisms less susceptible to phagocytosis invivo.

It was of interest to observe, however, thatdifferent strains varied in their potential to elabo-rate large amounts of slime layer under theseconditions; by utilizing the serum-soft agar tech-nique with modified 110 medium, those organismsin a mixed population which had the greatestpotential in this regard could be distinguishedand selected for.

Isolated crude slime layer material was shownto be nontoxic in milligram amounts, and pre-liminary analyses indicate that it is primarilypolysaccharide in nature. Further studies on thechemical, immunochemical, and protection-in-ducing properties of the slime layer material arein progress.

Finally, in preliminary experiments the diffuse-growing mutant organisms isolated from the 258strain by the modified 110 serum-soft agar selec-tion technique were shown to be more virulentthan organisms from the mixed culture. Furtherexperiments are in progress to confirm thisobservation and to determine whether it occurs ina group of strains recently isolated from clinicalmaterial.

ACKNOWLEDGMENTThis investigation was supported by Public Health

Service grant AI-07716-01 from the National Instituteof Allergy and Infectious Diseases.

LrrERATURE CITED

1. Cohn, Z. A. 1962. Determinants of infection inthe peritoneal cavity. I. Response to and fate ofStaphylococcus aureus and Staphylococcus albusin the mouse. Yale J. Biol. Med. 35:12-28.

2. Ekstedt, R. D. 1965. Mechanisms of resistance tostaphylococcal infection: natural and acquired.Ann. N.Y. Acad. Sci. 128:301-334.

3. Ekstedt, R. D. 1966. Studies on immunity tostaphylococcal infection in mice. IV. The roleof specific and non-specific immunity. J. Infect.Diseases 116:514-522.

4. Finkelstein, R. A., and S. E. Sulkin. 1958. Char-acteristics of coagulase positive and coagulasenegative staphylococci in serum-soft agar. J.Bacteriol. 75:339-344.

VOL. 96, 1968 907

on Novem

ber 9, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

908 YOSHIDA Al

5. Fisher, M. W., H. B. Devlin, and A. L. Erlandson.1963. A new staphylococcal antigen: its prepa-ration and immunizing activity against experi-mental infections. Nature 199:1074-1076.

6. Fisher, S. 1960. A heat stable protective staphylo-coccal antigen. Australian J. Exptl. Biol. Med.Sci. 38:479-485.

7. Hunt, G. A., and A. J. Moses. 1958. Acute infec-tion of mice with Smith strain of Staphylococcusaureus. Science 128:1574-1575.

8. Koenig, M. G. 1962. Factors relating to thevirulence of staphylococci. I. Comparative stud-ies on two colonial variants. Yale J. Biol. Med.34:537-559.

9. Koenig, M. G., and M. A. Melly. 1965. The im-portance of surface antigens in staphylococcalvirulence. Ann. N.Y. Acad. Sci. 128:231-248.

10. Koenig, M. G., M. A. Melly, and D. E. Rogers.1962. Factors relating to the virulence of staphy-lococci. III. Antibacterial versus antitoxic im-munity. J. Exptl. Med. 116:601-610.

11. Morse, S. I. 1962. Isolation and properties of asurface antigen of Staphylococcus aureus. J.Exptl. Med. 115:295-311.

N]D EKSTEDT J. BACTERIOL.

12. Mudd, S. 1965. Capsulation, pseudocapsulation,and the somatic antigens of the surface ofStaphylococcus aureus. Ann. N.Y. Acad. Sci.128:45-56.

13. Parker, S. E., P. Warner, and W. K. Slipetz. 1966.Estimation of virulence of Staphylococcusaureus for mice: the exceptional virulence ofthe Smith strain. Can. J. Microbiol. 12:753-762.

14. Sall, T., S. Mudd, and J. Taubler. 1961. Con-cerning the surfaces of cells of Staphylococcuspyogenes. 1. A pseudocapsulation phenomenonunder certain conditions. J. Exptl. Med. 113:693-700.

15. Warner, P., W. K. Slipetz, and J. Kroeker. 1966.Mouse-protective antibodies in human seraagainst the Smith strain of Staphylococcusaureus. Can. J. Microbiol. 12:949-955.

16. Wiley, B. B. 1961. A new virulence test for Staphy-lococcus aureus and its application to encapsu-lated strains. Can. J. Microbiol. 7:933-943.

17. Wiley, B. B., and N. H. Maverakis. 1968. Viru-lent and avirulent encapsulated variants ofStaphylococcus aureus. J. Bacteriol. 95:998-1002.

on Novem

ber 9, 2020 by guesthttp://jb.asm

.org/D

ownloaded from